research potential limitations

How to Write Limitations of the Study (with examples)

This blog emphasizes the importance of recognizing and effectively writing about limitations in research. It discusses the types of limitations, their significance, and provides guidelines for writing about them, highlighting their role in advancing scholarly research.

Updated on August 24, 2023

No matter how well thought out, every research endeavor encounters challenges. There is simply no way to predict all possible variances throughout the process.

These uncharted boundaries and abrupt constraints are known as limitations in research . Identifying and acknowledging limitations is crucial for conducting rigorous studies. Limitations provide context and shed light on gaps in the prevailing inquiry and literature.

This article explores the importance of recognizing limitations and discusses how to write them effectively. By interpreting limitations in research and considering prevalent examples, we aim to reframe the perception from shameful mistakes to respectable revelations.

What are limitations in research?

In the clearest terms, research limitations are the practical or theoretical shortcomings of a study that are often outside of the researcher’s control . While these weaknesses limit the generalizability of a study’s conclusions, they also present a foundation for future research.

Sometimes limitations arise from tangible circumstances like time and funding constraints, or equipment and participant availability. Other times the rationale is more obscure and buried within the research design. Common types of limitations and their ramifications include:

• Theoretical: limits the scope, depth, or applicability of a study.
• Methodological: limits the quality, quantity, or diversity of the data.
• Empirical: limits the representativeness, validity, or reliability of the data.
• Analytical: limits the accuracy, completeness, or significance of the findings.
• Ethical: limits the access, consent, or confidentiality of the data.

Regardless of how, when, or why they arise, limitations are a natural part of the research process and should never be ignored . Like all other aspects, they are vital in their own purpose.

Why is identifying limitations important?

Whether to seek acceptance or avoid struggle, humans often instinctively hide flaws and mistakes. Merging this thought process into research by attempting to hide limitations, however, is a bad idea. It has the potential to negate the validity of outcomes and damage the reputation of scholars.

By identifying and addressing limitations throughout a project, researchers strengthen their arguments and curtail the chance of peer censure based on overlooked mistakes. Pointing out these flaws shows an understanding of variable limits and a scrupulous research process.

Showing awareness of and taking responsibility for a project’s boundaries and challenges validates the integrity and transparency of a researcher. It further demonstrates the researchers understand the applicable literature and have thoroughly evaluated their chosen research methods.

Presenting limitations also benefits the readers by providing context for research findings. It guides them to interpret the project’s conclusions only within the scope of very specific conditions. By allowing for an appropriate generalization of the findings that is accurately confined by research boundaries and is not too broad, limitations boost a study’s credibility .

Limitations are true assets to the research process. They highlight opportunities for future research. When researchers identify the limitations of their particular approach to a study question, they enable precise transferability and improve chances for reproducibility. 

Simply stating a project’s limitations is not adequate for spurring further research, though. To spark the interest of other researchers, these acknowledgements must come with thorough explanations regarding how the limitations affected the current study and how they can potentially be overcome with amended methods.

How to write limitations

Typically, the information about a study’s limitations is situated either at the beginning of the discussion section to provide context for readers or at the conclusion of the discussion section to acknowledge the need for further research. However, it varies depending upon the target journal or publication guidelines. 

Don’t hide your limitations

It is also important to not bury a limitation in the body of the paper unless it has a unique connection to a topic in that section. If so, it needs to be reiterated with the other limitations or at the conclusion of the discussion section. Wherever it is included in the manuscript, ensure that the limitations section is prominently positioned and clearly introduced.

While maintaining transparency by disclosing limitations means taking a comprehensive approach, it is not necessary to discuss everything that could have potentially gone wrong during the research study. If there is no commitment to investigation in the introduction, it is unnecessary to consider the issue a limitation to the research. Wholly consider the term ‘limitations’ and ask, “Did it significantly change or limit the possible outcomes?” Then, qualify the occurrence as either a limitation to include in the current manuscript or as an idea to note for other projects. 

Writing limitations

Once the limitations are concretely identified and it is decided where they will be included in the paper, researchers are ready for the writing task. Including only what is pertinent, keeping explanations detailed but concise, and employing the following guidelines is key for crafting valuable limitations:

1) Identify and describe the limitations : Clearly introduce the limitation by classifying its form and specifying its origin. For example:

• An unintentional bias encountered during data collection
• An intentional use of unplanned post-hoc data analysis

2) Explain the implications : Describe how the limitation potentially influences the study’s findings and how the validity and generalizability are subsequently impacted. Provide examples and evidence to support claims of the limitations’ effects without making excuses or exaggerating their impact. Overall, be transparent and objective in presenting the limitations, without undermining the significance of the research. 

3) Provide alternative approaches for future studies : Offer specific suggestions for potential improvements or avenues for further investigation. Demonstrate a proactive approach by encouraging future research that addresses the identified gaps and, therefore, expands the knowledge base.

Whether presenting limitations as an individual section within the manuscript or as a subtopic in the discussion area, authors should use clear headings and straightforward language to facilitate readability. There is no need to complicate limitations with jargon, computations, or complex datasets.

Examples of common limitations

Limitations are generally grouped into two categories , methodology and research process .

Methodology limitations

Methodology may include limitations due to:

• Sample size
• Lack of available or reliable data
• Lack of prior research studies on the topic
• Measure used to collect the data
• Self-reported data

The researcher is addressing how the large sample size requires a reassessment of the measures used to collect and analyze the data.

Research process limitations

Limitations during the research process may arise from:

• Access to information
• Longitudinal effects
• Cultural and other biases
• Language fluency
• Time constraints

The author is pointing out that the model’s estimates are based on potentially biased observational studies.

Final thoughts

Successfully proving theories and touting great achievements are only two very narrow goals of scholarly research. The true passion and greatest efforts of researchers comes more in the form of confronting assumptions and exploring the obscure.

In many ways, recognizing and sharing the limitations of a research study both allows for and encourages this type of discovery that continuously pushes research forward. By using limitations to provide a transparent account of the project's boundaries and to contextualize the findings, researchers pave the way for even more robust and impactful research in the future.

Charla Viera, MS

See our "Privacy Policy"

Ensure your structure and ideas are consistent and clearly communicated

Pair your Premium Editing with our add-on service Presubmission Review for an overall assessment of your manuscript.

• Privacy Policy

Home » Limitations in Research – Types, Examples and Writing Guide

Limitations in Research – Types, Examples and Writing Guide

Table of Contents

Limitations in Research

Limitations in research refer to the factors that may affect the results, conclusions , and generalizability of a study. These limitations can arise from various sources, such as the design of the study, the sampling methods used, the measurement tools employed, and the limitations of the data analysis techniques.

Types of Limitations in Research

Types of Limitations in Research are as follows:

Sample Size Limitations

This refers to the size of the group of people or subjects that are being studied. If the sample size is too small, then the results may not be representative of the population being studied. This can lead to a lack of generalizability of the results.

Time Limitations

Time limitations can be a constraint on the research process . This could mean that the study is unable to be conducted for a long enough period of time to observe the long-term effects of an intervention, or to collect enough data to draw accurate conclusions.

Selection Bias

This refers to a type of bias that can occur when the selection of participants in a study is not random. This can lead to a biased sample that is not representative of the population being studied.

Confounding Variables

Confounding variables are factors that can influence the outcome of a study, but are not being measured or controlled for. These can lead to inaccurate conclusions or a lack of clarity in the results.

Measurement Error

This refers to inaccuracies in the measurement of variables, such as using a faulty instrument or scale. This can lead to inaccurate results or a lack of validity in the study.

Ethical Limitations

Ethical limitations refer to the ethical constraints placed on research studies. For example, certain studies may not be allowed to be conducted due to ethical concerns, such as studies that involve harm to participants.

Examples of Limitations in Research

Some Examples of Limitations in Research are as follows:

Research Title: “The Effectiveness of Machine Learning Algorithms in Predicting Customer Behavior”

Limitations:

• The study only considered a limited number of machine learning algorithms and did not explore the effectiveness of other algorithms.
• The study used a specific dataset, which may not be representative of all customer behaviors or demographics.
• The study did not consider the potential ethical implications of using machine learning algorithms in predicting customer behavior.

Research Title: “The Impact of Online Learning on Student Performance in Computer Science Courses”

• The study was conducted during the COVID-19 pandemic, which may have affected the results due to the unique circumstances of remote learning.
• The study only included students from a single university, which may limit the generalizability of the findings to other institutions.
• The study did not consider the impact of individual differences, such as prior knowledge or motivation, on student performance in online learning environments.

Research Title: “The Effect of Gamification on User Engagement in Mobile Health Applications”

• The study only tested a specific gamification strategy and did not explore the effectiveness of other gamification techniques.
• The study relied on self-reported measures of user engagement, which may be subject to social desirability bias or measurement errors.
• The study only included a specific demographic group (e.g., young adults) and may not be generalizable to other populations with different preferences or needs.

How to Write Limitations in Research

When writing about the limitations of a research study, it is important to be honest and clear about the potential weaknesses of your work. Here are some tips for writing about limitations in research:

• Identify the limitations: Start by identifying the potential limitations of your research. These may include sample size, selection bias, measurement error, or other issues that could affect the validity and reliability of your findings.
• Be honest and objective: When describing the limitations of your research, be honest and objective. Do not try to minimize or downplay the limitations, but also do not exaggerate them. Be clear and concise in your description of the limitations.
• Provide context: It is important to provide context for the limitations of your research. For example, if your sample size was small, explain why this was the case and how it may have affected your results. Providing context can help readers understand the limitations in a broader context.
• Discuss implications : Discuss the implications of the limitations for your research findings. For example, if there was a selection bias in your sample, explain how this may have affected the generalizability of your findings. This can help readers understand the limitations in terms of their impact on the overall validity of your research.
• Provide suggestions for future research : Finally, provide suggestions for future research that can address the limitations of your study. This can help readers understand how your research fits into the broader field and can provide a roadmap for future studies.

Purpose of Limitations in Research

There are several purposes of limitations in research. Here are some of the most important ones:

• To acknowledge the boundaries of the study : Limitations help to define the scope of the research project and set realistic expectations for the findings. They can help to clarify what the study is not intended to address.
• To identify potential sources of bias: Limitations can help researchers identify potential sources of bias in their research design, data collection, or analysis. This can help to improve the validity and reliability of the findings.
• To provide opportunities for future research: Limitations can highlight areas for future research and suggest avenues for further exploration. This can help to advance knowledge in a particular field.
• To demonstrate transparency and accountability: By acknowledging the limitations of their research, researchers can demonstrate transparency and accountability to their readers, peers, and funders. This can help to build trust and credibility in the research community.
• To encourage critical thinking: Limitations can encourage readers to critically evaluate the study’s findings and consider alternative explanations or interpretations. This can help to promote a more nuanced and sophisticated understanding of the topic under investigation.

When to Write Limitations in Research

Limitations should be included in research when they help to provide a more complete understanding of the study’s results and implications. A limitation is any factor that could potentially impact the accuracy, reliability, or generalizability of the study’s findings.

It is important to identify and discuss limitations in research because doing so helps to ensure that the results are interpreted appropriately and that any conclusions drawn are supported by the available evidence. Limitations can also suggest areas for future research, highlight potential biases or confounding factors that may have affected the results, and provide context for the study’s findings.

Generally, limitations should be discussed in the conclusion section of a research paper or thesis, although they may also be mentioned in other sections, such as the introduction or methods. The specific limitations that are discussed will depend on the nature of the study, the research question being investigated, and the data that was collected.

Examples of limitations that might be discussed in research include sample size limitations, data collection methods, the validity and reliability of measures used, and potential biases or confounding factors that could have affected the results. It is important to note that limitations should not be used as a justification for poor research design or methodology, but rather as a way to enhance the understanding and interpretation of the study’s findings.

Importance of Limitations in Research

Here are some reasons why limitations are important in research:

• Enhances the credibility of research: Limitations highlight the potential weaknesses and threats to validity, which helps readers to understand the scope and boundaries of the study. This improves the credibility of research by acknowledging its limitations and providing a clear picture of what can and cannot be concluded from the study.
• Facilitates replication: By highlighting the limitations, researchers can provide detailed information about the study’s methodology, data collection, and analysis. This information helps other researchers to replicate the study and test the validity of the findings, which enhances the reliability of research.
• Guides future research : Limitations provide insights into areas for future research by identifying gaps or areas that require further investigation. This can help researchers to design more comprehensive and effective studies that build on existing knowledge.
• Provides a balanced view: Limitations help to provide a balanced view of the research by highlighting both strengths and weaknesses. This ensures that readers have a clear understanding of the study’s limitations and can make informed decisions about the generalizability and applicability of the findings.

Advantages of Limitations in Research

Here are some potential advantages of limitations in research:

• Focus : Limitations can help researchers focus their study on a specific area or population, which can make the research more relevant and useful.
• Realism : Limitations can make a study more realistic by reflecting the practical constraints and challenges of conducting research in the real world.
• Innovation : Limitations can spur researchers to be more innovative and creative in their research design and methodology, as they search for ways to work around the limitations.
• Rigor : Limitations can actually increase the rigor and credibility of a study, as researchers are forced to carefully consider the potential sources of bias and error, and address them to the best of their abilities.
• Generalizability : Limitations can actually improve the generalizability of a study by ensuring that it is not overly focused on a specific sample or situation, and that the results can be applied more broadly.

About the author

Muhammad Hassan

Researcher, Academic Writer, Web developer

You may also like

How to Cite Research Paper – All Formats and...

Data Collection – Methods Types and Examples

Delimitations in Research – Types, Examples and...

Research Paper Format – Types, Examples and...

Research Process – Steps, Examples and Tips

Research Design – Types, Methods and Examples

• USC Libraries
• Research Guides

Organizing Your Social Sciences Research Paper

• Limitations of the Study
• Purpose of Guide
• Design Flaws to Avoid
• Independent and Dependent Variables
• Glossary of Research Terms
• Reading Research Effectively
• Narrowing a Topic Idea
• Broadening a Topic Idea
• Extending the Timeliness of a Topic Idea
• Academic Writing Style
• Applying Critical Thinking
• Choosing a Title
• Making an Outline
• Paragraph Development
• Research Process Video Series
• Executive Summary
• The C.A.R.S. Model
• Background Information
• The Research Problem/Question
• Theoretical Framework
• Citation Tracking
• Content Alert Services
• Evaluating Sources
• Primary Sources
• Secondary Sources
• Tiertiary Sources
• Scholarly vs. Popular Publications
• Qualitative Methods
• Quantitative Methods
• Insiderness
• Using Non-Textual Elements
• Common Grammar Mistakes
• Writing Concisely
• Avoiding Plagiarism
• Footnotes or Endnotes?
• Further Readings
• Generative AI and Writing
• USC Libraries Tutorials and Other Guides
• Bibliography

The limitations of the study are those characteristics of design or methodology that impacted or influenced the interpretation of the findings from your research. Study limitations are the constraints placed on the ability to generalize from the results, to further describe applications to practice, and/or related to the utility of findings that are the result of the ways in which you initially chose to design the study or the method used to establish internal and external validity or the result of unanticipated challenges that emerged during the study.

Price, James H. and Judy Murnan. “Research Limitations and the Necessity of Reporting Them.” American Journal of Health Education 35 (2004): 66-67; Theofanidis, Dimitrios and Antigoni Fountouki. "Limitations and Delimitations in the Research Process." Perioperative Nursing 7 (September-December 2018): 155-163. .

Importance of...

Always acknowledge a study's limitations. It is far better that you identify and acknowledge your study’s limitations than to have them pointed out by your professor and have your grade lowered because you appeared to have ignored them or didn't realize they existed.

Keep in mind that acknowledgment of a study's limitations is an opportunity to make suggestions for further research. If you do connect your study's limitations to suggestions for further research, be sure to explain the ways in which these unanswered questions may become more focused because of your study.

Acknowledgment of a study's limitations also provides you with opportunities to demonstrate that you have thought critically about the research problem, understood the relevant literature published about it, and correctly assessed the methods chosen for studying the problem. A key objective of the research process is not only discovering new knowledge but also to confront assumptions and explore what we don't know.

Claiming limitations is a subjective process because you must evaluate the impact of those limitations . Don't just list key weaknesses and the magnitude of a study's limitations. To do so diminishes the validity of your research because it leaves the reader wondering whether, or in what ways, limitation(s) in your study may have impacted the results and conclusions. Limitations require a critical, overall appraisal and interpretation of their impact. You should answer the question: do these problems with errors, methods, validity, etc. eventually matter and, if so, to what extent?

Price, James H. and Judy Murnan. “Research Limitations and the Necessity of Reporting Them.” American Journal of Health Education 35 (2004): 66-67; Structure: How to Structure the Research Limitations Section of Your Dissertation. Dissertations and Theses: An Online Textbook. Laerd.com.

Descriptions of Possible Limitations

All studies have limitations . However, it is important that you restrict your discussion to limitations related to the research problem under investigation. For example, if a meta-analysis of existing literature is not a stated purpose of your research, it should not be discussed as a limitation. Do not apologize for not addressing issues that you did not promise to investigate in the introduction of your paper.

Here are examples of limitations related to methodology and the research process you may need to describe and discuss how they possibly impacted your results. Note that descriptions of limitations should be stated in the past tense because they were discovered after you completed your research.

Possible Methodological Limitations

• Sample size -- the number of the units of analysis you use in your study is dictated by the type of research problem you are investigating. Note that, if your sample size is too small, it will be difficult to find significant relationships from the data, as statistical tests normally require a larger sample size to ensure a representative distribution of the population and to be considered representative of groups of people to whom results will be generalized or transferred. Note that sample size is generally less relevant in qualitative research if explained in the context of the research problem.
• Lack of available and/or reliable data -- a lack of data or of reliable data will likely require you to limit the scope of your analysis, the size of your sample, or it can be a significant obstacle in finding a trend and a meaningful relationship. You need to not only describe these limitations but provide cogent reasons why you believe data is missing or is unreliable. However, don’t just throw up your hands in frustration; use this as an opportunity to describe a need for future research based on designing a different method for gathering data.
• Lack of prior research studies on the topic -- citing prior research studies forms the basis of your literature review and helps lay a foundation for understanding the research problem you are investigating. Depending on the currency or scope of your research topic, there may be little, if any, prior research on your topic. Before assuming this to be true, though, consult with a librarian! In cases when a librarian has confirmed that there is little or no prior research, you may be required to develop an entirely new research typology [for example, using an exploratory rather than an explanatory research design ]. Note again that discovering a limitation can serve as an important opportunity to identify new gaps in the literature and to describe the need for further research.
• Measure used to collect the data -- sometimes it is the case that, after completing your interpretation of the findings, you discover that the way in which you gathered data inhibited your ability to conduct a thorough analysis of the results. For example, you regret not including a specific question in a survey that, in retrospect, could have helped address a particular issue that emerged later in the study. Acknowledge the deficiency by stating a need for future researchers to revise the specific method for gathering data.
• Self-reported data -- whether you are relying on pre-existing data or you are conducting a qualitative research study and gathering the data yourself, self-reported data is limited by the fact that it rarely can be independently verified. In other words, you have to the accuracy of what people say, whether in interviews, focus groups, or on questionnaires, at face value. However, self-reported data can contain several potential sources of bias that you should be alert to and note as limitations. These biases become apparent if they are incongruent with data from other sources. These are: (1) selective memory [remembering or not remembering experiences or events that occurred at some point in the past]; (2) telescoping [recalling events that occurred at one time as if they occurred at another time]; (3) attribution [the act of attributing positive events and outcomes to one's own agency, but attributing negative events and outcomes to external forces]; and, (4) exaggeration [the act of representing outcomes or embellishing events as more significant than is actually suggested from other data].

Possible Limitations of the Researcher

• Access -- if your study depends on having access to people, organizations, data, or documents and, for whatever reason, access is denied or limited in some way, the reasons for this needs to be described. Also, include an explanation why being denied or limited access did not prevent you from following through on your study.
• Longitudinal effects -- unlike your professor, who can literally devote years [even a lifetime] to studying a single topic, the time available to investigate a research problem and to measure change or stability over time is constrained by the due date of your assignment. Be sure to choose a research problem that does not require an excessive amount of time to complete the literature review, apply the methodology, and gather and interpret the results. If you're unsure whether you can complete your research within the confines of the assignment's due date, talk to your professor.
• Cultural and other type of bias -- we all have biases, whether we are conscience of them or not. Bias is when a person, place, event, or thing is viewed or shown in a consistently inaccurate way. Bias is usually negative, though one can have a positive bias as well, especially if that bias reflects your reliance on research that only support your hypothesis. When proof-reading your paper, be especially critical in reviewing how you have stated a problem, selected the data to be studied, what may have been omitted, the manner in which you have ordered events, people, or places, how you have chosen to represent a person, place, or thing, to name a phenomenon, or to use possible words with a positive or negative connotation. NOTE :   If you detect bias in prior research, it must be acknowledged and you should explain what measures were taken to avoid perpetuating that bias. For example, if a previous study only used boys to examine how music education supports effective math skills, describe how your research expands the study to include girls.
• Fluency in a language -- if your research focuses , for example, on measuring the perceived value of after-school tutoring among Mexican-American ESL [English as a Second Language] students and you are not fluent in Spanish, you are limited in being able to read and interpret Spanish language research studies on the topic or to speak with these students in their primary language. This deficiency should be acknowledged.

Aguinis, Hermam and Jeffrey R. Edwards. “Methodological Wishes for the Next Decade and How to Make Wishes Come True.” Journal of Management Studies 51 (January 2014): 143-174; Brutus, Stéphane et al. "Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations." Journal of Management 39 (January 2013): 48-75; Senunyeme, Emmanuel K. Business Research Methods. Powerpoint Presentation. Regent University of Science and Technology; ter Riet, Gerben et al. “All That Glitters Isn't Gold: A Survey on Acknowledgment of Limitations in Biomedical Studies.” PLOS One 8 (November 2013): 1-6.

Structure and Writing Style

Information about the limitations of your study are generally placed either at the beginning of the discussion section of your paper so the reader knows and understands the limitations before reading the rest of your analysis of the findings, or, the limitations are outlined at the conclusion of the discussion section as an acknowledgement of the need for further study. Statements about a study's limitations should not be buried in the body [middle] of the discussion section unless a limitation is specific to something covered in that part of the paper. If this is the case, though, the limitation should be reiterated at the conclusion of the section.

If you determine that your study is seriously flawed due to important limitations , such as, an inability to acquire critical data, consider reframing it as an exploratory study intended to lay the groundwork for a more complete research study in the future. Be sure, though, to specifically explain the ways that these flaws can be successfully overcome in a new study.

But, do not use this as an excuse for not developing a thorough research paper! Review the tab in this guide for developing a research topic . If serious limitations exist, it generally indicates a likelihood that your research problem is too narrowly defined or that the issue or event under study is too recent and, thus, very little research has been written about it. If serious limitations do emerge, consult with your professor about possible ways to overcome them or how to revise your study.

When discussing the limitations of your research, be sure to:

• Describe each limitation in detailed but concise terms;
• Explain why each limitation exists;
• Provide the reasons why each limitation could not be overcome using the method(s) chosen to acquire or gather the data [cite to other studies that had similar problems when possible];
• Assess the impact of each limitation in relation to the overall findings and conclusions of your study; and,
• If appropriate, describe how these limitations could point to the need for further research.

Remember that the method you chose may be the source of a significant limitation that has emerged during your interpretation of the results [for example, you didn't interview a group of people that you later wish you had]. If this is the case, don't panic. Acknowledge it, and explain how applying a different or more robust methodology might address the research problem more effectively in a future study. A underlying goal of scholarly research is not only to show what works, but to demonstrate what doesn't work or what needs further clarification.

Aguinis, Hermam and Jeffrey R. Edwards. “Methodological Wishes for the Next Decade and How to Make Wishes Come True.” Journal of Management Studies 51 (January 2014): 143-174; Brutus, Stéphane et al. "Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations." Journal of Management 39 (January 2013): 48-75; Ioannidis, John P.A. "Limitations are not Properly Acknowledged in the Scientific Literature." Journal of Clinical Epidemiology 60 (2007): 324-329; Pasek, Josh. Writing the Empirical Social Science Research Paper: A Guide for the Perplexed. January 24, 2012. Academia.edu; Structure: How to Structure the Research Limitations Section of Your Dissertation. Dissertations and Theses: An Online Textbook. Laerd.com; What Is an Academic Paper? Institute for Writing Rhetoric. Dartmouth College; Writing the Experimental Report: Methods, Results, and Discussion. The Writing Lab and The OWL. Purdue University.

Writing Tip

Don't Inflate the Importance of Your Findings!

After all the hard work and long hours devoted to writing your research paper, it is easy to get carried away with attributing unwarranted importance to what you’ve done. We all want our academic work to be viewed as excellent and worthy of a good grade, but it is important that you understand and openly acknowledge the limitations of your study. Inflating the importance of your study's findings could be perceived by your readers as an attempt hide its flaws or encourage a biased interpretation of the results. A small measure of humility goes a long way!

Another Writing Tip

Negative Results are Not a Limitation!

Negative evidence refers to findings that unexpectedly challenge rather than support your hypothesis. If you didn't get the results you anticipated, it may mean your hypothesis was incorrect and needs to be reformulated. Or, perhaps you have stumbled onto something unexpected that warrants further study. Moreover, the absence of an effect may be very telling in many situations, particularly in experimental research designs. In any case, your results may very well be of importance to others even though they did not support your hypothesis. Do not fall into the trap of thinking that results contrary to what you expected is a limitation to your study. If you carried out the research well, they are simply your results and only require additional interpretation.

Lewis, George H. and Jonathan F. Lewis. “The Dog in the Night-Time: Negative Evidence in Social Research.” The British Journal of Sociology 31 (December 1980): 544-558.

Yet Another Writing Tip

Sample Size Limitations in Qualitative Research

Sample sizes are typically smaller in qualitative research because, as the study goes on, acquiring more data does not necessarily lead to more information. This is because one occurrence of a piece of data, or a code, is all that is necessary to ensure that it becomes part of the analysis framework. However, it remains true that sample sizes that are too small cannot adequately support claims of having achieved valid conclusions and sample sizes that are too large do not permit the deep, naturalistic, and inductive analysis that defines qualitative inquiry. Determining adequate sample size in qualitative research is ultimately a matter of judgment and experience in evaluating the quality of the information collected against the uses to which it will be applied and the particular research method and purposeful sampling strategy employed. If the sample size is found to be a limitation, it may reflect your judgment about the methodological technique chosen [e.g., single life history study versus focus group interviews] rather than the number of respondents used.

Boddy, Clive Roland. "Sample Size for Qualitative Research." Qualitative Market Research: An International Journal 19 (2016): 426-432; Huberman, A. Michael and Matthew B. Miles. "Data Management and Analysis Methods." In Handbook of Qualitative Research . Norman K. Denzin and Yvonna S. Lincoln, eds. (Thousand Oaks, CA: Sage, 1994), pp. 428-444; Blaikie, Norman. "Confounding Issues Related to Determining Sample Size in Qualitative Research." International Journal of Social Research Methodology 21 (2018): 635-641; Oppong, Steward Harrison. "The Problem of Sampling in qualitative Research." Asian Journal of Management Sciences and Education 2 (2013): 202-210.

• << Previous: 8. The Discussion
• Next: 9. The Conclusion >>
• Last Updated: May 25, 2024 4:09 PM
• URL: https://libguides.usc.edu/writingguide

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Perspect Med Educ
• v.8(4); 2019 Aug

Limited by our limitations

Paula t. ross.

Medical School, University of Michigan, Ann Arbor, MI USA

Nikki L. Bibler Zaidi

Study limitations represent weaknesses within a research design that may influence outcomes and conclusions of the research. Researchers have an obligation to the academic community to present complete and honest limitations of a presented study. Too often, authors use generic descriptions to describe study limitations. Including redundant or irrelevant limitations is an ineffective use of the already limited word count. A meaningful presentation of study limitations should describe the potential limitation, explain the implication of the limitation, provide possible alternative approaches, and describe steps taken to mitigate the limitation. This includes placing research findings within their proper context to ensure readers do not overemphasize or minimize findings. A more complete presentation will enrich the readers’ understanding of the study’s limitations and support future investigation.

Introduction

Regardless of the format scholarship assumes, from qualitative research to clinical trials, all studies have limitations. Limitations represent weaknesses within the study that may influence outcomes and conclusions of the research. The goal of presenting limitations is to provide meaningful information to the reader; however, too often, limitations in medical education articles are overlooked or reduced to simplistic and minimally relevant themes (e.g., single institution study, use of self-reported data, or small sample size) [ 1 ]. This issue is prominent in other fields of inquiry in medicine as well. For example, despite the clinical implications, medical studies often fail to discuss how limitations could have affected the study findings and interpretations [ 2 ]. Further, observational research often fails to remind readers of the fundamental limitation inherent in the study design, which is the inability to attribute causation [ 3 ]. By reporting generic limitations or omitting them altogether, researchers miss opportunities to fully communicate the relevance of their work, illustrate how their work advances a larger field under study, and suggest potential areas for further investigation.

Goals of presenting limitations

Medical education scholarship should provide empirical evidence that deepens our knowledge and understanding of education [ 4 , 5 ], informs educational practice and process, [ 6 , 7 ] and serves as a forum for educating other researchers [ 8 ]. Providing study limitations is indeed an important part of this scholarly process. Without them, research consumers are pressed to fully grasp the potential exclusion areas or other biases that may affect the results and conclusions provided [ 9 ]. Study limitations should leave the reader thinking about opportunities to engage in prospective improvements [ 9 – 11 ] by presenting gaps in the current research and extant literature, thereby cultivating other researchers’ curiosity and interest in expanding the line of scholarly inquiry [ 9 ].

Presenting study limitations is also an ethical element of scientific inquiry [ 12 ]. It ensures transparency of both the research and the researchers [ 10 , 13 , 14 ], as well as provides transferability [ 15 ] and reproducibility of methods. Presenting limitations also supports proper interpretation and validity of the findings [ 16 ]. A study’s limitations should place research findings within their proper context to ensure readers are fully able to discern the credibility of a study’s conclusion, and can generalize findings appropriately [ 16 ].

Why some authors may fail to present limitations

As Price and Murnan [ 8 ] note, there may be overriding reasons why researchers do not sufficiently report the limitations of their study. For example, authors may not fully understand the importance and implications of their study’s limitations or assume that not discussing them may increase the likelihood of publication. Word limits imposed by journals may also prevent authors from providing thorough descriptions of their study’s limitations [ 17 ]. Still another possible reason for excluding limitations is a diffusion of responsibility in which some authors may incorrectly assume that the journal editor is responsible for identifying limitations. Regardless of reason or intent, researchers have an obligation to the academic community to present complete and honest study limitations.

A guide to presenting limitations

The presentation of limitations should describe the potential limitations, explain the implication of the limitations, provide possible alternative approaches, and describe steps taken to mitigate the limitations. Too often, authors only list the potential limitations, without including these other important elements.

Describe the limitations

When describing limitations authors should identify the limitation type to clearly introduce the limitation and specify the origin of the limitation. This helps to ensure readers are able to interpret and generalize findings appropriately. Here we outline various limitation types that can occur at different stages of the research process.

Study design

Some study limitations originate from conscious choices made by the researcher (also known as delimitations) to narrow the scope of the study [ 1 , 8 , 18 ]. For example, the researcher may have designed the study for a particular age group, sex, race, ethnicity, geographically defined region, or some other attribute that would limit to whom the findings can be generalized. Such delimitations involve conscious exclusionary and inclusionary decisions made during the development of the study plan, which may represent a systematic bias intentionally introduced into the study design or instrument by the researcher [ 8 ]. The clear description and delineation of delimitations and limitations will assist editors and reviewers in understanding any methodological issues.

Data collection

Study limitations can also be introduced during data collection. An unintentional consequence of human subjects research is the potential of the researcher to influence how participants respond to their questions. Even when appropriate methods for sampling have been employed, some studies remain limited by the use of data collected only from participants who decided to enrol in the study (self-selection bias) [ 11 , 19 ]. In some cases, participants may provide biased input by responding to questions they believe are favourable to the researcher rather than their authentic response (social desirability bias) [ 20 – 22 ]. Participants may influence the data collected by changing their behaviour when they are knowingly being observed (Hawthorne effect) [ 23 ]. Researchers—in their role as an observer—may also bias the data they collect by allowing a first impression of the participant to be influenced by a single characteristic or impression of another characteristic either unfavourably (horns effect) or favourably (halo effort) [ 24 ].

Data analysis

Study limitations may arise as a consequence of the type of statistical analysis performed. Some studies may not follow the basic tenets of inferential statistical analyses when they use convenience sampling (i.e. non-probability sampling) rather than employing probability sampling from a target population [ 19 ]. Another limitation that can arise during statistical analyses occurs when studies employ unplanned post-hoc data analyses that were not specified before the initial analysis [ 25 ]. Unplanned post-hoc analysis may lead to statistical relationships that suggest associations but are no more than coincidental findings [ 23 ]. Therefore, when unplanned post-hoc analyses are conducted, this should be clearly stated to allow the reader to make proper interpretation and conclusions—especially when only a subset of the original sample is investigated [ 23 ].

Study results

The limitations of any research study will be rooted in the validity of its results—specifically threats to internal or external validity [ 8 ]. Internal validity refers to reliability or accuracy of the study results [ 26 ], while external validity pertains to the generalizability of results from the study’s sample to the larger, target population [ 8 ].

Examples of threats to internal validity include: effects of events external to the study (history), changes in participants due to time instead of the studied effect (maturation), systematic reduction in participants related to a feature of the study (attrition), changes in participant responses due to repeatedly measuring participants (testing effect), modifications to the instrument (instrumentality) and selecting participants based on extreme scores that will regress towards the mean in repeat tests (regression to the mean) [ 27 ].

Threats to external validity include factors that might inhibit generalizability of results from the study’s sample to the larger, target population [ 8 , 27 ]. External validity is challenged when results from a study cannot be generalized to its larger population or to similar populations in terms of the context, setting, participants and time [ 18 ]. Therefore, limitations should be made transparent in the results to inform research consumers of any known or potentially hidden biases that may have affected the study and prevent generalization beyond the study parameters.

Explain the implication(s) of each limitation

Authors should include the potential impact of the limitations (e.g., likelihood, magnitude) [ 13 ] as well as address specific validity implications of the results and subsequent conclusions [ 16 , 28 ]. For example, self-reported data may lead to inaccuracies (e.g. due to social desirability bias) which threatens internal validity [ 19 ]. Even a researcher’s inappropriate attribution to a characteristic or outcome (e.g., stereotyping) can overemphasize (either positively or negatively) unrelated characteristics or outcomes (halo or horns effect) and impact the internal validity [ 24 ]. Participants’ awareness that they are part of a research study can also influence outcomes (Hawthorne effect) and limit external validity of findings [ 23 ]. External validity may also be threatened should the respondents’ propensity for participation be correlated with the substantive topic of study, as data will be biased and not represent the population of interest (self-selection bias) [ 29 ]. Having this explanation helps readers interpret the results and generalize the applicability of the results for their own setting.

Provide potential alternative approaches and explanations

Often, researchers use other studies’ limitations as the first step in formulating new research questions and shaping the next phase of research. Therefore, it is important for readers to understand why potential alternative approaches (e.g. approaches taken by others exploring similar topics) were not taken. In addition to alternative approaches, authors can also present alternative explanations for their own study’s findings [ 13 ]. This information is valuable coming from the researcher because of the direct, relevant experience and insight gained as they conducted the study. The presentation of alternative approaches represents a major contribution to the scholarly community.

Describe steps taken to minimize each limitation

No research design is perfect and free from explicit and implicit biases; however various methods can be employed to minimize the impact of study limitations. Some suggested steps to mitigate or minimize the limitations mentioned above include using neutral questions, randomized response technique, force choice items, or self-administered questionnaires to reduce respondents’ discomfort when answering sensitive questions (social desirability bias) [ 21 ]; using unobtrusive data collection measures (e.g., use of secondary data) that do not require the researcher to be present (Hawthorne effect) [ 11 , 30 ]; using standardized rubrics and objective assessment forms with clearly defined scoring instructions to minimize researcher bias, or making rater adjustments to assessment scores to account for rater tendencies (halo or horns effect) [ 24 ]; or using existing data or control groups (self-selection bias) [ 11 , 30 ]. When appropriate, researchers should provide sufficient evidence that demonstrates the steps taken to mitigate limitations as part of their study design [ 13 ].

In conclusion, authors may be limiting the impact of their research by neglecting or providing abbreviated and generic limitations. We present several examples of limitations to consider; however, this should not be considered an exhaustive list nor should these examples be added to the growing list of generic and overused limitations. Instead, careful thought should go into presenting limitations after research has concluded and the major findings have been described. Limitations help focus the reader on key findings, therefore it is important to only address the most salient limitations of the study [ 17 , 28 ] related to the specific research problem, not general limitations of most studies [ 1 ]. It is important not to minimize the limitations of study design or results. Rather, results, including their limitations, must help readers draw connections between current research and the extant literature.

The quality and rigor of our research is largely defined by our limitations [ 31 ]. In fact, one of the top reasons reviewers report recommending acceptance of medical education research manuscripts involves limitations—specifically how the study’s interpretation accounts for its limitations [ 32 ]. Therefore, it is not only best for authors to acknowledge their study’s limitations rather than to have them identified by an editor or reviewer, but proper framing and presentation of limitations can actually increase the likelihood of acceptance. Perhaps, these issues could be ameliorated if academic and research organizations adopted policies and/or expectations to guide authors in proper description of limitations.

What are the limitations in research and how to write them?

Learn about the potential limitations in research and how to appropriately address them in order to deliver honest and ethical research.

It is fairly uncommon for researchers to stumble into the term research limitations when working on their research paper. Limitations in research can arise owing to constraints on design, methods, materials, and so on, and these aspects, unfortunately, may have an influence on your subject’s findings.

In this Mind The Graph’s article, we’ll discuss some recommendations for writing limitations in research , provide examples of various common types of limitations, and suggest how to properly present this information.

What are the limitations in research?

The limitations in research are the constraints in design, methods or even researchers’ limitations that affect and influence the interpretation of your research’s ultimate findings. These are limitations on the generalization and usability of findings that emerge from the design of the research and/or the method employed to ensure validity both internally and externally. 

Researchers are usually cautious to acknowledge the limitations of their research in their publications for fear of undermining the research’s scientific validity. No research is faultless or covers every possible angle. As a result, addressing the constraints of your research exhibits honesty and integrity .

Why should include limitations of research in my paper?

Though limitations tackle potential flaws in research, commenting on them at the conclusion of your paper, by demonstrating that you are aware of these limitations and explaining how they impact the conclusions that may be taken from the research, improves your research by disclosing any issues before other researchers or reviewers do . 

Additionally, emphasizing research constraints implies that you have thoroughly investigated the ramifications of research shortcomings and have a thorough understanding of your research problem. 

Limits exist in any research; being honest about them and explaining them would impress researchers and reviewers more than disregarding them. 

Remember that acknowledging a research’s shortcomings offers a chance to provide ideas for future research, but be careful to describe how your study may help to concentrate on these outstanding problems.

Possible limitations examples

Here are some limitations connected to methodology and the research procedure that you may need to explain and discuss in connection to your findings.

Methodological limitations

Sample size.

The number of units of analysis used in your study is determined by the sort of research issue being investigated. It is important to note that if your sample is too small, finding significant connections in the data will be challenging, as statistical tests typically require a larger sample size to ensure a fair representation and this can be limiting. 

Lack of available or reliable data

A lack of data or trustworthy data will almost certainly necessitate limiting the scope of your research or the size of your sample, or it can be a substantial impediment to identifying a pattern and a relevant connection.

Lack of prior research on the subject

Citing previous research papers forms the basis of your literature review and aids in comprehending the research subject you are researching. Yet there may be little if any, past research on your issue.

The measure used to collect data

After finishing your analysis of the findings, you realize that the method you used to collect data limited your capacity to undertake a comprehensive evaluation of the findings. Recognize the flaw by mentioning that future researchers should change the specific approach for data collection.

Issues with research samples and selection

Sampling inaccuracies arise when a probability sampling method is employed to choose a sample, but that sample does not accurately represent the overall population or the relevant group. As a result, your study suffers from “sampling bias” or “selection bias.”

Limitations of the research

When your research requires polling certain persons or a specific group, you may have encountered the issue of limited access to these interviewees. Because of the limited access, you may need to reorganize or rearrange your research. In this scenario, explain why access is restricted and ensure that your findings are still trustworthy and valid despite the constraint.

Time constraints

Practical difficulties may limit the amount of time available to explore a research issue and monitor changes as they occur. If time restrictions have any detrimental influence on your research, recognize this impact by expressing the necessity for a future investigation.

Due to their cultural origins or opinions on observed events, researchers may carry biased opinions, which can influence the credibility of a research. Furthermore, researchers may exhibit biases toward data and conclusions that only support their hypotheses or arguments.

The structure of the limitations section 

The limitations of your research are usually stated at the beginning of the discussion section of your paper so that the reader is aware of and comprehends the limitations prior to actually reading the rest of your findings, or they are stated at the end of the discussion section as an acknowledgment of the need for further research.

The ideal way is to divide your limitations section into three steps: 

1. Identify the research constraints; 

2. Describe in great detail how they affect your research; 

3. Mention the opportunity for future investigations and give possibilities. 

By following this method while addressing the constraints of your research, you will be able to effectively highlight your research’s shortcomings without jeopardizing the quality and integrity of your research.

Present your research or paper in an innovative way

If you want your readers to be engaged and participate in your research, try Mind The Graph tool to add visual assets to your content. Infographics may improve comprehension and are easy to read, just as the Mind The Graph tool is simple to use and offers a variety of templates from which you can select the one that best suits your information.

Subscribe to our newsletter

Exclusive high quality content about effective visual communication in science.

Unlock Your Creativity

Create infographics, presentations and other scientifically-accurate designs without hassle — absolutely free for 7 days!

About Jessica Abbadia

Jessica Abbadia is a lawyer that has been working in Digital Marketing since 2020, improving organic performance for apps and websites in various regions through ASO and SEO. Currently developing scientific and intellectual knowledge for the community's benefit. Jessica is an animal rights activist who enjoys reading and drinking strong coffee.

Content tags

Writing Limitations of Research Study — 4 Reasons Why It Is Important!

It is not unusual for researchers to come across the term limitations of research during their academic paper writing. More often this is interpreted as something terrible. However, when it comes to research study, limitations can help structure the research study better. Therefore, do not underestimate significance of limitations of research study.

Allow us to take you through the context of how to evaluate the limits of your research and conclude an impactful relevance to your results.

Table of Contents

What Are the Limitations of a Research Study?

Every research has its limit and these limitations arise due to restrictions in methodology or research design.  This could impact your entire research or the research paper you wish to publish. Unfortunately, most researchers choose not to discuss their limitations of research fearing it will affect the value of their article in the eyes of readers.

However, it is very important to discuss your study limitations and show it to your target audience (other researchers, journal editors, peer reviewers etc.). It is very important that you provide an explanation of how your research limitations may affect the conclusions and opinions drawn from your research. Moreover, when as an author you state the limitations of research, it shows that you have investigated all the weaknesses of your study and have a deep understanding of the subject. Being honest could impress your readers and mark your study as a sincere effort in research.

Why and Where Should You Include the Research Limitations?

The main goal of your research is to address your research objectives. Conduct experiments, get results and explain those results, and finally justify your research question . It is best to mention the limitations of research in the discussion paragraph of your research article.

At the very beginning of this paragraph, immediately after highlighting the strengths of the research methodology, you should write down your limitations. You can discuss specific points from your research limitations as suggestions for further research in the conclusion of your thesis.

1. Common Limitations of the Researchers

Limitations that are related to the researcher must be mentioned. This will help you gain transparency with your readers. Furthermore, you could provide suggestions on decreasing these limitations in you and your future studies.

2. Limited Access to Information

Your work may involve some institutions and individuals in research, and sometimes you may have problems accessing these institutions. Therefore, you need to redesign and rewrite your work. You must explain your readers the reason for limited access.

3. Limited Time

All researchers are bound by their deadlines when it comes to completing their studies. Sometimes, time constraints can affect your research negatively. However, the best practice is to acknowledge it and mention a requirement for future study to solve the research problem in a better way.

4. Conflict over Biased Views and Personal Issues

Biased views can affect the research. In fact, researchers end up choosing only those results and data that support their main argument, keeping aside the other loose ends of the research.

Types of Limitations of Research

Before beginning your research study, know that there are certain limitations to what you are testing or possible research results. There are different types that researchers may encounter, and they all have unique characteristics, such as:

1. Research Design Limitations

Certain restrictions on your research or available procedures may affect your final results or research outputs. You may have formulated research goals and objectives too broadly. However, this can help you understand how you can narrow down the formulation of research goals and objectives, thereby increasing the focus of your study.

2. Impact Limitations

Even if your research has excellent statistics and a strong design, it can suffer from the influence of the following factors:

• Presence of increasing findings as researched
• Being population specific
• A strong regional focus.

3. Data or statistical limitations

In some cases, it is impossible to collect sufficient data for research or very difficult to get access to the data. This could lead to incomplete conclusion to your study. Moreover, this insufficiency in data could be the outcome of your study design. The unclear, shabby research outline could produce more problems in interpreting your findings.

How to Correctly Structure Your Research Limitations?

There are strict guidelines for narrowing down research questions, wherein you could justify and explain potential weaknesses of your academic paper. You could go through these basic steps to get a well-structured clarity of research limitations:

• Declare that you wish to identify your limitations of research and explain their importance,
• Provide the necessary depth, explain their nature, and justify your study choices.
• Write how you are suggesting that it is possible to overcome them in the future.

In this section, your readers will see that you are aware of the potential weaknesses in your business, understand them and offer effective solutions, and it will positively strengthen your article as you clarify all limitations of research to your target audience.

Know that you cannot be perfect and there is no individual without flaws. You could use the limitations of research as a great opportunity to take on a new challenge and improve the future of research. In a typical academic paper, research limitations may relate to:

1. Formulating your goals and objectives

If you formulate goals and objectives too broadly, your work will have some shortcomings. In this case, specify effective methods or ways to narrow down the formula of goals and aim to increase your level of study focus.

2. Application of your data collection methods in research

If you do not have experience in primary data collection, there is a risk that there will be flaws in the implementation of your methods. It is necessary to accept this, and learn and educate yourself to understand data collection methods.

3. Sample sizes

This depends on the nature of problem you choose. Sample size is of a greater importance in quantitative studies as opposed to qualitative ones. If your sample size is too small, statistical tests cannot identify significant relationships or connections within a given data set.

You could point out that other researchers should base the same study on a larger sample size to get more accurate results.

4. The absence of previous studies in the field you have chosen

Writing a literature review is an important step in any scientific study because it helps researchers determine the scope of current work in the chosen field. It is a major foundation for any researcher who must use them to achieve a set of specific goals or objectives.

However, if you are focused on the most current and evolving research problem or a very narrow research problem, there may be very little prior research on your topic. For example, if you chose to explore the role of Bitcoin as the currency of the future, you may not find tons of scientific papers addressing the research problem as Bitcoins are only a new phenomenon.

It is important that you learn to identify research limitations examples at each step. Whatever field you choose, feel free to add the shortcoming of your work. This is mainly because you do not have many years of experience writing scientific papers or completing complex work. Therefore, the depth and scope of your discussions may be compromised at different levels compared to academics with a lot of expertise. Include specific points from limitations of research. Use them as suggestions for the future.

Have you ever faced a challenge of writing the limitations of research study in your paper? How did you overcome it? What ways did you follow? Were they beneficial? Let us know in the comments below!

Frequently Asked Questions

Setting limitations in our study helps to clarify the outcomes drawn from our research and enhance understanding of the subject. Moreover, it shows that the author has investigated all the weaknesses in the study.

Scope is the range and limitations of a research project which are set to define the boundaries of a project. Limitations are the impacts on the overall study due to the constraints on the research design.

Limitation in research is an impact of a constraint on the research design in the overall study. They are the flaws or weaknesses in the study, which may influence the outcome of the research.

1. Limitations in research can be written as follows: Formulate your goals and objectives 2. Analyze the chosen data collection method and the sample sizes 3. Identify your limitations of research and explain their importance 4. Provide the necessary depth, explain their nature, and justify your study choices 5. Write how you are suggesting that it is possible to overcome them in the future

Excellent article ,,,it has helped me big

This is very helpful information. It has given me an insight on how to go about my study limitations.

Good comments and helpful

Rate this article Cancel Reply

Your email address will not be published.

Enago Academy's Most Popular Articles

• Diversity and Inclusion
• Trending Now

The Silent Struggle: Confronting gender bias in science funding

In the 1990s, Dr. Katalin Kariko’s pioneering mRNA research seemed destined for obscurity, doomed by…

• Promoting Research

Plain Language Summary — Communicating your research to bridge the academic-lay gap

Science can be complex, but does that mean it should not be accessible to the…

Addressing Barriers in Academia: Navigating unconscious biases in the Ph.D. journey

In the journey of academia, a Ph.D. marks a transitional phase, like that of a…

• Manuscripts & Grants
• Reporting Research

Unraveling Research Population and Sample: Understanding their role in statistical inference

Research population and sample serve as the cornerstones of any scientific inquiry. They hold the…

• Manuscript Preparation
• Publishing Research

Research Problem Statement — Find out how to write an impactful one!

What Is a Research Problem Statement? A research problem statement is a clear, concise, and…

How to Develop a Good Research Question? — Types & Examples

5 Effective Ways to Avoid Ghostwriting for Busy Researchers

Top 5 Key Differences Between Methods and Methodology

Sign-up to read more

Subscribe for free to get unrestricted access to all our resources on research writing and academic publishing including:

• 2000+ blog articles
• 50+ Webinars
• 10+ Expert podcasts
• 50+ Infographics
• 10+ Checklists
• Research Guides

We hate spam too. We promise to protect your privacy and never spam you.

I am looking for Editing/ Proofreading services for my manuscript Tentative date of next journal submission:

As a researcher, what do you consider most when choosing an image manipulation detector?

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

• PLOS Biology
• PLOS Climate
• PLOS Complex Systems
• PLOS Computational Biology
• PLOS Digital Health
• PLOS Genetics
• PLOS Global Public Health
• PLOS Medicine
• PLOS Mental Health
• PLOS Neglected Tropical Diseases
• PLOS Pathogens
• PLOS Sustainability and Transformation
• PLOS Collections
• How to Write Discussions and Conclusions

The discussion section contains the results and outcomes of a study. An effective discussion informs readers what can be learned from your experiment and provides context for the results.

What makes an effective discussion?

When you’re ready to write your discussion, you’ve already introduced the purpose of your study and provided an in-depth description of the methodology. The discussion informs readers about the larger implications of your study based on the results. Highlighting these implications while not overstating the findings can be challenging, especially when you’re submitting to a journal that selects articles based on novelty or potential impact. Regardless of what journal you are submitting to, the discussion section always serves the same purpose: concluding what your study results actually mean.

A successful discussion section puts your findings in context. It should include:

• the results of your research,
• a discussion of related research, and
• a comparison between your results and initial hypothesis.

Tip: Not all journals share the same naming conventions.

You can apply the advice in this article to the conclusion, results or discussion sections of your manuscript.

Our Early Career Researcher community tells us that the conclusion is often considered the most difficult aspect of a manuscript to write. To help, this guide provides questions to ask yourself, a basic structure to model your discussion off of and examples from published manuscripts. 

Questions to ask yourself:

• Was my hypothesis correct?
• If my hypothesis is partially correct or entirely different, what can be learned from the results? 
• How do the conclusions reshape or add onto the existing knowledge in the field? What does previous research say about the topic? 
• Why are the results important or relevant to your audience? Do they add further evidence to a scientific consensus or disprove prior studies? 
• How can future research build on these observations? What are the key experiments that must be done? 
• What is the “take-home” message you want your reader to leave with?

How to structure a discussion

Trying to fit a complete discussion into a single paragraph can add unnecessary stress to the writing process. If possible, you’ll want to give yourself two or three paragraphs to give the reader a comprehensive understanding of your study as a whole. Here’s one way to structure an effective discussion:

Writing Tips

While the above sections can help you brainstorm and structure your discussion, there are many common mistakes that writers revert to when having difficulties with their paper. Writing a discussion can be a delicate balance between summarizing your results, providing proper context for your research and avoiding introducing new information. Remember that your paper should be both confident and honest about the results! 

• Read the journal’s guidelines on the discussion and conclusion sections. If possible, learn about the guidelines before writing the discussion to ensure you’re writing to meet their expectations. 
• Begin with a clear statement of the principal findings. This will reinforce the main take-away for the reader and set up the rest of the discussion. 
• Explain why the outcomes of your study are important to the reader. Discuss the implications of your findings realistically based on previous literature, highlighting both the strengths and limitations of the research. 
• State whether the results prove or disprove your hypothesis. If your hypothesis was disproved, what might be the reasons? 
• Introduce new or expanded ways to think about the research question. Indicate what next steps can be taken to further pursue any unresolved questions. 
• If dealing with a contemporary or ongoing problem, such as climate change, discuss possible consequences if the problem is avoided. 
• Be concise. Adding unnecessary detail can distract from the main findings. 

Don’t

• Rewrite your abstract. Statements with “we investigated” or “we studied” generally do not belong in the discussion. 
• Include new arguments or evidence not previously discussed. Necessary information and evidence should be introduced in the main body of the paper. 
• Apologize. Even if your research contains significant limitations, don’t undermine your authority by including statements that doubt your methodology or execution. 
• Shy away from speaking on limitations or negative results. Including limitations and negative results will give readers a complete understanding of the presented research. Potential limitations include sources of potential bias, threats to internal or external validity, barriers to implementing an intervention and other issues inherent to the study design. 
• Overstate the importance of your findings. Making grand statements about how a study will fully resolve large questions can lead readers to doubt the success of the research. 

Snippets of Effective Discussions:

Consumer-based actions to reduce plastic pollution in rivers: A multi-criteria decision analysis approach

Identifying reliable indicators of fitness in polar bears

• How to Write a Great Title
• How to Write an Abstract
• How to Write Your Methods
• How to Report Statistics
• How to Edit Your Work

The contents of the Peer Review Center are also available as a live, interactive training session, complete with slides, talking points, and activities. …

The contents of the Writing Center are also available as a live, interactive training session, complete with slides, talking points, and activities. …

There’s a lot to consider when deciding where to submit your work. Learn how to choose a journal that will help your study reach its audience, while reflecting your values as a researcher…

• Affiliate Program

• UNITED STATES
• 台灣 (TAIWAN)
• TÜRKIYE (TURKEY)
• Academic Editing Services
• - Research Paper
• - Journal Manuscript
• - Dissertation
• - College & University Assignments
• Admissions Editing Services
• - Application Essay
• - Personal Statement
• - Recommendation Letter
• - Cover Letter
• - CV/Resume
• Business Editing Services
• - Business Documents
• - Report & Brochure
• - Website & Blog
• Writer Editing Services
• - Script & Screenplay
• Our Editors
• Client Reviews
• Editing & Proofreading Prices
• Wordvice Points
• Partner Discount
• Plagiarism Checker

APA Citation Generator

MLA Citation Generator

Chicago Citation Generator

Vancouver Citation Generator

• - APA Style
• - MLA Style
• - Chicago Style
• - Vancouver Style
• Writing & Editing Guide
• Academic Resources
• Admissions Resources

How to Present the Limitations of the Study Examples

What are the limitations of a study?

The limitations of a study are the elements of methodology or study design that impact the interpretation of your research results. The limitations essentially detail any flaws or shortcomings in your study. Study limitations can exist due to constraints on research design, methodology, materials, etc., and these factors may impact the findings of your study. However, researchers are often reluctant to discuss the limitations of their study in their papers, feeling that bringing up limitations may undermine its research value in the eyes of readers and reviewers.

In spite of the impact it might have (and perhaps because of it) you should clearly acknowledge any limitations in your research paper in order to show readers—whether journal editors, other researchers, or the general public—that you are aware of these limitations and to explain how they affect the conclusions that can be drawn from the research.

In this article, we provide some guidelines for writing about research limitations, show examples of some frequently seen study limitations, and recommend techniques for presenting this information. And after you have finished drafting and have received manuscript editing for your work, you still might want to follow this up with academic editing before submitting your work to your target journal.

Why do I need to include limitations of research in my paper?

Although limitations address the potential weaknesses of a study, writing about them toward the end of your paper actually strengthens your study by identifying any problems before other researchers or reviewers find them.

Furthermore, pointing out study limitations shows that you’ve considered the impact of research weakness thoroughly and have an in-depth understanding of your research topic. Since all studies face limitations, being honest and detailing these limitations will impress researchers and reviewers more than ignoring them.

Where should I put the limitations of the study in my paper?

Some limitations might be evident to researchers before the start of the study, while others might become clear while you are conducting the research. Whether these limitations are anticipated or not, and whether they are due to research design or to methodology, they should be clearly identified and discussed in the discussion section —the final section of your paper. Most journals now require you to include a discussion of potential limitations of your work, and many journals now ask you to place this “limitations section” at the very end of your article. 

Some journals ask you to also discuss the strengths of your work in this section, and some allow you to freely choose where to include that information in your discussion section—make sure to always check the author instructions of your target journal before you finalize a manuscript and submit it for peer review .

Limitations of the Study Examples

There are several reasons why limitations of research might exist. The two main categories of limitations are those that result from the methodology and those that result from issues with the researcher(s).

Common Methodological Limitations of Studies

Limitations of research due to methodological problems can be addressed by clearly and directly identifying the potential problem and suggesting ways in which this could have been addressed—and SHOULD be addressed in future studies. The following are some major potential methodological issues that can impact the conclusions researchers can draw from the research.

Issues with research samples and selection

Sampling errors occur when a probability sampling method is used to select a sample, but that sample does not reflect the general population or appropriate population concerned. This results in limitations of your study known as “sample bias” or “selection bias.”

For example, if you conducted a survey to obtain your research results, your samples (participants) were asked to respond to the survey questions. However, you might have had limited ability to gain access to the appropriate type or geographic scope of participants. In this case, the people who responded to your survey questions may not truly be a random sample.

Insufficient sample size for statistical measurements

When conducting a study, it is important to have a sufficient sample size in order to draw valid conclusions. The larger the sample, the more precise your results will be. If your sample size is too small, it will be difficult to identify significant relationships in the data.

Normally, statistical tests require a larger sample size to ensure that the sample is considered representative of a population and that the statistical result can be generalized to a larger population. It is a good idea to understand how to choose an appropriate sample size before you conduct your research by using scientific calculation tools—in fact, many journals now require such estimation to be included in every manuscript that is sent out for review.

Lack of previous research studies on the topic

Citing and referencing prior research studies constitutes the basis of the literature review for your thesis or study, and these prior studies provide the theoretical foundations for the research question you are investigating. However, depending on the scope of your research topic, prior research studies that are relevant to your thesis might be limited.

When there is very little or no prior research on a specific topic, you may need to develop an entirely new research typology. In this case, discovering a limitation can be considered an important opportunity to identify literature gaps and to present the need for further development in the area of study.

Methods/instruments/techniques used to collect the data

After you complete your analysis of the research findings (in the discussion section), you might realize that the manner in which you have collected the data or the ways in which you have measured variables has limited your ability to conduct a thorough analysis of the results.

For example, you might realize that you should have addressed your survey questions from another viable perspective, or that you were not able to include an important question in the survey. In these cases, you should acknowledge the deficiency or deficiencies by stating a need for future researchers to revise their specific methods for collecting data that includes these missing elements.

Common Limitations of the Researcher(s)

Study limitations that arise from situations relating to the researcher or researchers (whether the direct fault of the individuals or not) should also be addressed and dealt with, and remedies to decrease these limitations—both hypothetically in your study, and practically in future studies—should be proposed.

Limited access to data

If your research involved surveying certain people or organizations, you might have faced the problem of having limited access to these respondents. Due to this limited access, you might need to redesign or restructure your research in a different way. In this case, explain the reasons for limited access and be sure that your finding is still reliable and valid despite this limitation.

Time constraints

Just as students have deadlines to turn in their class papers, academic researchers might also have to meet deadlines for submitting a manuscript to a journal or face other time constraints related to their research (e.g., participants are only available during a certain period; funding runs out; collaborators move to a new institution). The time available to study a research problem and to measure change over time might be constrained by such practical issues. If time constraints negatively impacted your study in any way, acknowledge this impact by mentioning a need for a future study (e.g., a longitudinal study) to answer this research problem.

Conflicts arising from cultural bias and other personal issues

Researchers might hold biased views due to their cultural backgrounds or perspectives of certain phenomena, and this can affect a study’s legitimacy. Also, it is possible that researchers will have biases toward data and results that only support their hypotheses or arguments. In order to avoid these problems, the author(s) of a study should examine whether the way the research problem was stated and the data-gathering process was carried out appropriately.

Steps for Organizing Your Study Limitations Section

When you discuss the limitations of your study, don’t simply list and describe your limitations—explain how these limitations have influenced your research findings. There might be multiple limitations in your study, but you only need to point out and explain those that directly relate to and impact how you address your research questions.

We suggest that you divide your limitations section into three steps: (1) identify the study limitations; (2) explain how they impact your study in detail; and (3) propose a direction for future studies and present alternatives. By following this sequence when discussing your study’s limitations, you will be able to clearly demonstrate your study’s weakness without undermining the quality and integrity of your research.

Step 1. Identify the limitation(s) of the study

• This part should comprise around 10%-20% of your discussion of study limitations.

The first step is to identify the particular limitation(s) that affected your study. There are many possible limitations of research that can affect your study, but you don’t need to write a long review of all possible study limitations. A 200-500 word critique is an appropriate length for a research limitations section. In the beginning of this section, identify what limitations your study has faced and how important these limitations are.

You only need to identify limitations that had the greatest potential impact on: (1) the quality of your findings, and (2) your ability to answer your research question.

Step 2. Explain these study limitations in detail

• This part should comprise around 60-70% of your discussion of limitations.

After identifying your research limitations, it’s time to explain the nature of the limitations and how they potentially impacted your study. For example, when you conduct quantitative research, a lack of probability sampling is an important issue that you should mention. On the other hand, when you conduct qualitative research, the inability to generalize the research findings could be an issue that deserves mention.

Explain the role these limitations played on the results and implications of the research and justify the choice you made in using this “limiting” methodology or other action in your research. Also, make sure that these limitations didn’t undermine the quality of your dissertation .

Step 3. Propose a direction for future studies and present alternatives (optional)

• This part should comprise around 10-20% of your discussion of limitations.

After acknowledging the limitations of the research, you need to discuss some possible ways to overcome these limitations in future studies. One way to do this is to present alternative methodologies and ways to avoid issues with, or “fill in the gaps of” the limitations of this study you have presented.  Discuss both the pros and cons of these alternatives and clearly explain why researchers should choose these approaches.

Make sure you are current on approaches used by prior studies and the impacts they have had on their findings. Cite review articles or scientific bodies that have recommended these approaches and why. This might be evidence in support of the approach you chose, or it might be the reason you consider your choices to be included as limitations. This process can act as a justification for your approach and a defense of your decision to take it while acknowledging the feasibility of other approaches.

P hrases and Tips for Introducing Your Study Limitations in the Discussion Section

The following phrases are frequently used to introduce the limitations of the study:

• “There may be some possible limitations in this study.”
• “The findings of this study have to be seen in light of some limitations.”
•  “The first is the…The second limitation concerns the…”
•  “The empirical results reported herein should be considered in the light of some limitations.”
• “This research, however, is subject to several limitations.”
• “The primary limitation to the generalization of these results is…”
• “Nonetheless, these results must be interpreted with caution and a number of limitations should be borne in mind.”
• “As with the majority of studies, the design of the current study is subject to limitations.”
• “There are two major limitations in this study that could be addressed in future research. First, the study focused on …. Second ….”

For more articles on research writing and the journal submissions and publication process, visit Wordvice’s Academic Resources page.

And be sure to receive professional English editing and proofreading services , including paper editing services , for your journal manuscript before submitting it to journal editors.

Wordvice Resources

Proofreading & Editing Guide

Writing the Results Section for a Research Paper

How to Write a Literature Review

Research Writing Tips: How to Draft a Powerful Discussion Section

How to Captivate Journal Readers with a Strong Introduction

Tips That Will Make Your Abstract a Success!

APA In-Text Citation Guide for Research Writing

Additional Resources

• Diving Deeper into Limitations and Delimitations (PhD student)
• Organizing Your Social Sciences Research Paper: Limitations of the Study (USC Library)
• Research Limitations (Research Methodology)
• How to Present Limitations and Alternatives (UMASS)

Article References

Pearson-Stuttard, J., Kypridemos, C., Collins, B., Mozaffarian, D., Huang, Y., Bandosz, P.,…Micha, R. (2018). Estimating the health and economic effects of the proposed US Food and Drug Administration voluntary sodium reformulation: Microsimulation cost-effectiveness analysis. PLOS. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002551

Xu, W.L, Pedersen, N.L., Keller, L., Kalpouzos, G., Wang, H.X., Graff, C,. Fratiglioni, L. (2015). HHEX_23 AA Genotype Exacerbates Effect of Diabetes on Dementia and Alzheimer Disease: A Population-Based Longitudinal Study. PLOS. Retrieved from https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001853

21 Research Limitations Examples

Research limitations refer to the potential weaknesses inherent in a study. All studies have limitations of some sort, meaning declaring limitations doesn’t necessarily need to be a bad thing, so long as your declaration of limitations is well thought-out and explained.

Rarely is a study perfect. Researchers have to make trade-offs when developing their studies, which are often based upon practical considerations such as time and monetary constraints, weighing the breadth of participants against the depth of insight, and choosing one methodology or another.

In research, studies can have limitations such as limited scope, researcher subjectivity, and lack of available research tools.

Acknowledging the limitations of your study should be seen as a strength. It demonstrates your willingness for transparency, humility, and submission to the scientific method and can bolster the integrity of the study. It can also inform future research direction.

Typically, scholars will explore the limitations of their study in either their methodology section, their conclusion section, or both.

Research Limitations Examples

Qualitative and quantitative research offer different perspectives and methods in exploring phenomena, each with its own strengths and limitations. So, I’ve split the limitations examples sections into qualitative and quantitative below.

Qualitative Research Limitations

Qualitative research seeks to understand phenomena in-depth and in context. It focuses on the ‘why’ and ‘how’ questions.

It’s often used to explore new or complex issues, and it provides rich, detailed insights into participants’ experiences, behaviors, and attitudes. However, these strengths also create certain limitations, as explained below.

1. Subjectivity

Qualitative research often requires the researcher to interpret subjective data. One researcher may examine a text and identify different themes or concepts as more dominant than others.

Close qualitative readings of texts are necessarily subjective – and while this may be a limitation, qualitative researchers argue this is the best way to deeply understand everything in context.

Suggested Solution and Response: To minimize subjectivity bias, you could consider cross-checking your own readings of themes and data against other scholars’ readings and interpretations. This may involve giving the raw data to a supervisor or colleague and asking them to code the data separately, then coming together to compare and contrast results.

2. Researcher Bias

The concept of researcher bias is related to, but slightly different from, subjectivity.

Researcher bias refers to the perspectives and opinions you bring with you when doing your research.

For example, a researcher who is explicitly of a certain philosophical or political persuasion may bring that persuasion to bear when interpreting data.

In many scholarly traditions, we will attempt to minimize researcher bias through the utilization of clear procedures that are set out in advance or through the use of statistical analysis tools.

However, in other traditions, such as in postmodern feminist research , declaration of bias is expected, and acknowledgment of bias is seen as a positive because, in those traditions, it is believed that bias cannot be eliminated from research, so instead, it is a matter of integrity to present it upfront.

Suggested Solution and Response: Acknowledge the potential for researcher bias and, depending on your theoretical framework , accept this, or identify procedures you have taken to seek a closer approximation to objectivity in your coding and analysis.

3. Generalizability

If you’re struggling to find a limitation to discuss in your own qualitative research study, then this one is for you: all qualitative research, of all persuasions and perspectives, cannot be generalized.

This is a core feature that sets qualitative data and quantitative data apart.

The point of qualitative data is to select case studies and similarly small corpora and dig deep through in-depth analysis and thick description of data.

Often, this will also mean that you have a non-randomized sample size.

While this is a positive – you’re going to get some really deep, contextualized, interesting insights – it also means that the findings may not be generalizable to a larger population that may not be representative of the small group of people in your study.

Suggested Solution and Response: Suggest future studies that take a quantitative approach to the question.

4. The Hawthorne Effect

The Hawthorne effect refers to the phenomenon where research participants change their ‘observed behavior’ when they’re aware that they are being observed.

This effect was first identified by Elton Mayo who conducted studies of the effects of various factors ton workers’ productivity. He noticed that no matter what he did – turning up the lights, turning down the lights, etc. – there was an increase in worker outputs compared to prior to the study taking place.

Mayo realized that the mere act of observing the workers made them work harder – his observation was what was changing behavior.

So, if you’re looking for a potential limitation to name for your observational research study , highlight the possible impact of the Hawthorne effect (and how you could reduce your footprint or visibility in order to decrease its likelihood).

Suggested Solution and Response: Highlight ways you have attempted to reduce your footprint while in the field, and guarantee anonymity to your research participants.

5. Replicability

Quantitative research has a great benefit in that the studies are replicable – a researcher can get a similar sample size, duplicate the variables, and re-test a study. But you can’t do that in qualitative research.

Qualitative research relies heavily on context – a specific case study or specific variables that make a certain instance worthy of analysis. As a result, it’s often difficult to re-enter the same setting with the same variables and repeat the study.

Furthermore, the individual researcher’s interpretation is more influential in qualitative research, meaning even if a new researcher enters an environment and makes observations, their observations may be different because subjectivity comes into play much more. This doesn’t make the research bad necessarily (great insights can be made in qualitative research), but it certainly does demonstrate a weakness of qualitative research.

6. Limited Scope

“Limited scope” is perhaps one of the most common limitations listed by researchers – and while this is often a catch-all way of saying, “well, I’m not studying that in this study”, it’s also a valid point.

No study can explore everything related to a topic. At some point, we have to make decisions about what’s included in the study and what is excluded from the study.

So, you could say that a limitation of your study is that it doesn’t look at an extra variable or concept that’s certainly worthy of study but will have to be explored in your next project because this project has a clearly and narrowly defined goal.

Suggested Solution and Response: Be clear about what’s in and out of the study when writing your research question.

7. Time Constraints

This is also a catch-all claim you can make about your research project: that you would have included more people in the study, looked at more variables, and so on. But you’ve got to submit this thing by the end of next semester! You’ve got time constraints.

And time constraints are a recognized reality in all research.

But this means you’ll need to explain how time has limited your decisions. As with “limited scope”, this may mean that you had to study a smaller group of subjects, limit the amount of time you spent in the field, and so forth.

Suggested Solution and Response: Suggest future studies that will build on your current work, possibly as a PhD project.

8. Resource Intensiveness

Qualitative research can be expensive due to the cost of transcription, the involvement of trained researchers, and potential travel for interviews or observations.

So, resource intensiveness is similar to the time constraints concept. If you don’t have the funds, you have to make decisions about which tools to use, which statistical software to employ, and how many research assistants you can dedicate to the study.

Suggested Solution and Response: Suggest future studies that will gain more funding on the back of this ‘ exploratory study ‘.

9. Coding Difficulties

Data analysis in qualitative research often involves coding, which can be subjective and complex, especially when dealing with ambiguous or contradicting data.

After naming this as a limitation in your research, it’s important to explain how you’ve attempted to address this. Some ways to ‘limit the limitation’ include:

• Triangulation: Have 2 other researchers code the data as well and cross-check your results with theirs to identify outliers that may need to be re-examined, debated with the other researchers, or removed altogether.
• Procedure: Use a clear coding procedure to demonstrate reliability in your coding process. I personally use the thematic network analysis method outlined in this academic article by Attride-Stirling (2001).

Suggested Solution and Response: Triangulate your coding findings with colleagues, and follow a thematic network analysis procedure.

10. Risk of Non-Responsiveness

There is always a risk in research that research participants will be unwilling or uncomfortable sharing their genuine thoughts and feelings in the study.

This is particularly true when you’re conducting research on sensitive topics, politicized topics, or topics where the participant is expressing vulnerability .

This is similar to the Hawthorne effect (aka participant bias), where participants change their behaviors in your presence; but it goes a step further, where participants actively hide their true thoughts and feelings from you.

Suggested Solution and Response: One way to manage this is to try to include a wider group of people with the expectation that there will be non-responsiveness from some participants.

11. Risk of Attrition

Attrition refers to the process of losing research participants throughout the study.

This occurs most commonly in longitudinal studies , where a researcher must return to conduct their analysis over spaced periods of time, often over a period of years.

Things happen to people over time – they move overseas, their life experiences change, they get sick, change their minds, and even die. The more time that passes, the greater the risk of attrition.

Suggested Solution and Response: One way to manage this is to try to include a wider group of people with the expectation that there will be attrition over time.

12. Difficulty in Maintaining Confidentiality and Anonymity

Given the detailed nature of qualitative data , ensuring participant anonymity can be challenging.

If you have a sensitive topic in a specific case study, even anonymizing research participants sometimes isn’t enough. People might be able to induce who you’re talking about.

Sometimes, this will mean you have to exclude some interesting data that you collected from your final report. Confidentiality and anonymity come before your findings in research ethics – and this is a necessary limiting factor.

Suggested Solution and Response: Highlight the efforts you have taken to anonymize data, and accept that confidentiality and accountability place extremely important constraints on academic research.

13. Difficulty in Finding Research Participants

A study that looks at a very specific phenomenon or even a specific set of cases within a phenomenon means that the pool of potential research participants can be very low.

Compile on top of this the fact that many people you approach may choose not to participate, and you could end up with a very small corpus of subjects to explore. This may limit your ability to make complete findings, even in a quantitative sense.

You may need to therefore limit your research question and objectives to something more realistic.

Suggested Solution and Response: Highlight that this is going to limit the study’s generalizability significantly.

14. Ethical Limitations

Ethical limitations refer to the things you cannot do based on ethical concerns identified either by yourself or your institution’s ethics review board.

This might include threats to the physical or psychological well-being of your research subjects, the potential of releasing data that could harm a person’s reputation, and so on.

Furthermore, even if your study follows all expected standards of ethics, you still, as an ethical researcher, need to allow a research participant to pull out at any point in time, after which you cannot use their data, which demonstrates an overlap between ethical constraints and participant attrition.

Suggested Solution and Response: Highlight that these ethical limitations are inevitable but important to sustain the integrity of the research.

For more on Qualitative Research, Explore my Qualitative Research Guide

Quantitative Research Limitations

Quantitative research focuses on quantifiable data and statistical, mathematical, or computational techniques. It’s often used to test hypotheses, assess relationships and causality, and generalize findings across larger populations.

Quantitative research is widely respected for its ability to provide reliable, measurable, and generalizable data (if done well!). Its structured methodology has strengths over qualitative research, such as the fact it allows for replication of the study, which underpins the validity of the research.

However, this approach is not without it limitations, explained below.

1. Over-Simplification

Quantitative research is powerful because it allows you to measure and analyze data in a systematic and standardized way. However, one of its limitations is that it can sometimes simplify complex phenomena or situations.

In other words, it might miss the subtleties or nuances of the research subject.

For example, if you’re studying why people choose a particular diet, a quantitative study might identify factors like age, income, or health status. But it might miss other aspects, such as cultural influences or personal beliefs, that can also significantly impact dietary choices.

When writing about this limitation, you can say that your quantitative approach, while providing precise measurements and comparisons, may not capture the full complexity of your subjects of study.

Suggested Solution and Response: Suggest a follow-up case study using the same research participants in order to gain additional context and depth.

2. Lack of Context

Another potential issue with quantitative research is that it often focuses on numbers and statistics at the expense of context or qualitative information.

Let’s say you’re studying the effect of classroom size on student performance. You might find that students in smaller classes generally perform better. However, this doesn’t take into account other variables, like teaching style , student motivation, or family support.

When describing this limitation, you might say, “Although our research provides important insights into the relationship between class size and student performance, it does not incorporate the impact of other potentially influential variables. Future research could benefit from a mixed-methods approach that combines quantitative analysis with qualitative insights.”

3. Applicability to Real-World Settings

Oftentimes, experimental research takes place in controlled environments to limit the influence of outside factors.

This control is great for isolation and understanding the specific phenomenon but can limit the applicability or “external validity” of the research to real-world settings.

For example, if you conduct a lab experiment to see how sleep deprivation impacts cognitive performance, the sterile, controlled lab environment might not reflect real-world conditions where people are dealing with multiple stressors.

Therefore, when explaining the limitations of your quantitative study in your methodology section, you could state:

“While our findings provide valuable information about [topic], the controlled conditions of the experiment may not accurately represent real-world scenarios where extraneous variables will exist. As such, the direct applicability of our results to broader contexts may be limited.”

Suggested Solution and Response: Suggest future studies that will engage in real-world observational research, such as ethnographic research.

4. Limited Flexibility

Once a quantitative study is underway, it can be challenging to make changes to it. This is because, unlike in grounded research, you’re putting in place your study in advance, and you can’t make changes part-way through.

Your study design, data collection methods, and analysis techniques need to be decided upon before you start collecting data.

For example, if you are conducting a survey on the impact of social media on teenage mental health, and halfway through, you realize that you should have included a question about their screen time, it’s generally too late to add it.

When discussing this limitation, you could write something like, “The structured nature of our quantitative approach allows for consistent data collection and analysis but also limits our flexibility to adapt and modify the research process in response to emerging insights and ideas.”

Suggested Solution and Response: Suggest future studies that will use mixed-methods or qualitative research methods to gain additional depth of insight.

5. Risk of Survey Error

Surveys are a common tool in quantitative research, but they carry risks of error.

There can be measurement errors (if a question is misunderstood), coverage errors (if some groups aren’t adequately represented), non-response errors (if certain people don’t respond), and sampling errors (if your sample isn’t representative of the population).

For instance, if you’re surveying college students about their study habits , but only daytime students respond because you conduct the survey during the day, your results will be skewed.

In discussing this limitation, you might say, “Despite our best efforts to develop a comprehensive survey, there remains a risk of survey error, including measurement, coverage, non-response, and sampling errors. These could potentially impact the reliability and generalizability of our findings.”

Suggested Solution and Response: Suggest future studies that will use other survey tools to compare and contrast results.

6. Limited Ability to Probe Answers

With quantitative research, you typically can’t ask follow-up questions or delve deeper into participants’ responses like you could in a qualitative interview.

For instance, imagine you are surveying 500 students about study habits in a questionnaire. A respondent might indicate that they study for two hours each night. You might want to follow up by asking them to elaborate on what those study sessions involve or how effective they feel their habits are.

However, quantitative research generally disallows this in the way a qualitative semi-structured interview could.

When discussing this limitation, you might write, “Given the structured nature of our survey, our ability to probe deeper into individual responses is limited. This means we may not fully understand the context or reasoning behind the responses, potentially limiting the depth of our findings.”

Suggested Solution and Response: Suggest future studies that engage in mixed-method or qualitative methodologies to address the issue from another angle.

7. Reliance on Instruments for Data Collection

In quantitative research, the collection of data heavily relies on instruments like questionnaires, surveys, or machines.

The limitation here is that the data you get is only as good as the instrument you’re using. If the instrument isn’t designed or calibrated well, your data can be flawed.

For instance, if you’re using a questionnaire to study customer satisfaction and the questions are vague, confusing, or biased, the responses may not accurately reflect the customers’ true feelings.

When discussing this limitation, you could say, “Our study depends on the use of questionnaires for data collection. Although we have put significant effort into designing and testing the instrument, it’s possible that inaccuracies or misunderstandings could potentially affect the validity of the data collected.”

Suggested Solution and Response: Suggest future studies that will use different instruments but examine the same variables to triangulate results.

8. Time and Resource Constraints (Specific to Quantitative Research)

Quantitative research can be time-consuming and resource-intensive, especially when dealing with large samples.

It often involves systematic sampling, rigorous design, and sometimes complex statistical analysis.

If resources and time are limited, it can restrict the scale of your research, the techniques you can employ, or the extent of your data analysis.

For example, you may want to conduct a nationwide survey on public opinion about a certain policy. However, due to limited resources, you might only be able to survey people in one city.

When writing about this limitation, you could say, “Given the scope of our research and the resources available, we are limited to conducting our survey within one city, which may not fully represent the nationwide public opinion. Hence, the generalizability of the results may be limited.”

Suggested Solution and Response: Suggest future studies that will have more funding or longer timeframes.

How to Discuss Your Research Limitations

1. in your research proposal and methodology section.

In the research proposal, which will become the methodology section of your dissertation, I would recommend taking the four following steps, in order:

• Be Explicit about your Scope – If you limit the scope of your study in your research question, aims, and objectives, then you can set yourself up well later in the methodology to say that certain questions are “outside the scope of the study.” For example, you may identify the fact that the study doesn’t address a certain variable, but you can follow up by stating that the research question is specifically focused on the variable that you are examining, so this limitation would need to be looked at in future studies.
• Acknowledge the Limitation – Acknowledging the limitations of your study demonstrates reflexivity and humility and can make your research more reliable and valid. It also pre-empts questions the people grading your paper may have, so instead of them down-grading you for your limitations; they will congratulate you on explaining the limitations and how you have addressed them!
• Explain your Decisions – You may have chosen your approach (despite its limitations) for a very specific reason. This might be because your approach remains, on balance, the best one to answer your research question. Or, it might be because of time and monetary constraints that are outside of your control.
• Highlight the Strengths of your Approach – Conclude your limitations section by strongly demonstrating that, despite limitations, you’ve worked hard to minimize the effects of the limitations and that you have chosen your specific approach and methodology because it’s also got some terrific strengths. Name the strengths.

Overall, you’ll want to acknowledge your own limitations but also explain that the limitations don’t detract from the value of your study as it stands.

2. In the Conclusion Section or Chapter

In the conclusion of your study, it is generally expected that you return to a discussion of the study’s limitations. Here, I recommend the following steps:

• Acknowledge issues faced – After completing your study, you will be increasingly aware of issues you may have faced that, if you re-did the study, you may have addressed earlier in order to avoid those issues. Acknowledge these issues as limitations, and frame them as recommendations for subsequent studies.
• Suggest further research – Scholarly research aims to fill gaps in the current literature and knowledge. Having established your expertise through your study, suggest lines of inquiry for future researchers. You could state that your study had certain limitations, and “future studies” can address those limitations.
• Suggest a mixed methods approach – Qualitative and quantitative research each have pros and cons. So, note those ‘cons’ of your approach, then say the next study should approach the topic using the opposite methodology or could approach it using a mixed-methods approach that could achieve the benefits of quantitative studies with the nuanced insights of associated qualitative insights as part of an in-study case-study.

Overall, be clear about both your limitations and how those limitations can inform future studies.

In sum, each type of research method has its own strengths and limitations. Qualitative research excels in exploring depth, context, and complexity, while quantitative research excels in examining breadth, generalizability, and quantifiable measures. Despite their individual limitations, each method contributes unique and valuable insights, and researchers often use them together to provide a more comprehensive understanding of the phenomenon being studied.

Attride-Stirling, J. (2001). Thematic networks: an analytic tool for qualitative research. Qualitative research , 1 (3), 385-405. ( Source )

Atkinson, P., Delamont, S., Cernat, A., Sakshaug, J., & Williams, R. A. (2021).  SAGE research methods foundations . London: Sage Publications.

Clark, T., Foster, L., Bryman, A., & Sloan, L. (2021).  Bryman’s social research methods . Oxford: Oxford University Press.

Köhler, T., Smith, A., & Bhakoo, V. (2022). Templates in qualitative research methods: Origins, limitations, and new directions.  Organizational Research Methods ,  25 (2), 183-210. ( Source )

Lenger, A. (2019). The rejection of qualitative research methods in economics.  Journal of Economic Issues ,  53 (4), 946-965. ( Source )

Taherdoost, H. (2022). What are different research approaches? Comprehensive review of qualitative, quantitative, and mixed method research, their applications, types, and limitations.  Journal of Management Science & Engineering Research ,  5 (1), 53-63. ( Source )

Walliman, N. (2021).  Research methods: The basics . New York: Routledge.

Chris Drew (PhD)

Dr. Chris Drew is the founder of the Helpful Professor. He holds a PhD in education and has published over 20 articles in scholarly journals. He is the former editor of the Journal of Learning Development in Higher Education. [Image Descriptor: Photo of Chris]

• Chris Drew (PhD) https://helpfulprofessor.com/author/chris-drew-phd/ 15 Top Stakeholders in Education
• Chris Drew (PhD) https://helpfulprofessor.com/author/chris-drew-phd/ The Six Principles of Andragogy (Malcolm Knowles)
• Chris Drew (PhD) https://helpfulprofessor.com/author/chris-drew-phd/ What are Pedagogical Skills? - 15 Examples
• Chris Drew (PhD) https://helpfulprofessor.com/author/chris-drew-phd/ 44 Maslow’s Hierarchy of Needs Examples

Leave a Comment Cancel Reply

Your email address will not be published. Required fields are marked *

Stating the Obvious: Writing Assumptions, Limitations, and Delimitations

During the process of writing your thesis or dissertation, you might suddenly realize that your research has inherent flaws. Don’t worry! Virtually all projects contain restrictions to your research. However, being able to recognize and accurately describe these problems is the difference between a true researcher and a grade-school kid with a science-fair project. Concerns with truthful responding, access to participants, and survey instruments are just a few of examples of restrictions on your research. In the following sections, the differences among delimitations, limitations, and assumptions of a dissertation will be clarified.

Delimitations

Delimitations are the definitions you set as the boundaries of your own thesis or dissertation, so delimitations are in your control. Delimitations are set so that your goals do not become impossibly large to complete. Examples of delimitations include objectives, research questions, variables, theoretical objectives that you have adopted, and populations chosen as targets to study. When you are stating your delimitations, clearly inform readers why you chose this course of study. The answer might simply be that you were curious about the topic and/or wanted to improve standards of a professional field by revealing certain findings. In any case, you should clearly list the other options available and the reasons why you did not choose these options immediately after you list your delimitations. You might have avoided these options for reasons of practicality, interest, or relativity to the study at hand. For example, you might have only studied Hispanic mothers because they have the highest rate of obese babies. Delimitations are often strongly related to your theory and research questions. If you were researching whether there are different parenting styles between unmarried Asian, Caucasian, African American, and Hispanic women, then a delimitation of your study would be the inclusion of only participants with those demographics and the exclusion of participants from other demographics such as men, married women, and all other ethnicities of single women (inclusion and exclusion criteria). A further delimitation might be that you only included closed-ended Likert scale responses in the survey, rather than including additional open-ended responses, which might make some people more willing to take and complete your survey. Remember that delimitations are not good or bad. They are simply a detailed description of the scope of interest for your study as it relates to the research design. Don’t forget to describe the philosophical framework you used throughout your study, which also delimits your study.

Limitations

Limitations of a dissertation are potential weaknesses in your study that are mostly out of your control, given limited funding, choice of research design, statistical model constraints, or other factors. In addition, a limitation is a restriction on your study that cannot be reasonably dismissed and can affect your design and results. Do not worry about limitations because limitations affect virtually all research projects, as well as most things in life. Even when you are going to your favorite restaurant, you are limited by the menu choices. If you went to a restaurant that had a menu that you were craving, you might not receive the service, price, or location that makes you enjoy your favorite restaurant. If you studied participants’ responses to a survey, you might be limited in your abilities to gain the exact type or geographic scope of participants you wanted. The people whom you managed to get to take your survey may not truly be a random sample, which is also a limitation. If you used a common test for data findings, your results are limited by the reliability of the test. If your study was limited to a certain amount of time, your results are affected by the operations of society during that time period (e.g., economy, social trends). It is important for you to remember that limitations of a dissertation are often not something that can be solved by the researcher. Also, remember that whatever limits you also limits other researchers, whether they are the largest medical research companies or consumer habits corporations. Certain kinds of limitations are often associated with the analytical approach you take in your research, too. For example, some qualitative methods like heuristics or phenomenology do not lend themselves well to replicability. Also, most of the commonly used quantitative statistical models can only determine correlation, but not causation.

Assumptions

Assumptions are things that are accepted as true, or at least plausible, by researchers and peers who will read your dissertation or thesis. In other words, any scholar reading your paper will assume that certain aspects of your study is true given your population, statistical test, research design, or other delimitations. For example, if you tell your friend that your favorite restaurant is an Italian place, your friend will assume that you don’t go there for the sushi. It’s assumed that you go there to eat Italian food. Because most assumptions are not discussed in-text, assumptions that are discussed in-text are discussed in the context of the limitations of your study, which is typically in the discussion section. This is important, because both assumptions and limitations affect the inferences you can draw from your study. One of the more common assumptions made in survey research is the assumption of honesty and truthful responses. However, for certain sensitive questions this assumption may be more difficult to accept, in which case it would be described as a limitation of the study. For example, asking people to report their criminal behavior in a survey may not be as reliable as asking people to report their eating habits. It is important to remember that your limitations and assumptions should not contradict one another. For instance, if you state that generalizability is a limitation of your study given that your sample was limited to one city in the United States, then you should not claim generalizability to the United States population as an assumption of your study. Statistical models in quantitative research designs are accompanied with assumptions as well, some more strict than others. These assumptions generally refer to the characteristics of the data, such as distributions, correlational trends, and variable type, just to name a few. Violating these assumptions can lead to drastically invalid results, though this often depends on sample size and other considerations.

Click here to cancel reply.

You must be logged in to post a comment.

Copyright © 2024 PhDStudent.com. All rights reserved. Designed by Divergent Web Solutions, LLC .

Organizing Academic Research Papers: Limitations of the Study

• Purpose of Guide
• Design Flaws to Avoid
• Glossary of Research Terms
• Narrowing a Topic Idea
• Broadening a Topic Idea
• Extending the Timeliness of a Topic Idea
• Academic Writing Style
• Choosing a Title
• Making an Outline
• Paragraph Development
• Executive Summary
• Background Information
• The Research Problem/Question
• Theoretical Framework
• Citation Tracking
• Content Alert Services
• Evaluating Sources
• Primary Sources
• Secondary Sources
• Tertiary Sources
• What Is Scholarly vs. Popular?
• Qualitative Methods
• Quantitative Methods
• Using Non-Textual Elements
• Limitations of the Study
• Common Grammar Mistakes
• Avoiding Plagiarism
• Footnotes or Endnotes?
• Further Readings
• Annotated Bibliography
• Dealing with Nervousness
• Using Visual Aids
• Grading Someone Else's Paper
• How to Manage Group Projects
• Multiple Book Review Essay
• Reviewing Collected Essays
• About Informed Consent
• Writing Field Notes
• Writing a Policy Memo
• Writing a Research Proposal
• Acknowledgements

The limitations of the study are those characteristics of design or methodology that impacted or influenced the application or interpretation of the results of your study. They are the constraints on generalizability and utility of findings that are the result of the ways in which you chose to design the study and/or the method used to establish internal and external validity.

Importance of...

Always acknowledge a study's limitations. It is far better for you to identify and acknowledge your study’s limitations than to have them pointed out by your professor and be graded down because you appear to have ignored them.

Keep in mind that acknowledgement of a study's limitations is an opportunity to make suggestions for further research. If you do connect your study's limitations to suggestions for further research, be sure to explain the ways in which these unanswered questions may become more focused because of your study.

Acknowledgement of a study's limitations also provides you with an opportunity to demonstrate to your professor that you have thought critically about the research problem, understood the relevant literature published about it, and correctly assessed the methods chosen for studying the problem. A key objective of the research process is not only discovering new knowledge but also to confront assumptions and explore what we don't know.

Claiming limitiations is a subjective process because you must evaluate the impact of those limitations . Don't just list key weaknesses and the magnitude of a study's limitations. To do so diminishes the validity of your research because it leaves the reader wondering whether, or in what ways, limitation(s) in your study may have impacted the findings and conclusions. Limitations require a critical, overall appraisal and interpretation of their impact. You should answer the question: do these problems with errors, methods, validity, etc. eventually matter and, if so, to what extent?

Structure: How to Structure the Research Limitations Section of Your Dissertation . Dissertations and Theses: An Online Textbook. Laerd.com.

Descriptions of Possible Limitations

All studies have limitations . However, it is important that you restrict your discussion to limitations related to the research problem under investigation. For example, if a meta-analysis of existing literature is not a stated purpose of your research, it should not be discussed as a limitation. Do not apologize for not addressing issues that you did not promise to investigate in your paper.

Here are examples of limitations you may need to describe and to discuss how they possibly impacted your findings. Descriptions of limitations should be stated in the past tense.

Possible Methodological Limitations

• Sample size -- the number of the units of analysis you use in your study is dictated by the type of research problem you are investigating. Note that, if your sample size is too small, it will be difficult to find significant relationships from the data, as statistical tests normally require a larger sample size to ensure a representative distribution of the population and to be considered representative of groups of people to whom results will be generalized or transferred.
• Lack of available and/or reliable data -- a lack of data or of reliable data will likely require you to limit the scope of your analysis, the size of your sample, or it can be a significant obstacle in finding a trend and a meaningful relationship. You need to not only describe these limitations but to offer reasons why you believe data is missing or is unreliable. However, don’t just throw up your hands in frustration; use this as an opportunity to describe the need for future research.
• Lack of prior research studies on the topic -- citing prior research studies forms the basis of your literature review and helps lay a foundation for understanding the research problem you are investigating. Depending on the currency or scope of your research topic, there may be little, if any, prior research on your topic. Before assuming this to be true, consult with a librarian! In cases when a librarian has confirmed that there is a lack of prior research, you may be required to develop an entirely new research typology [for example, using an exploratory rather than an explanatory research design]. Note that this limitation can serve as an important opportunity to describe the need for further research.
• Measure used to collect the data -- sometimes it is the case that, after completing your interpretation of the findings, you discover that the way in which you gathered data inhibited your ability to conduct a thorough analysis of the results. For example, you regret not including a specific question in a survey that, in retrospect, could have helped address a particular issue that emerged later in the study. Acknowledge the deficiency by stating a need in future research to revise the specific method for gathering data.
• Self-reported data -- whether you are relying on pre-existing self-reported data or you are conducting a qualitative research study and gathering the data yourself, self-reported data is limited by the fact that it rarely can be independently verified. In other words, you have to take what people say, whether in interviews, focus groups, or on questionnaires, at face value. However, self-reported data contain several potential sources of bias that should be noted as limitations: (1) selective memory (remembering or not remembering experiences or events that occurred at some point in the past); (2) telescoping [recalling events that occurred at one time as if they occurred at another time]; (3) attribution [the act of attributing positive events and outcomes to one's own agency but attributing negative events and outcomes to external forces]; and, (4) exaggeration [the act of representing outcomes or embellishing events as more significant than is actually suggested from other data].

Possible Limitations of the Researcher

• Access -- if your study depends on having access to people, organizations, or documents and, for whatever reason, access is denied or otherwise limited, the reasons for this need to be described.
• Longitudinal effects -- unlike your professor, who can literally devote years [even a lifetime] to studying a single research problem, the time available to investigate a research problem and to measure change or stability within a sample is constrained by the due date of your assignment. Be sure to choose a topic that does not require an excessive amount of time to complete the literature review, apply the methodology, and gather and interpret the results. If you're unsure, talk to your professor.
• Cultural and other type of bias -- we all have biases, whether we are conscience of them or not. Bias is when a person, place, or thing is viewed or shown in a consistently inaccurate way. It is usually negative, though one can have a positive bias as well. When proof-reading your paper, be especially critical in reviewing how you have stated a problem, selected the data to be studied, what may have been omitted, the manner in which you have ordered events, people, or places and how you have chosen to represent a person, place, or thing, to name a phenomenon, or to use possible words with a positive or negative connotation. Note that if you detect bias in prior research, it must be acknowledged and you should explain what measures were taken to avoid perpetuating bias.
• Fluency in a language -- if your research focuses on measuring the perceived value of after-school tutoring among Mexican-American ESL [English as a Second Language] students, for example, and you are not fluent in Spanish, you are limited in being able to read and interpret Spanish language research studies on the topic. This deficiency should be acknowledged.

Brutus, Stéphane et al. Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations. Journal of Management 39 (January 2013): 48-75; Senunyeme, Emmanuel K. Business Research Methods . Powerpoint Presentation. Regent University of Science and Technology.

Structure and Writing Style

Information about the limitations of your study are generally placed either at the beginning of the discussion section of your paper so the reader knows and understands the limitations before reading the rest of your analysis of the findings, or, the limitations are outlined at the conclusion of the discussion section as an acknowledgement of the need for further study. Statements about a study's limitations should not be buried in the body [middle] of the discussion section unless a limitation is specific to something covered in that part of the paper. If this is the case, though, the limitation should be reiterated at the conclusion of the section.

If you determine that your study is seriously flawed due to important limitations , such as, an inability to acquire critical data, consider reframing it as a pilot study intended to lay the groundwork for a more complete research study in the future. Be sure, though, to specifically explain the ways that these flaws can be successfully overcome in later studies.

But, do not use this as an excuse for not developing a thorough research paper! Review the tab in this guide for developing a research topic . If serious limitations exist, it generally indicates a likelihood that your research problem is too narrowly defined or that the issue or event under study  is too recent and, thus, very little research has been written about it. If serious limitations do emerge, consult with your professor about possible ways to overcome them or how to reframe your study.

When discussing the limitations of your research, be sure to:

• Describe each limitation in detailed but concise terms;
• Explain why each limitation exists;
• Provide the reasons why each limitation could not be overcome using the method(s) chosen to gather the data [cite to other studies that had similar problems when possible];
• Assess the impact of each limitation in relation to  the overall findings and conclusions of your study; and,
• If appropriate, describe how these limitations could point to the need for further research.

Remember that the method you chose may be the source of a significant limitation that has emerged during your interpretation of the results [for example, you didn't ask a particular question in a survey that you later wish you had]. If this is the case, don't panic. Acknowledge it, and explain how applying a different or more robust methodology might address the research problem more effectively in any future study. A underlying goal of scholarly research is not only to prove what works, but to demonstrate what doesn't work or what needs further clarification.

Brutus, Stéphane et al. Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations. Journal of Management 39 (January 2013): 48-75; Ioannidis, John P.A. Limitations are not Properly Acknowledged in the Scientific Literature. Journal of Clinical Epidemiology 60 (2007): 324-329; Pasek, Josh. Writing the Empirical Social Science Research Paper: A Guide for the Perplexed . January 24, 2012. Academia.edu; Structure: How to Structure the Research Limitations Section of Your Dissertation . Dissertations and Theses: An Online Textbook. Laerd.com; What Is an Academic Paper? Institute for Writing Rhetoric. Dartmouth College; Writing the Experimental Report: Methods, Results, and Discussion. The Writing Lab and The OWL. Purdue University.

Writing Tip

Don't Inflate the Importance of Your Findings! After all the hard work and long hours devoted to writing your research paper, it is easy to get carried away with attributing unwarranted importance to what you’ve done. We all want our academic work to be viewed as excellent and worthy of a good grade, but it is important that you understand and openly acknowledge the limitiations of your study. Inflating of the importance of your study's findings in an attempt hide its flaws is a big turn off to your readers. A measure of humility goes a long way!

Another Writing Tip

Negative Results are Not a Limitation!

Negative evidence refers to findings that unexpectedly challenge rather than support your hypothesis. If you didn't get the results you anticipated, it may mean your hypothesis was incorrect and needs to be reformulated, or, perhaps you have stumbled onto something unexpected that warrants further study. Moreover, the absence of an effect may be very telling in many situations, particularly in experimental research designs. In any case, your results may be of importance to others even though they did not support your hypothesis. Do not fall into the trap of thinking that results contrary to what you expected is a limitation to your study. If you carried out the research well, they are simply your results and only require additional interpretation.

Yet Another Writing Tip

A Note about Sample Size Limitations in Qualitative Research

Sample sizes are typically smaller in qualitative research because, as the study goes on, acquiring more data does not necessarily lead to more information. This is because one occurrence of a piece of data, or a code, is all that is necessary to ensure that it becomes part of the analysis framework. However, it remains true that sample sizes that are too small cannot adequately support claims of having achieved valid conclusions and sample sizes that are too large do not permit the deep, naturalistic, and inductive analysis that defines qualitative inquiry. Determining adequate sample size in qualitative research is ultimately a matter of judgment and experience in evaluating the quality of the information collected against the uses to which it will be applied and the particular research method and purposeful sampling strategy employed. If the sample size is found to be a limitation, it may reflect your judgement about the methodological technique chosen [e.g., single life history study versus focus group interviews] rather than the number of respondents used.

Huberman, A. Michael and Matthew B. Miles. Data Management and Analysis Methods. In Handbook of Qualitative Research. Norman K. Denzin and Yvonna S. Lincoln, eds. (Thousand Oaks, CA: Sage, 1994), pp. 428-444.

• << Previous: 8. The Discussion
• Next: 9. The Conclusion >>
• Last Updated: Jul 18, 2023 11:58 AM
• URL: https://library.sacredheart.edu/c.php?g=29803
• QuickSearch
• Library Catalog
• Databases A-Z
• Publication Finder
• Course Reserves
• Citation Linker
• Digital Commons
• Our Website

Research Support

• Ask a Librarian
• Appointments
• Interlibrary Loan (ILL)
• Research Guides
• Databases by Subject
• Citation Help

Using the Library

• Reserve a Group Study Room
• Renew Books
• Honors Study Rooms
• Off-Campus Access
• Library Policies
• Library Technology

User Information

• Grad Students
• Online Students
• COVID-19 Updates
• Staff Directory
• News & Announcements
• Library Newsletter

My Accounts

• Interlibrary Loan
• Staff Site Login

FIND US ON  

Research Limitations & Delimitations

What they are and how they’re different (with examples)

By: Derek Jansen (MBA) | Expert Reviewed By: David Phair (PhD) | September 2022

If you’re new to the world of research, you’ve probably heard the terms “ research limitations ” and “ research delimitations ” being thrown around, often quite loosely. In this post, we’ll unpack what both of these mean, how they’re similar and how they’re different – so that you can write up these sections the right way.

Overview: Limitations vs Delimitations

• Are they the same?
• What are research limitations
• What are research delimitations
• Limitations vs delimitations

First things first…

Let’s start with the most important takeaway point of this post – research limitations and research delimitations are not the same – but they are related to each other (we’ll unpack that a little later). So, if you hear someone using these two words interchangeably, be sure to share this post with them!

Research Limitations

Research limitations are, at the simplest level, the weaknesses of the study , based on factors that are often outside of your control as the researcher. These factors could include things like time , access to funding, equipment , data or participants . For example, if you weren’t able to access a random sample of participants for your study and had to adopt a convenience sampling strategy instead, that would impact the generalizability of your findings and therefore reflect a limitation of your study.

Research limitations can also emerge from the research design itself . For example, if you were undertaking a correlational study, you wouldn’t be able to infer causality (since correlation doesn’t mean certain causation). Similarly, if you utilised online surveys to collect data from your participants, you naturally wouldn’t be able to get the same degree of rich data that you would from in-person interviews .

Simply put, research limitations reflect the shortcomings of a study , based on practical (or theoretical) constraints that the researcher faced. These shortcomings limit what you can conclude from a study, but at the same time, present a foundation for future research . Importantly, all research has limitations , so there’s no need to hide anything here – as long as you discuss how the limitations might affect your findings, it’s all good.

Research Delimitations

Alright, now that we’ve unpacked the limitations, let’s move on to the delimitations .

Research delimitations are similar to limitations in that they also “ limit ” the study, but their focus is entirely different. Specifically, the delimitations of a study refer to the scope of the research aims and research questions . In other words, delimitations reflect the choices you, as the researcher, intentionally make in terms of what you will and won’t try to achieve with your study. In other words, what your research aims and research questions will and won’t include.

As we’ve spoken about many times before, it’s important to have a tight, narrow focus for your research, so that you can dive deeply into your topic, apply your energy to one specific area and develop meaningful insights. If you have an overly broad scope or unfocused topic, your research will often pull in multiple, even opposing directions, and you’ll just land up with a muddy mess of findings .

So, the delimitations section is where you’ll clearly state what your research aims and research questions will focus on – and just as importantly, what they will exclude . For example, you might investigate a widespread phenomenon, but choose to focus your study on a specific age group, ethnicity or gender. Similarly, your study may focus exclusively on one country, city or even organization. As long as the scope is well justified (in other words, it represents a novel, valuable research topic), this is perfectly acceptable – in fact, it’s essential. Remember, focus is your friend.

Need a helping hand?

Conclusion: Limitations vs Delimitations

Ok, so let’s recap.

Research limitations and research delimitations are related in that they both refer to “limits” within a study. But, they are distinctly different. Limitations reflect the shortcomings of your study, based on practical or theoretical constraints that you faced.

Contrasted to that, delimitations reflect the choices that you made in terms of the focus and scope of your research aims and research questions. If you want to learn more about research aims and questions, you can check out this video post , where we unpack those concepts in detail.

Psst... there’s more!

This post was based on one of our popular Research Bootcamps . If you're working on a research project, you'll definitely want to check this out ...

You Might Also Like:

18 Comments

Good clarification of ideas on how a researcher ought to do during Process of choice

Thank you so much for this very simple but explicit explanation on limitation and delimitation. It has so helped me to develop my masters proposal. hope to recieve more from your site as time progresses

Thank you for this explanation – very clear.

Thanks for the explanation, really got it well.

This website is really helpful for my masters proposal

Thank you very much for helping to explain these two terms

I spent almost the whole day trying to figure out the differences

when I came across your notes everything became very clear

thanks for the clearly outlined explanation on the two terms, limitation and delimitation.

Very helpful Many thanks 🙏

Excellent it resolved my conflict .

I would like you to assist me please. If in my Research, I interviewed some participants and I submitted Questionnaires to other participants to answered to the questions, in the same organization, Is this a Qualitative methodology , a Quantitative Methodology or is it a Mixture Methodology I have used in my research? Please help me

How do I cite this article in APA format

Really so great ,finally have understood it’s difference now

Getting more clear regarding Limitations and Delimitation and concepts

I really appreciate your apt and precise explanation of the two concepts namely ; Limitations and Delimitations.

This is a good sources of research information for learners.

thank you for this, very helpful to researchers

Very good explained

Great and clear explanation, after a long confusion period on the two words, i can now explain to someone with ease.

Submit a Comment Cancel reply

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

• Print Friendly

Scientific Research and Methodology : An introduction to quantitative research and statistics

9 research design limitations.

So far, you have learnt to ask a RQ and designs studies. In this chapter , you will learn to identify:

• limitations to internally valid.
• limitations to externally valid.
• limitations to ecologically valid.

9.1 Introduction

The type of study and the research design determine how the results of the study should be interpreted. Ideally, a study would be perfectly externally and internally valid; in practice this is very difficult to achieve. Practically every study has limitations. The results of a study should be interpreted in light of these limitations. Limitations are not necessarily problems .

Limitations generally can be discussed through three components:

• Internal validity (Sect. 6.1 ): Discuss any limitations to internal validity due to the research design (such as identifying possible confounding variables). This is related to the effectiveness of the study within the sample (Sect. 9.2 ).
• External validity (Sect. 5.1 ): Discuss how well the sample represents the intended population. This is related to the generalisability of the study to the intended population (Sect. 9.3 ).
• Ecological validity : Discuss how well the study methods, materials and context approximate the real situation being studied. This is related to the practicality of the results to real life (Sect. 9.4 ).

Some of these limitations are imposed by the type of study. All these issues should be addressed when considering the study limitations.

Almost every study has limitations. Identifying potential limitations, and discussing the likely impact they have on the interpretation of the study results, is important and ethical.

Example 9.1 Delarue et al. ( 2019 ) discuss studies where subjects rate the taste of new food products. They note that taste-testing studies should (p. 78):

... allow generalizing the conclusions obtained with a consumer sample [...] to the general targeted population [i.e., external validity]... tests should be reliable in terms of accuracy and replicability [i.e., internal validity].

However, even with good internal and external validity, these studies often result in a 'high rate of failures of new launched products'. That is, the studies do not replicate the real world, and so lack ecological validity .

9.2 Limitations: internal validity

Internal validity refers to the extent to which a cause-and-effect relationship can be established in a study, eliminating other possible explanations (Sect. 6.1 ). A discussion of the limitations of internal validity should cover, as appropriate: possible confounding variables; the impact of the Hawthorne, observer, placebo and carry-over effects; the impact of any other design decisions.

If any of these issues are likely to compromise internal validity, the implications on the interpretation of the results should be discussed. For example, if the participants were not blinded, this should be clearly stated, and the conclusion should indicate that the individuals in the study may have behaved differently than usual.

Example 9.2 (Study limitations) Axmann et al. ( 2020 ) randomly allocated Ugandan farmers to receive, or not receive, hybrid maize seeds. One potential threat to internal validity was that farmers receiving the hybrid seeds could share their seeds with their neighbours.

Hence, the researchers contacted the \(75\) farmers allocated to receive the hybrid seeds; none of the contacted farmers reported selling or giving seeds to other farmers. This extra step increased the internal validity of the study.

Maximizing internal validity in observational studies is more difficult than in experimental studies (e.g., random allocation is not possible). The internal validity of experimental studies involving people is often compromised because people must be informed that they are participating in a study.

Example 9.3 (Internal validity) In a study of the hand-hygiene practices of paramedics ( Barr et al. 2017 ) , self -reported hand-hygiene practices were very different than what was reported by peers . That is, how people self-report their behaviours may not align with how they actually behave, which influenced the internal validity of the study.

A study evaluated using a new therapy on elderly men, and listed some limitations of their study:

... the researcher was not blinded and had prior knowledge of the research aims, disease status, and intervention. As such, these could all have influenced data recording [...] The potential of reporting bias and observer bias could be reduced by implementing blinding in future studies. --- Kabata-Piżuch et al. ( 2021 ) , p. 10

9.3 Limitations: external validity

External validity refers to the ability to generalise the findings made from the sample to the entire intended population (Sect.  5.1 ). For a study to be externally valid, it must first be internally valid: if the study of not effective in the sample studied (i.e., internally valid), the results may not apply to the intended population either.

External validity refers to how well the sample is likely to represent the intended population in the RQ.

If the population is Iowans, then the study is externally valid if the sample is representative of Iowans The results do not have to apply to people in the rest of the United States (though this can be commented on, too). The intended population is Iowans .

External validity depends on how the sample was obtained. Results from random samples (Sects.  5.5 to  5.9 ) are likely to generalise to the population and be externally valid. (The analyses in this book assume all samples are simple random samples .) Furthermore, results from approximately representative samples (Sect.  5.10 ) may generalise to the population and be externally valid if those in the study are not obviously different than those not in the study.

Example 9.4 (External validity) A New Zealand study ( Gammon et al. 2012 ) identified (for well-documented reasons) a population of interest: 'women of South Asian origin living in New Zealand' (p. 21). The women in the sample were 'women of South Asian origin [...] recruited using a convenience sample method throughout Auckland' (p. 21).

The results may not generalise to the intended population ( all women of South Asian origin living in New Zealand) because all the women in the sample came from Auckland, and the sample was not a random sample from this population anyway. The study was still useful however!

Example 9.5 (Using biochar) Farrar et al. ( 2018 ) studied growing ginger using biochar on one farm at Mt Mellum, Australia. The results may only generalise to growing ginger at Mt Mellum, but since ginger is usually grown in similar types of climates and soils, the results may apply to other ginger farms also.

9.4 Limitations: ecological validity

The likely practicality of the study results in the real world should also be discussed. This is called ecological validity .

Definition 9.1 (Ecological validity) A study is ecologically valid if the study methods, materials and context closely approximate the real situation of interest.

Studies don't need to be ecologically valid to be useful; much can be learnt under special conditions, as long as the potential limitations are understood when applying the results to the real world. The ecological validity of experimental studies may be compromised because the experimental conditions are sometimes artificially controlled (for good reason).

Example 9.6 (Ecological validity) Consider a study to determine the proportion of people that buy coffee in a reusable cup. People could be asked about their behaviour. This study may not be ecologically valid, as how people act may not align with how they say they will act.

An alternative study could watch people buy coffees at various coffee shops, and record what people do in practice. This second study is more likely to be ecologically valid , as real-world behaviour is observed.

A study observed the effect of using high-mounted rear brake lights ( Kahane and Hertz 1998 ) , which are now commonplace. The American study showed that such lights reduced rear-end collisions by about \(50\) %. However, after making these lights mandatory, rear-end collisions reduced by only \(5\) %. Why?

9.5 Limitations: study types

Experimental studies, in general, have higher internal validity than observational studies, since more of the research design in under the control of the researchers; for example, random allocation of treatments is possible to minimise confounding.

Only well-conducted experimental studies can show cause-and-effect relationships.

However, experimental studies may suffer from poor ecological validity; for instance, laboratory experiments are often conducted under controlled temperature and humidity. Many experiments also require that people be told about being in a study (due to ethics), and so internal validity may be comprised (the Hawthorne effect).

Example 9.7 (Retrofitting) giandomenico2022systematic studied retro-fitting houses with energy-saving devices, and found large discrepancies in savings for observational studies ( \(12.2\) %) and experimental studies ( \(6.2\) %). The authors say that 'this finding reinforces the importance of using study designs with high internal validity to evaluate program savings' (p. 692).

9.6 Chapter summary

The limitations in a study need to be identified, and may be related to:

• internal validity (effectiveness): how well the study is conducted within the sample, isolating the relationship of interest.
• external validity (generalisability): how well the sample results are likely to apply to the intended population.
• ecological validity (practicality): how well the results may apply to the real-world situation.

Many of the limitations are a results of the type of study.

9.7 Quick review questions

Are the following statements true or false ?

• When interpreting the results of a study, the steps taken to maximize internal validity should be evaluated TRUE FALSE
• If studies are not externally valid, then they are not useful. TRUE FALSE
• When interpreting the results of a study, the steps taken to maximize external validity do not need to be evaluated TRUE FALSE
• When interpreting the results of a study, ecological validity is about the impact of the study on the environment. TRUE FALSE

9.8 Exercises

Answers to odd-numbered exercises are available in App.  E .

Exercise 9.1 A research study examined how people can save energy through lighting choices ( Gentile 2022 ) . The study states (p. 9) that the results 'are limited to the specific study and cannot be easily projected to other similar settings'.

What type of validity is being discussed here?

Exercise 9.2 Fill the blanks with the correct word: internal , external or ecological .

When interpreting the results of studies, we consider the practicality ( internal external ecological validity), the generalizability ( internal external ecological validity) and the effectiveness ( internal external ecological validity).

Exercise 9.3 A student project asked if 'the percentage of word retention higher in male students than female students?' When discussing external validity , the students stated:

We cannot say whether or not the general public have better or worse word retention compared to the students that we will be studying.

Why is the statement not relevant in a discussion of external validity?

Exercise 9.4 Yeh et al. ( 2018 ) conducted an experimental study to 'determine if using a parachute prevents death or major traumatic injury when jumping from an aircraft'.

The researchers randomised \(23\) volunteers into one of two groups: wearing a parachute, or wearing an empty backpack. The response variable was a measurement of death or major traumatic injury upon landing. From the study, death or major injury was the same in both groups (0% for each group). However, the study used 'small stationary aircraft on the ground, suggesting cautious extrapolation to high altitude jumps' (p. 1).

Comment on the internal, external and ecological validity.

Exercise 9.5 A study examined how well hospital patients sleep at night ( Delaney et al. 2018 ) . The researchers state that 'convenience sampling was used to recruit patients' (p. 2). Later, the researchers state (p. 7):

... while most healthy individuals sleep primarily or exclusively at night, it is important to consider that patients requiring hospitalization will likely require some daytime nap periods. This study looks at sleep only in the night-time period \(22\) : \(00\) -- \(07\) : \(00\) h, without the context of daytime sleep considered.

Discuss these issues using the language introduced in this chapter.

Exercise 9.6 Botelho et al. ( 2019 ) examined the food choices made when subjects were asked to shop for ingredients to make a last-minute meal. Half were told to prepare a 'healthy meal', and the other half told just to prepare a 'meal'. The authors stated (p. 436):

Another limitation is that results report findings from a simulated purchase. As participants did not have to pay for their selection, actual choices could be different. Participants may also have not behaved in their usual manner since they were taking part in a research study, a situation known as the Hawthorne effect.

What type of limitation is being discussed?

Exercise 9.7 Johnson et al. ( 2018 ) studied the use of over-the-counter menthol cough-drops in people with a cough. One conclusion from the observational study of \(548\) people was that, taking 'too many cough drops [...] may actually make coughs more severe', as one author explained in an interview about the study Critique this statement.

Research Limitations

It is for sure that your research will have some limitations and it is normal. However, it is critically important for you to be striving to minimize the range of scope of limitations throughout the research process.  Also, you need to provide the acknowledgement of your research limitations in conclusions chapter honestly.

It is always better to identify and acknowledge shortcomings of your work, rather than to leave them pointed out to your by your dissertation assessor. While discussing your research limitations, don’t just provide the list and description of shortcomings of your work. It is also important for you to explain how these limitations have impacted your research findings.

Your research may have multiple limitations, but you need to discuss only those limitations that directly relate to your research problems. For example, if conducting a meta-analysis of the secondary data has not been stated as your research objective, no need to mention it as your research limitation.

Research limitations in a typical dissertation may relate to the following points:

1. Formulation of research aims and objectives . You might have formulated research aims and objectives too broadly. You can specify in which ways the formulation of research aims and objectives could be narrowed so that the level of focus of the study could be increased.

2. Implementation of data collection method . Because you do not have an extensive experience in primary data collection (otherwise you would not be reading this book), there is a great chance that the nature of implementation of data collection method is flawed.

3. Sample size. Sample size depends on the nature of the research problem. If sample size is too small, statistical tests would not be able to identify significant relationships within data set. You can state that basing your study in larger sample size could have generated more accurate results. The importance of sample size is greater in quantitative studies compared to qualitative studies.

4. Lack of previous studies in the research area . Literature review is an important part of any research, because it helps to identify the scope of works that have been done so far in research area. Literature review findings are used as the foundation for the researcher to be built upon to achieve her research objectives.

However, there may be little, if any, prior research on your topic if you have focused on the most contemporary and evolving research problem or too narrow research problem. For example, if you have chosen to explore the role of Bitcoins as the future currency, you may not be able to find tons of scholarly paper addressing the research problem, because Bitcoins are only a recent phenomenon.

5. Scope of discussions . You can include this point as a limitation of your research regardless of the choice of the research area. Because (most likely) you don’t have many years of experience of conducing researches and producing academic papers of such a large size individually, the scope and depth of discussions in your paper is compromised in many levels compared to the works of experienced scholars.

You can discuss certain points from your research limitations as the suggestion for further research at conclusions chapter of your dissertation.

My e-book,  The Ultimate Guide to Writing a Dissertation in Business Studies: a step by step assistance  offers practical assistance to complete a dissertation with minimum or no stress. The e-book covers all stages of writing a dissertation starting from the selection to the research area to submitting the completed version of the work within the deadline. John Dudovskiy

• Open access
• Published: 27 May 2024

Optimizing double-layered convolutional neural networks for efficient lung cancer classification through hyperparameter optimization and advanced image pre-processing techniques

• M. Mohamed Musthafa 1 ,
• I. Manimozhi 2 ,
• T. R. Mahesh 3 &
• Suresh Guluwadi 4  

BMC Medical Informatics and Decision Making volume  24 , Article number:  142 ( 2024 ) Cite this article

Metrics details

Lung cancer remains a leading cause of cancer-related mortality globally, with prognosis significantly dependent on early-stage detection. Traditional diagnostic methods, though effective, often face challenges regarding accuracy, early detection, and scalability, being invasive, time-consuming, and prone to ambiguous interpretations. This study proposes an advanced machine learning model designed to enhance lung cancer stage classification using CT scan images, aiming to overcome these limitations by offering a faster, non-invasive, and reliable diagnostic tool. Utilizing the IQ-OTHNCCD lung cancer dataset, comprising CT scans from various stages of lung cancer and healthy individuals, we performed extensive preprocessing including resizing, normalization, and Gaussian blurring. A Convolutional Neural Network (CNN) was then trained on this preprocessed data, and class imbalance was addressed using Synthetic Minority Over-sampling Technique (SMOTE). The model’s performance was evaluated through metrics such as accuracy, precision, recall, F1-score, and ROC curve analysis. The results demonstrated a classification accuracy of 99.64%, with precision, recall, and F1-score values exceeding 98% across all categories. SMOTE significantly enhanced the model’s ability to classify underrepresented classes, contributing to the robustness of the diagnostic tool. These findings underscore the potential of machine learning in transforming lung cancer diagnostics, providing high accuracy in stage classification, which could facilitate early detection and tailored treatment strategies, ultimately improving patient outcomes.

Peer Review reports

Introduction

Lung cancer stands as a formidable global health challenge, consistently ranking as one of the leading causes of cancer-related mortality worldwide. It is characterized by the uncontrolled growth of abnormal cells in one or both lungs, typically in the cells lining the air passages. Unlike normal cells, these cancerous cells do not develop into healthy lung tissue; instead, they divide rapidly and form tumors that disrupt the lung’s primary function: oxygen exchange.

The global impact of lung cancer is staggering, with millions of new cases diagnosed annually. Its high mortality rate is primarily due to late-stage detection, where the cancer has progressed to an advanced stage or metastasized to other body parts, significantly diminishing the effectiveness of treatment modalities. Thus, early and accurate diagnosis of lung cancer is paramount in improving patient prognoses, extending survival rates, and enhancing the quality of life for affected individuals.

The primary cause of lung cancer is cigarette smoking, which exposes the lungs to carcinogenic substances that can damage the cells’ DNA and lead to cancer. Other risk factors for lung cancer include exposure to secondhand smoke, radon gas, asbestos, air pollution, and a family history of lung cancer.

Symptoms of lung cancer can vary but may include persistent coughing, chest pain, shortness of breath, hoarseness, coughing up blood, unexplained weight loss, and fatigue. However, lung cancer may not cause symptoms in its initial stages, which is why early detection through screening is crucial for improving outcomes.

Diagnosis of lung cancer typically involves imaging tests such as chest X-rays, CT scans, and PET scans to visualize the lungs and detect any abnormalities. A biopsy, where a small sample of lung tissue is taken and examined under a microscope, is usually needed to confirm the diagnosis.

Treatment options for lung cancer depend on several factors, including the type and stage of the cancer, as well as the patient’s overall health and preferences. Treatment may include surgery to remove the tumor, chemotherapy, radiation therapy, targeted therapy, immunotherapy, or a combination of these approaches.

Lung cancer is a critical condition that necessitates immediate medical care. Detecting it early, along with improvements in treatment methods, has enhanced the prognosis for numerous patients. However, the most effective strategy to avoid lung cancer is to stop smoking and minimize contact with additional risk elements. Figure  1 displays some example images of lung cancer tests.

Sample images of lung cancer

Current diagnostic techniques for lung cancer involve various approaches, such as biopsies, CT scans, chest X-rays, PET scans, and MRI, among others [ 1 ]. While these methods are invaluable in the diagnostic process, they come with certain limitations. For instance, biopsies, while definitive, are invasive and carry risks of complications. Less invasive imaging methods such as X-rays or CT scans might produce false positives or negatives, potentially causing unwarranted stress or delays in treatment.

Moreover, the interpretation of these diagnostic tests heavily relies on the expertise of the clinician, introducing a degree of subjectivity and potential for human error. There’s also the challenge of early-stage lung cancer, which often presents very subtle changes not always detectable with conventional imaging techniques [ 2 ].

This context highlights the critical need for advanced diagnostic tools capable of overcoming these challenges. This study aims to address these issues by developing a machine learning model using Convolutional Neural Networks (CNNs) to enhance the precision and effectiveness of lung cancer stage classification from CT scans. By automating and refining the diagnostic process, the proposed model seeks to mitigate the limitations of traditional methods, offering a faster, non-invasive, and more reliable diagnostic alternative.

The impact of this study is significant: the model’s high accuracy in classifying lung cancer stages promises to revolutionize clinical diagnostics, facilitating early detection and enabling tailored treatment strategies. This advancement has the potential to improve patient outcomes by allowing for timely intervention and more effective management of lung cancer, ultimately contributing to reduced mortality rates and enhanced patient care.

The objective of this research paper is to:

Develop a machine learning model utilizing Convolutional Neural Networks (CNNs) for lung cancer stage classification based on CT scans.

Bridge existing diagnostic deficiencies by providing clinicians with a tool for expedited and precise decision-making in lung cancer management.

Contribute to improved patient outcomes through enhanced diagnostic accuracy and early detection capabilities.

The paper is organized as follows: Initially, the Literature Review explores existing research on lung cancer diagnostics, highlighting advancements and limitations, and sets the foundation for the proposed methodology. Subsequently, the Materials and methods section describes the dataset, preprocessing steps, model architecture, training process, and evaluation metrics in detail. The Results section then presents the study’s findings, including model performance metrics and comparative analysis with existing methods. This is followed by the Discussion, which interprets the results, discusses implications for clinical practice, addresses limitations, and suggests future research directions. Finally, the Conclusion summarizes the main findings and their relevance within the broader scope of lung cancer diagnostics, supported by a comprehensive list of References to provide credit and enable readers to explore the research background further.

Through this structured approach, the paper aims to contribute meaningful insights to the field of medical imaging and machine learning, offering a novel tool for the early and accurate diagnosis of lung cancer.

Literature review

The literature surrounding lung cancer diagnostics encompasses various methodologies, ranging from traditional imaging techniques to more advanced approaches such as machine learning. This review aims to explore existing research in this area, highlighting both the advancements made and the limitations faced, ultimately setting the foundation for the proposed machine learning-based methodology.

Diagnosis of lung cancer using CT scans

The utilization of Computed Tomography (CT) scans in lung cancer diagnosis has been a cornerstone in the medical field, offering high-resolution images that are pivotal for detecting and monitoring various stages of lung tumors [ 3 ]. Over the years, numerous studies have underscored the importance of CT scans in identifying nodules that could potentially be malignant, with a particular focus on low-dose CT scans, which have become a standard in screening programs, especially for high-risk populations. Such studies underscore the superior sensitivity of CT scans in identifying early-stage lung cancer, a significant advancement over other imaging methods like chest X-rays, which may overlook smaller, subtler lesions.

Despite the advancements, the interpretation of CT scans remains a significant challenge. Radiologists need to discern between benign and malignant nodules, an endeavor complicated by the presence of various artifacts and benign conditions like scars or inflammatory diseases, which can mimic the appearance of cancerous nodules [ 4 , 5 ].

Machine learning approaches in lung cancer detection and classification

The integration of machine learning, particularly deep learning techniques, into the analysis of CT images has established a groundbreaking paradigm in the identification and classification of lung cancer. Convolutional Neural Networks (CNNs) are spearheading this transformation by providing a framework for automated extraction and categorization of features directly from the images. This advancement marks a substantial stride in augmenting the accuracy and effectiveness of lung cancer diagnostics, thus facilitating more precise and timely interventions.

Binary classification models

Early studies primarily focused on binary classification, distinguishing between malignant and non-malignant nodules. CNNs, through their layered architecture, have demonstrated the ability to learn complex patterns in imaging data, surpassing traditional computer vision techniques in accuracy and reliability [ 6 , 7 ].

Multi-class classification models

Recent advancements have moved towards more nuanced multi-class classification models that categorize nodules into various cancer stages or types. This granularity is crucial for treatment planning and prognosis, offering a more detailed understanding of the disease’s progression [ 8 ].

Transfer learning

Given the challenges of assembling large annotated medical imaging datasets, transfer learning has become a popular approach. Models pre-trained on vast, non-medical image datasets are fine-tuned on smaller medical imaging datasets, leveraging learned features to improve performance in the medical domain [ 9 ].

Data augmentation

To address the issue of restricted training data, strategies such as rotation, scaling, and flipping are commonly employed for data augmentation, effectively expanding the training dataset artificially. These methods bolster the model’s resilience and its ability to generalize from a limited number of examples [ 10 ].

Segmentation models

Deep learning models extend their utility beyond mere classification; they are also employed in segmentation tasks, delineating the precise boundaries of nodules, which is vital for assessing tumor size and growth over time. U-Net, a type of CNN, is particularly noted for its effectiveness in medical image segmentation [ 11 ].

In Table 1  a few of the studies which have been done in this field are given.

Gaps in current research

Despite significant advancements in lung cancer diagnostics, several critical gaps remain in the current research landscape. Many existing models are trained on datasets lacking diversity in demographics, scanner types, and image acquisition parameters, which can limit their generalizability across different populations and clinical settings. This limitation underscores the need for more comprehensive and diverse datasets to enhance the robustness of diagnostic models. Additionally, the “black box” nature of deep learning models poses a challenge for clinical adoption, as there is a growing demand for models that not only predict accurately but also provide insights into the reasoning behind their predictions. This issue of interpretability is crucial for gaining the trust of clinicians and integrating these models into clinical workflows effectively. Furthermore, the transition from research to clinical practice is slow, with models requiring not just technological solutions but also addressing regulatory, ethical, and practical considerations to facilitate their integration into routine medical care. Another critical gap is the need for models capable of longitudinal analysis, which can analyze changes in lung nodules over time, providing a dynamic assessment that aligns more closely with clinical needs. Addressing these gaps, this study introduces a comprehensive CNN model trained on a diverse and extensive dataset, encompassing various stages of lung cancer. The model is designed for multi-class classification, offering detailed insights critical for personalized treatment strategies. Emphasis is placed on the interpretability of the model, aiming to provide clinicians with understandable and actionable information. By demonstrating the model’s effectiveness in a clinical setting, this research contributes to the ongoing effort to integrate advanced machine learning techniques into the realm of lung cancer diagnosis and treatment.

Addressing these gaps, this study introduces a comprehensive CNN model trained on a diverse and extensive dataset, encompassing various stages of lung cancer. The model is designed for multi-class classification, offering detailed insights critical for personalized treatment strategies. Emphasis is placed on the interpretability of the model, aiming to provide clinicians with understandable and actionable information. By demonstrating the model’s effectiveness in a clinical setting, this research contributes to the ongoing effort to integrate advanced machine learning techniques into the realm of lung cancer diagnosis and treatment.

Materials and methods

This section delineates the comprehensive methodology employed to construct and validate a convolutional neural network (CNN) model for the classification of lung cancer stages using the IQ-OTHNCCD lung cancer dataset. The approach encompasses dataset acquisition, application of preprocessing methodologies, formulation of the model architecture, delineation of training procedures, and determination of evaluation metrics to ensure a comprehensive and reliable analysis. The workflow of the proposed model is visually depicted in Fig.  2 .

Workflow of the proposed model

Dataset description and preprocessing

The IQ-OTHNCCD lung cancer dataset, integral to this study, is painstakingly curated to facilitate the creation and validation of machine learning models aimed at identifying and classifying lung cancer stages. This dataset encompasses a vast collection of CT scan images essential for advancing diagnostic capabilities in the field of lung cancer.

This dataset comprises CT scan images, comprising a diverse and comprehensive range of cases, covering various stages of lung cancer, including benign, malignant, and normal cases. This diversity is essential for training robust models capable of generalizing well across the spectrum of lung cancer manifestations, enabling effective diagnostic applications. In Table 2  a brief description of the dataset has been given.

Based on Table  2 , to provide visual insights of the data Fig.  3 delves into the same aspects.

Dataset description

Annotating and labeling each image meticulously, medical professionals from the Iraq-Oncology Teaching Hospital/National Center for Cancer Diseases have ensured the dataset’s reliability. Annotations categorize images into one of three classes: benign, malignant, or normal. Such granular labeling establishes a solid ground truth essential for training and assessing the model, enhancing the dataset’s utility in research and clinical applications.

Characterized by high quality and consistency, the CT scans adhere to standardized imaging protocols, guaranteeing reliability and accuracy. However, variations in image dimensions necessitate preprocessing to standardize inputs for neural networks. These steps ensure that the model processes uniform data, enhancing its performance and generalizability across diverse datasets. The images of the dataset ratio are checked using Eq.  1 .

Preprocessing steps are pivotal in preparing data for effective model training, including:

Resizing: Resizing images to a uniform dimension ensures consistency in input size for CNNs, optimizing model performance.

Normalization: Normalizing pixel values to a scale of 0 to 1 expedites model convergence during training, facilitating efficient learning. It is achieved using Eq.  2 .

Augmentation: Utilizing data augmentation methods like rotation, flipping, and scaling improves the model’s robustness and helps prevent overfitting by effectively enlarging the dataset size.

Splitting: Partitioning the dataset into training, validation, and test sets is crucial for facilitating effective model training and evaluation, thereby ensuring the model’s ability to generalize and perform accurately on unseen data.

In this process, CNN is trained using the preprocessed dataset to adeptly extract features from CT scan images and accurately classify the stages of lung cancer. The dataset’s diversity and quality are pivotal in enabling the model to learn nuanced features and patterns associated with various lung cancer stages, underscoring its significance in advancing diagnostic accuracy and efficiency.

The IQ-OTHNCCD lung cancer dataset serves as the cornerstone for developing machine learning models that enhance early detection and classification of lung cancer. Through meticulous curation and rigorous preprocessing, this dataset showcases the transformative potential of AI in healthcare, underscoring its role in improving diagnostic accuracy and efficiency.

• Image preprocessing

The preprocessing of images stands as a pivotal stage in the pipeline of developing a machine learning model, especially when handling medical imaging data like the IQ-OTHNCCD lung cancer dataset. This procedure comprises several crucial steps, each tailored to convert the raw CT scan images into a format conducive to effective analysis by a convolutional neural network (CNN).

Initially, image resizing is conducted. Given the inherent variability in the dimensions of CT scans, it is imperative to standardize the size of all images to ensure consistent input to the CNN. Resizing is performed while preserving the aspect ratio to avoid distortion, typically scaling down to a fixed size (e.g., 256 × 256 pixels). This uniformity is vital for the neural network to process and glean insights from the data effectively, as it necessitates a consistent input size [ 21 ].

Some pre-processed images to enhance the accessibility has been provided in Fig.  4 .

Pre-processed images

Following resizing, normalization of pixel values is performed. CT scans, by nature, contain a wide range of pixel intensities, which can adversely affect the training process of a CNN due to the varying scales of image brightness and contrast. Normalization is a crucial preprocessing step in image analysis that adjusts the pixel values to fall within a specific range, commonly 0 to 1 or -1 to 1. This adjustment is typically achieved by dividing the pixel values by the maximum possible value, which is 255 for 8-bit images. Such a normalization process ensures that the model can train faster and more efficiently. This step ensures that the model trains faster and more effectively, as small, standardized values facilitate quicker convergence during the optimization process.

Gaussian blur is then applied as an additional preprocessing step. This technique, which employs a Gaussian kernel to smooth the image, is instrumental in reducing image noise and mitigating the effects of minor variations and artifacts in the scans. By doing so, the model’s focus is directed toward the salient features relevant to lung cancer classification, rather than being distracted by irrelevant noise or details. Gaussian blur operates by convolving the image with a Gaussian function, effectively averaging the pixel values within a specified radius. This process smoothens the image, reducing high-frequency components and noise, which can otherwise lead to overfitting or distraction during the training of the CNN.

In the context of lung cancer CT scans, Gaussian blur helps to highlight the important structural elements of the lungs and nodules while suppressing irrelevant details that could complicate the model’s learning process. By smoothing the images, Gaussian blur enhances the model’s ability to generalize by focusing on the more significant, lower-frequency features of the image, such as the shape and size of nodules, rather than being confounded by small variations or noise. This is particularly beneficial in medical imaging, where the presence of noise and artifacts can obscure critical diagnostic features.

The application of Gaussian blur can also aid in generalizing the model, preventing overfitting to the high-frequency noise present in the training set. It is achieved using Eq.  3 and the SMOTE ratio through Eq.  4 .

These are the preprocessing steps collectively enhance the quality and consistency of the input data, enabling the CNN to focus on learning meaningful, discriminative features from the CT images [ 22 ]. By ensuring that the images are appropriately resized, normalized, and filtered, the model is better equipped to identify the subtle nuances associated with different stages of lung cancer, thereby improving its diagnostic accuracy and reliability. Through meticulous image preprocessing, the foundation is laid for developing a robust machine learning model capable of contributing significantly to the field of medical imaging and diagnostics.

Deep learning model

The model architecture utilized in this study is a Convolutional Neural Network (CNN), renowned for its effectiveness in various image analysis tasks, notably in the domain of medical image processing. In this study, we utilized a Convolutional Neural Network (CNN) architecture, known for its effectiveness in analyzing images, particularly in medical contexts like lung cancer diagnosis from CT scans. Let’s break down how it works in simpler terms. First, the input layer takes in images resized to a standard size of 256 × 256 pixels, in black and white. This consistency helps the CNN learn efficiently. Then comes the first convolutional layer, where the model looks for basic patterns like edges and textures using small 3 × 3 filters. After that, a process called max pooling reduces the image’s size, focusing on the most important features. This step helps the model generalize better and ignore noise. We repeat this process with another convolutional layer to capture more complex patterns. The flattened layer turns the extracted features into a format the model can understand. Next, a fully connected layer reasons based on these features, helping with the final classification. The output layer then gives probabilities for each class (benign, malignant, or normal). Throughout training, we used the Adam optimizer to adjust learning rates and manage gradients effectively. Additionally, we applied a technique called SMOTE to balance our dataset, ensuring the model learned from all classes equally. By carefully designing our CNN architecture and incorporating these steps, we aimed to create a model that can accurately classify lung cancer stages from CT scans.

Input layer : The input layer accepts images resized to 256 × 256 pixels, maintaining a single channel (grayscale), resulting in an input shape of (256, 256, 1).

First convolutional layer : This layer consists of 64 filters of size 3 × 3, using a ReLU (Rectified Linear Unit) activation function. The choice of 64 filters is aimed at capturing a broad array of features from the input image, while the 3 × 3 filter size is standard for capturing spatial relationships in the image data. The equation involved are given in Eqs.  5 and 6 .

First max pooling layer : Following the convolutional layer, the model incorporates a max pooling layer with a 2 × 2 pool size. This layer serves to decrease the spatial dimensions of the feature maps, which not only helps in reducing the computational load but also enhances the model’s generalization capabilities. By focusing on the most prominent features, max pooling ensures that the model does not overfit to the noise in the training data. It is done using Eq.  7 .

Second convolutional layer : Another set of 64 filters is applied, like the first convolutional layer, to further refine the feature extraction. This layer also uses a 3 × 3 kernel and is followed by a ReLU activation. It is achieved using Eqs.  8 and 9 .

Second max pooling layer : This layer additionally decreases the size of the feature maps, aiding in the prevention of overfitting and lessening the computational burden.

Flattening : The feature maps are flattened into a single vector to prepare for the fully connected layers, facilitating the transition from convolutional layers to dense layers.

Fully connected layer : A dense layer with 16 neurons is used, providing a high-level reasoning based on the extracted features. This layer utilizes a linear activation function to allow for a range of linear responses. The equations helping in this are given in Eqs.  10 and 11 .

Output layer : The final layer of the model contains three neurons, each representing one of the classes: benign, malignant, and normal. It uses a SoftMax activation function, which is selected because it provides a probability distribution across these three classes, making it . involved are given in Eqs.  12 and 13 .

Optimizer : The Adam optimizer is used due to its effectiveness in managing sparse gradients and its ability to adapt learning rates, which enhance the convergence speed during training. The equation involved in this is given in Eq.  14 .

CNN is chosen for its proven efficacy in image classification tasks, particularly its ability to learn hierarchical patterns in data. In medical imaging, CNNs have demonstrated success in identifying subtle patterns that are indicative of various pathologies, making them ideal for this application. The sequential model with convolutional layers followed by pooling layers allows for the extraction and down sampling of features, which is critical for capturing relevant information from medical images.

The Synthetic Minority Over-Sampling Technique (SMOTE) represents an innovative strategy devised to address the issue of class imbalance within the dataset. Class imbalance poses a substantial risk of biasing the model’s performance, particularly in medical datasets where one class may be underrepresented. SMOTE functions by creating synthetic samples within the feature space of the minority class, drawing inspiration from the feature space of its nearest neighbors. This process aids in rectifying class imbalances and ensuring more equitable representation during model training.

Filter mapping of a sample image is shown in Fig.  5 to make it more sound about the interoperability of the model.

In this research:

Application of SMOTE : SMOTE is applied only to the training data to prevent information leakage and to promote robust generalization on unseen data. It balances the dataset by augmenting the minority classes, ensuring that the model does not become biased toward the majority class.

Impact on model performance : By addressing the class imbalance, SMOTE helps in improving the model’s sensitivity towards the minority class, which is crucial in medical diagnostics, as overlooking a positive case can have serious implications.

Considerations : While SMOTE can significantly improve model performance in cases of class imbalance, it’s essential to monitor for overfitting, as the synthetic samples may cause the model to overgeneralize from the minority class.

The algorithm for the proposed model is presented in Algorithm 1.

Algorithm 1: Proposed algorithm for the methodology

As per the algorithm in the initial convolutional layers of the model, two sets of convolutional layers followed by max-pooling layers play a pivotal role in feature detection. Utilizing a standard 3 × 3 kernel size allows the model to discern small, localized features within CT scan images. By stacking these convolutional layers before applying max pooling, the model effectively captures intricate patterns such as edges, textures, and shapes, crucial for distinguishing between benign, malignant, and normal lung tissue. The ReLU activation function is employed in these convolutional layers due to its effectiveness in introducing non-linearity, enabling the model to learn complex patterns efficiently. Additionally, max pooling is utilized to downsample the feature maps, reducing computational load and enhancing robustness to image variations, thereby improving translational invariance. Following feature extraction, the model flattens the output and transitions to dense layers, condensing learned information into abstract representations. The final layer consists of three neurons, representing the three classes under consideration, employing the SoftMax activation function to transform logits into probabilities, thereby providing insights into the model’s confidence regarding each class. Throughout the compilation and training phases, the Adam optimizer and sparse categorical crossentropy loss function, as depicted by Eq.  15 , are chosen due to their adaptive learning rate features and appropriateness for classification objectives. Validation on an independent dataset is crucial for detecting overfitting and refining hyperparameters.

In the training phase, SMOTE is strategically applied to create a balanced dataset representative of all classes, crucial for generalizing well across various lung tissue conditions, especially in medical datasets where class imbalance may exist.

Training and validation

Throughout the training and validation phases of the deep learning model, meticulous steps are taken to ensure that the model not only learns effectively from the training data but also demonstrates robust generalization capabilities when presented with new, unseen data. This phase plays a pivotal role in evaluating the model’s proficiency in accurately classifying lung cancer stages from CT scans.

The training process initiates with the segmentation of the dataset into distinct training and validation subsets. This segmentation is performed in a stratified manner to guarantee that each subset encompasses a balanced representation of the various classes. Such stratification is essential for maintaining consistency and mitigating biases, particularly in light of the class imbalance addressed by SMOTE during training. Approximately 80% of the data is allocated for training purposes, while the remaining 20% is reserved for validation.

Subsequent to the data segmentation, the training commences with the utilization of a batch size of 8. The selection of a smaller batch size is deliberate, aiming to facilitate more precise and nuanced updates to the model’s weights during each iteration, thereby potentially enhancing generalization. Nonetheless, it is imperative to strike a balance between this granularity and computational efficiency, as smaller batch sizes may prolong the training duration.

The number of epochs is predetermined to be 12, indicating the total number of complete passes that the learning algorithm will undertake across the entire training dataset. This choice represents a delicate balance between underfitting and overfitting; insufficient epochs may hinder the model’s learning process, whereas excessive epochs may result in the model memorizing the training data, consequently impairing its ability to generalize effectively. The progression of training and validation loss and accuracy across epochs is visualized in Fig.  6 .

Training and validation loss and accuracy

During training, the model’s performance is continuously evaluated using a comprehensive set of performance metrics assessed against the validation set. These metrics encompass accuracy, precision, recall, and F1-score, all of which are instrumental in comprehending the model’s strengths and weaknesses in classifying each lung cancer stage. Accuracy furnishes a broad overview of the model’s overall performance, while precision and recall delve deeper into its class-specific performance, a critical consideration in medical diagnostics where false negatives and false positives carry significant consequences. The F1-score serves to harmonize precision and recall, furnishing a unified metric to gauge the model’s equilibrium between these two facets.

Moreover, the validation process incorporates a confusion matrix and ROC curves to furnish a more granular analysis of the model’s performance across diverse thresholds and classes. The confusion matrix delineates the model’s true positives, false positives, false negatives, and true negatives, offering a snapshot of its classification capabilities. Meanwhile, ROC curves and the corresponding AUC (Area Under the Curve) provide insights into the model’s capacity to discriminate between classes at varying threshold settings, a crucial consideration for refining the model’s decision boundary.

In our quest to maximize the performance of our Convolutional Neural Network (CNN) model for lung cancer classification, we meticulously fine-tuned several critical hyperparameters that play pivotal roles in shaping the learning process and ultimately, the model’s accuracy. Specifically, we focused on optimizing the learning rate, batch size, number of filters in each convolutional layer, filter size, and dropout rate. Firstly, we delved into exploring a spectrum of learning rates to pinpoint the optimal value that ensures swift convergence towards the minimum of the loss function without overshooting. Next, we scrutinized various batch sizes to strike a delicate balance between training time and the stability of the gradient descent process. Moving forward, we embarked on an exploration of different combinations of the number of filters and filter sizes in the convolutional layers, aiming to unearth the configuration most adept at extracting salient features from the intricate CT scan images. Additionally, to combat overfitting and foster model robustness, we meticulously optimized the dropout rate, discerning the precise proportion of neurons to deactivate during training. Our methodology embraced a meticulous grid search strategy, systematically traversing through predefined sets of values for each hyperparameter while evaluating the model’s performance using cross-validation. This exhaustive search enabled us to pinpoint the hyperparameter combination that not only elevated the model’s classification accuracy but also bolstered its generalization capabilities. Subsequently, the efficacy of the selected hyperparameters was meticulously validated using a distinct validation set, underscoring the robustness and reliability of our chosen parameters. Through this systematic and rigorous approach to hyperparameter tuning, we achieved remarkable strides in fortifying the performance and stability of our lung cancer classification model, thereby augmenting its potential for real-world clinical applications.

The training and validation phases operate iteratively, with refinements made to the model’s architecture, hyperparameters, or training methodology based on the validation outcomes. This iterative refinement persists until the model achieves a satisfactory equilibrium of accuracy, generalizability, and robustness, thereby ensuring its efficacy and reliability in clinical settings for lung cancer stage classification.

Statistical methods

In the analysis of the IQ-OTH/NCCD lung cancer dataset, various statistical and machine learning techniques were employed to ensure a comprehensive evaluation of the data. The primary focus was on classification metrics to assess the performance of the predictive models.

Confusion matrix : The confusion matrix serves as a pivotal component in our analysis, furnishing a visual representation of the model’s performance. It succinctly presents the counts of true positives, true negatives, false positives, and false negatives, thereby offering a lucid comprehension of the model’s classification accuracy and any instances of misclassification.

Accuracy : The accuracy metric was calculated by dividing the number of correctly predicted observations by the total number of observations, providing a straightforward measure for assessing the model’s overall performance. However, relying solely on accuracy can be deceptive, particularly in datasets with imbalanced class distributions. Therefore, it is imperative to incorporate additional metrics for a more comprehensive evaluation. It is achieved by Eq.  16 .

Precision (positive predictive value) : Precision was utilized to assess the accuracy of positive predictions, quantified as the ratio of true positives to the sum of true positives and false positives. This metric bears significant relevance in scenarios where the repercussions of false positives are considerable. It is achieved by Eq.  17 .

Recall (sensitivity or true positive rate) : Recall assesses the model’s ability to detect positive instances, calculated as the ratio of true positives to the sum of true positives and false negatives. This metric holds particular importance in medical diagnostics, where failing to identify a positive case can lead to severe consequences. It is achieved by Eq.  18 .

F1-score : The F1-score, which is the harmonic mean of precision and recall, was used to provide a balance between the two metrics, particularly valuable in situations of class imbalance. It is a more robust measure than accuracy in scenarios where false negatives and false positives have different implications. It is achieved by Eq.  19 .

Cohen’s kappa : The Cohen’s Kappa statistic was applied to assess the agreement between observed and predicted classifications, accounting for chance agreement. This statistic offers a nuanced understanding of the model’s performance, which is particularly valuable in scenarios involving imbalanced datasets. It is achieved by Eq.  20 .

Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) : MSE (Mean Squared Error) and RMSE (Root Mean Squared Error) were calculated to evaluate the average squared difference and the square root of the average squared differences, respectively, between predicted and actual classification categories. These metrics are instrumental in understanding the variance of prediction errors. MSE and RMSE are achieved using Eqs.  21 and 22 , respectively.

Mean Absolute Error (MAE) : MAE (Mean Absolute Error) measures the average magnitude of errors in a set of predictions, regardless of their direction. It is a linear score, meaning that all individual differences are equally weighted in the average. It is achieved using Eq.  23 .

Receiver Operating Characteristic (ROC) Curve and Area Under the Curve (AUC) : The ROC curve graphically illustrates the diagnostic ability of the model by plotting the true positive rate against the false positive rate at various threshold settings. The AUC (Area Under the Curve) provides a single scalar value summarizing the overall performance of the model across all possible classification thresholds. It is achieved using Eq.  24 .

F2-score : The F2-score was calculated to weigh recall higher than precision, useful in scenarios where missing positive predictions is more detrimental than making false positives. It is achieved using Eq.  25 .

These statistical methods and metrics provided a multifaceted evaluation of the model’s performance, ensuring a robust analysis of the predictive capabilities and reliability in classifying the cases within the IQ-OTH/NCCD lung cancer dataset.

The evaluation of the IQ-OTH/NCCD lung cancer dataset through our predictive model yielded detailed insights across various statistical metrics, showcasing the model’s efficacy in classifying lung cancer stages. Here we delve into a comprehensive analysis of each metric:

Confusion matrix : The confusion matrix offered a detailed perspective on the model’s classification performance, unveiling a notable count of true positives and true negatives, reflecting precise predictions. Notably, there were minimal occurrences of false positives and false negatives, underscoring the model’s accuracy in discerning between benign, malignant, and normal cases. The same is visualized in Fig.  7 .

Confusion matrix

Accuracy : The overall model accuracy was noted at 99.64%, highlighting the model’s robust capacity to accurately identify and classify instances within the dataset. This exceptional accuracy rate underscores the model’s reliability in clinical diagnostic settings, establishing a solid basis for subsequent validation and potential clinical implementation. To provide visual insight of this Fig.  8 gives truly classified instances.

Correctly classified instances

Precision : The precision metric provided valuable insights into the model’s predictive reliability. It attained a precision of 96.77% for benign cases, signifying a high probability that a case predicted as benign is indeed benign. Moreover, for malignant and normal cases, the precision reached 100%, demonstrating the model’s outstanding ability to predict these categories accurately without any false positives.

Recall : The recall scores were equally remarkable, achieving 100% for both benign and malignant cases, and 99.04% for normal cases. These findings underscore the model’s sensitivity and its capability to accurately detect all true positive cases, thereby mitigating the risk of false negatives as a pivotal consideration in medical diagnostics.

F1-score : The F1-scores, which strike a balance between precision and recall, were 98.36% for benign, 100% for malignant, and 99.52% for normal cases. These scores signify the model’s balanced performance, guaranteeing both the accuracy of positive predictions and the reduction of false negatives. To enhance the visualization of the classification report, Table  3 provides a statistical representation.

Based on Table 3 a heatmap to visualize the same detail is provided in Fig.  9 for better insights.

Classification report

Cohen’s kappa : With a Cohen’s Kappa score of 0.9938, the model exhibited perfect agreement with the actual classifications, surpassing the performance expected by chance alone. This underscores an elevated level of consistency in the model’s predictions, thus reinforcing its reliability.

Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) : The model reported an MSE of 0.0145 and an RMSE of 0.1206, indicating minimal variance and bias in the prediction errors. These low values suggest that the model’s predictions are consistently close to the actual values, enhancing trust in its predictive power.

Mean Absolute Error (MAE) : With an MAE of 0.0073, the model exhibited minimal average error magnitude in its predictions, signifying high predictive accuracy. This metric further reinforces the model’s suitability for clinical settings where precision is crucial. To visualize the error metrics, a bar chart is given in Fig.  10 .

Error metrics barh chart

Receiver Operating Characteristic (ROC) Curve and Area Under the Curve (AUC) : The ROC curves and corresponding AUC values were exceptional, achieving AUCs of 1.00 for malignant, benign, and normal cases. These results indicate the model’s outstanding discrimination ability between different classes across various threshold settings. The roc-auc curve is provided in Fig.  11 .

F2-score : The F2-score of 0.9964, which places more emphasis on recall, indicates the model’s strong ability to identify positive cases. This is particularly important in the medical field, where failing to detect a condition could have profound consequences. The visual representation of performance score is given in Fig.  12 .

Performance scores

The detailed results across these metrics provide a comprehensive picture of the model’s performance, highlighting its precision, reliability, and robustness in classifying lung cancer stages from the IQ-OTH/NCCD dataset. The findings demonstrate the model’s potential as a diagnostic tool, supporting its further investigation and potential integration into clinical practice.

The analysis of the IQ-OTH/NCCD lung cancer dataset with our model reveals a groundbreaking level of performance in medical image classification. With an accuracy of 99.64% and exceptional precision and recall metrics across the three categories (benign, malignant, and normal), the model emerges as a highly reliable diagnostic aid. The significance of these results extends beyond the high metric scores; it lies in the model’s capability to accurately distinguish between benign and malignant cases, a critical aspect for patient management and treatment planning.

The high F1-score underscores the model’s balanced consideration of precision and recall, thereby minimizing the risk of misdiagnosis. Additionally, the emphasis on recall in the F2-score holds particular significance in the medical domain, where overlooking a positive case (false negative) can have more severe consequences than erroneously identifying a case as positive (false positive). The comparison between the baseline models and proposed model has been given in Table  4 .

In the realm of lung cancer detection, many existing models focus predominantly on binary classification, often neglecting the nuanced differentiation between benign and malignant cases [ 37 ]. Our model’s tri-classification capability sets a new benchmark, offering a more detailed diagnostic tool compared to the binary classifiers. When juxtaposed with existing methods, our model’s performance underscores its advanced detection capabilities, potentially offering a more nuanced and informative diagnostic perspective than currently available tools.

For clinical practice, the integration of such a high-performing model could revolutionize lung cancer diagnostics [ 22 , 38 ]. It can augment radiologists’ capabilities, reducing diagnostic time and increasing throughput. The ability to accurately classify lung nodules as benign, malignant, or normal could significantly reduce unnecessary interventions, minimizing patient exposure to invasive procedures and associated risks. Additionally, it can streamline the patient pathway, ensuring rapid treatment initiation for malignant cases and appropriate follow-up for benign conditions [ 39 , 40 ].

While the results are promising, the study’s limitations warrant consideration. The model’s training on a dataset from a specific demographic and geographic area raises questions about its applicability to broader populations. Additionally, the model’s performance in a controlled study environment might not fully translate to the diverse and unpredictable nature of clinical settings. The black-box nature of deep learning models also poses a challenge in clinical contexts, where understanding the rationale behind a diagnosis is as crucial as the diagnosis itself [ 41 ]. To make it more clear in Fig.  13 some misclassified instances has been shown.

Misclassified instances

When evaluating our CNN model’s performance on the lung cancer dataset, we noticed some errors in classification. These mistakes can happen for various reasons. Firstly, some features in the CT scans may look similar between benign and malignant nodules, making it hard for the model to tell them apart. Also, noise and artifacts in the scans can confuse the model by hiding important details. Even though we tried to balance the classes, rare cases could still be challenging for the model to recognize. Plus, early-stage cancer might look very similar to normal tissue, making it tricky for the model to spot. Differences in how scans are taken can also affect the model’s understanding, leading to errors. Lastly, if the model learns too much from the training data, it might not perform well on new, unseen images. To fix these issues, we’re planning to use better techniques for preparing the data, like removing noise more effectively and making the model more flexible to different imaging conditions. We also aim to combine multiple models and use more diverse data to improve accuracy. By addressing these challenges, we hope to make our model better at classifying lung cancer stages.

While the IQ-OTHNCCD lung cancer dataset has been instrumental in developing and validating our model, it is important to recognize its limitations, particularly concerning demographic and geographic diversity. The dataset predominantly represents a specific population, which may not capture the full spectrum of variations seen in global populations. This limitation poses challenges for the model’s generalizability, as differences in demographics, such as age, ethnicity, and underlying health conditions, can influence the presentation of lung cancer in CT scans.

To address these limitations, future research should focus on expanding the dataset to include a more diverse range of CT scan images from various demographic groups and geographic regions. This expansion can be facilitated through collaborations with international medical institutions and accessing publicly available medical imaging repositories. Additionally, incorporating advanced data augmentation techniques that simulate variations in demographic characteristics, such as age and gender, can further enhance the dataset’s diversity. By broadening the dataset, we aim to improve the model’s robustness and ensure its applicability across different populations, ultimately enhancing the utility and reliability of our diagnostic tool in diverse clinical settings. This approach will contribute to developing a more inclusive and universally applicable model for lung cancer diagnosis.

Sensitivity analysis of precision, recall, and F1-score

In our endeavor to comprehensively assess the performance of our Convolutional Neural Network (CNN) model for lung cancer diagnosis, we conducted a sensitivity analysis focusing on precision, recall, and the F1-score. Precision sensitivity involved systematically adjusting the threshold values used for classification to observe its impact on false positive rates and the model’s conservatism in identifying positive cases. As precision increased, indicating a more stringent classification approach, false positives decreased, but the risk of false negatives rose, necessitating a delicate balance in medical diagnostics. Conversely, recall sensitivity entailed modifying the model’s sensitivity to detect positive cases, thereby influencing its ability to minimize false negatives. Heightened recall improved the identification of true positives, crucial for early diagnosis and treatment, albeit with potential increases in false positives, mandating cautious management. Additionally, analyzing the F1-score, a harmonic mean of precision and recall, elucidated its role in balancing false positives and false negatives. Optimizing for a high F1-score underscored a balanced approach, ensuring robust performance across both precision and recall metrics. Overall, the sensitivity analysis underscored the significance of striking a delicate balance between precision, recall, and the F1-score to optimize the model’s performance in clinical settings. By navigating and managing these trade-offs effectively, we can bolster the reliability and efficacy of our model in diagnosing lung cancer, thereby contributing to improved patient outcomes.

Regulatory considerations for clinical application

Implementing machine learning models in clinical settings involves navigating a complex landscape of regulatory requirements to ensure patient safety, data security, and efficacy. One of the primary regulatory hurdles is obtaining approval from medical device regulatory bodies such as the U.S. Food and Drug Administration (FDA), the European Medicines Agency (EMA), or other relevant national authorities. These regulatory agencies require extensive validation studies to demonstrate the model’s accuracy, reliability, and safety in diagnosing lung cancer. This involves rigorous testing on diverse datasets to ensure the model’s generalizability and performance across different patient populations and clinical scenarios.

Additionally, regulatory guidelines mandate that machine learning models used in healthcare must provide a level of interpretability and transparency. Clinicians need to understand the decision-making process of the model to trust and effectively integrate it into clinical workflows. This requirement for explainability poses a challenge for deep learning models, which are often considered “black boxes.” Therefore, developing methods to elucidate the model’s reasoning, such as feature importance analysis or visual explanations, is crucial for meeting regulatory standards.

Data privacy and security are also significant regulatory concerns, particularly with the implementation of regulations like the General Data Protection Regulation (GDPR) in Europe and the Health Insurance Portability and Accountability Act (HIPAA) in the United States. Ensuring that patient data is anonymized, securely stored, and used ethically is essential for compliance. This includes implementing robust data encryption, access controls, and audit trails to protect sensitive health information from unauthorized access and breaches.

Moreover, post-market surveillance is a critical component of regulatory compliance, requiring continuous monitoring of the model’s performance in real-world clinical settings. This involves tracking the model’s diagnostic accuracy, identifying potential biases, and updating the model as needed to maintain its efficacy and safety over time. Establishing a framework for ongoing evaluation and improvement is essential to meet regulatory requirements and ensure the model’s long-term success in clinical applications.

Addressing these regulatory hurdles necessitates close collaboration between developers, healthcare providers, and regulatory bodies to ensure that machine learning models are safe, effective, and aligned with clinical needs. By adhering to these regulatory frameworks, we can facilitate the successful integration of advanced diagnostic tools into healthcare, ultimately enhancing patient outcomes and advancing the field of medical diagnostics.

Future research directions should focus on external validation of the model across various populations and healthcare settings to ascertain its universality and robustness. Integrating multimodal data, encompassing patient history, genetic information, and other diagnostic results, could enhance the model’s diagnostic precision. Addressing the interpretability of deep learning models could foster greater trust and integration into clinical decision-making processes. Additionally, prospective studies assessing the model’s impact on clinical outcomes, patient satisfaction, and healthcare efficiency would provide invaluable insights into its practical benefits and potential areas for improvement.

This study presented a comprehensive analysis of the IQ-OTH/NCCD lung cancer dataset using a sophisticated machine learning model, which demonstrated exceptional performance in classifying lung cancer stages. Key findings include a near-perfect accuracy rate of 99.64%, alongside impressive precision and recall metrics across benign, malignant, and normal case classifications. The model’s balanced F1-score and the emphasis on recall in the F2-score further highlight its diagnostic precision and sensitivity. These results signify a substantial advancement in the model’s ability to differentiate between nuanced lung cancer stages, providing a critical tool for early and accurate diagnosis.

The implications of these discoveries on the field of lung cancer diagnostics are profound. The model’s precision in classifying lung cancer stages holds the promise of substantially enhancing diagnostic protocols, thereby refining the accuracy and efficiency of lung cancer detection. This advancement has the potential to facilitate earlier treatment interventions, potentially enhancing patient outcomes and survival rates. Moreover, the model’s capability to differentiate between benign and malignant nodules could mitigate the need for unnecessary invasive procedures, consequently reducing patient risk and healthcare expenditures.

Future research should focus on external validation of the model to ensure its effectiveness across diverse populations and clinical settings. The exploration of model interpretability is crucial for clinical adoption, where understanding the basis for diagnostic decisions is essential. Additionally, integrating the model with other diagnostic data and clinical workflows could enhance its utility and impact.

Prospective studies are needed to evaluate the model’s real-world clinical impact, particularly its ability to improve patient outcomes, streamline diagnostic pathways, and reduce healthcare costs. The potential for the model to be adapted or extended to other types of cancers or medical imaging modalities also represents an exciting avenue for future research.

This study highlights the potential of advanced machine learning models to transform lung cancer diagnostics, providing a more precise, effective, and nuanced approach to detecting and classifying lung cancer. The ongoing advancement and incorporation of such models into clinical settings hold the promise of catalyzing substantial progress in patient care and outcomes within the field of oncology.

Availability of data and materials

Data used for the findings are publicly available at https://www.kaggle.com/datasets/hamdallak/the-iqothnccd-lung-cancer-dataset .

Nooreldeen R. Current and future development in lung cancer diagnosis. Int J Mol Sci. 2021;22:8661.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Rea G, et al. Beyond visual interpretation: quantitative analysis and artificial intelligence in interstitial lung disease diagnosis expanding horizons in radiology. Diagnostics. 2023;13:2333.

Article   PubMed   PubMed Central   Google Scholar  

Rajasekar V, et al. Lung cancer disease prediction with CT scan and histopathological images feature analysis using deep learning techniques. Results Eng. 2023;18:101111.

Article   CAS   Google Scholar  

Lanjewar MG, Kamini G, Panchbhai, Panem Charanarur. Lung cancer detection from CT scans using modified DenseNet with feature selection methods and ML classifiers. Expert Syst Appl. 2023;224:119961.

Article   Google Scholar  

Raza R, et al. Lung-EffNet: lung cancer classification using EfficientNet from CT-scan images. Eng Appl Artif Intell. 2023;126:106902.

Chaunzwa TL, et al. Deep learning classification of lung cancer histology using CT images. Sci Rep. 2021;11(1):1–12.

Chaturvedi P, Jhamb A, Vanani M, Nemade V. Prediction and Classification of Lung Cancer Using Machine Learning Techniques. IOP Conference Series: Materials Science and Engineering. 2021;1099:012059. https://doi.org/10.1088/1757-899X/1099/1/012059 .

Hong M, et al. Multi-class classification of lung diseases using CNN models. Appl Sci. 2021;11:9289.

Phankokkruad M. Ensemble transfer learning for lung cancer detection. 2021 4th international conference on data science and information technology. 2021.

Google Scholar  

Ren Z, Zhang Y, Wang S. LCDAE: data augmented ensemble framework for lung cancer classification. Technology Cancer Research Treatment. 2022;21:15330338221124372.

Protonotarios NE, et al. A few-shot U-Net deep learning model for lung cancer lesion segmentation via PET/CT imaging. Biomedical Physics Engineering Express. 2022;8(2):025019.

Heuvelmans MA, van Ooijen PM, Ather S, Silva CF, Han D, Heussel CP, Oudkerk M. Lung cancer prediction by Deep Learning to identify benign lung nodules. Lung Cancer. 2021;154:1–4.

Article   PubMed   Google Scholar  

Le NQK, Kha QH, Nguyen VH, Chen YC, Cheng SJ, Chen CY. Machine learning-based radiomics signatures for EGFR and KRAS mutations prediction in non-small-cell lung cancer. Int J Mol Sci. 2021;22(17):9254.

Xie Y, Meng WY, Li RZ, Wang YW, Qian X, Chan C, Leung ELH. Early lung cancer diagnostic biomarker discovery by machine learning methods. Transl Oncol. 2021;14(1):907.

Li Z, et al. Deep Learning Methods for Lung Cancer Segmentation in Whole-Slide Histopathology Images—The ACDC@LungHP Challenge 2019. IEEE J Biomed Health Inform. 2021;25(2):429–40.

Narvekar S, Shirodkar M, Raut T, Vainganka P, Chaman Kumar KM, Aswale S. A Survey on Detection of Lung Cancer Using Different Image Processing Techniques. London, United Kingdom: 2022 3rd International Conference on Intelligent Engineering and Management (ICIEM); 2022. p. 13–8. https://doi.org/10.1109/ICIEM54221.2022.9853190 .

Book   Google Scholar  

Aharonu M, Kumar RL. Convolutional Neural Network based Framework for Automatic Lung Cancer Detection from Lung CT Images. Bangalore, India: 2022 International Conference on Smart Generation Computing, Communication and Networking (SMART GENCON); 2022. p. 1–7. https://doi.org/10.1109/SMARTGENCON56628.2022.10084235 .

Kavitha BC, Naveen KB. Image Acquisition and Pre-processing for Detection of Lung Cancer using Neural Network. Mandya, India: 2022 Fourth International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT); 2022. p. 1–4.

Causey JL, et al. Spatial pyramid pooling with 3D convolution improves Lung Cancer Detection, in IEEE/ACM transactions on Computational Biology and Bioinformatics . 1 March-April. 2022;19(2):1165–72. https://doi.org/10.1109/TCBB.2020.3027744 .

Ahmed I, Chehri A, Jeon G, Piccialli F. Automated Pulmonary Nodule Classification and Detection Using Deep Learning Architecture. IEEE/ACM Trans Comput Biol Bioinform. 2023;20(4):2445–56. https://doi.org/10.1109/TCBB.2022.3192139 .

Thakur A, Gupta M, Sinha DK, Mishra KK, Venkatesan VK, Guluwadi S. Transformative breast Cancer diagnosis using CNNs with optimized ReduceLROnPlateau and Early stopping Enhancements. Int J Comput Intell Syst. 2024;17(1):14.

Albalawi E, Thakur A, Ramakrishna MT, Khan B, Sankaranarayanan S, Almarri SB, Aldhyani T. Oral squamous cell carcinoma detection using EfficientNet on histopathological images. Front Med. 2024;10:1349336.

Shah AA, Malik HAM, Muhammad A, Alourani A, Butt ZA. Deep learning ensemble 2D CNN approach towards the detection of lung cancer. Sci Rep. 2023;13(1):2987.

Alzubaidi MA, Otoom M, Jaradat H. Comprehensive and Comparative Global and Local Feature Extraction Framework for Lung Cancer Detection Using CT Scan Images, in IEEE Access . 2021;9:158140–54. https://doi.org/10.1109/ACCESS.2021.3129597 .

Mathio D, Johansen JS, Cristiano S, Medina JE, Phallen J, Larsen KR, Velculescu E. Detection and characterization of lung cancer using cell-free DNA fragmentomes. Nat Commu. 2021;12(1):5060.

Mehmood S et al. Malignancy Detection in Lung and Colon Histopathology Images Using Transfer Learning With Class Selective Image Processing, in IEEE Access , vol. 10, pp. 25657–25668, 2022, https://doi.org/10.1109/ACCESS.2022.3150924 .

Dritsas E, Trigka M. Lung cancer risk prediction with machine learning models. Big Data Cogn Comput. 2022;6(4):139.

Masud M, Sikder N, Nahid AA, Bairagi AK, AlZain MA. A machine learning approach to diagnosing lung and colon cancer using a deep learning-based classification framework. Sensors. 2021;21(3):748.

Naseer S, Akram T, Masood M, Rashid, Jaffar A. Lung Cancer Classification Using Modified U-Net Based Lobe Segmentation and Nodule Detection, in IEEE Access , vol. 11, pp. 60279–60291, 2023, https://doi.org/10.1109/ACCESS.2023.3285821 .

Bharathy S, Pavithra R. Lung Cancer Detection using Machine Learning. In 2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC). 2022. p. 539–43 IEEE.

Kasinathan G, Jayakumar S. Cloud based lung tumor detection and stage classification using deep learning techniques. BioMed Res Int. 2022;2022:4185835.

Das S, et al. Automated prediction of Lung Cancer using Deep Learning algorithms. Applied Artificial Intelligence. CRC; 2023. pp. 93–120.

Chapter   Google Scholar  

Tasnim N, et al. A Deep Learning Based Image Processing Technique for Early Lung Cancer Prediction. 2024 ASU International Conference in Emerging Technologies for Sustainability and Intelligent Systems (ICETSIS). 2024. IEEE.

Safta W. Advancing pulmonary nodule diagnosis by integrating Engineered and Deep features extracted from CT scans. Algorithms. 2024;17(4):161.

Khaliq K, et al. LCCNet: a deep learning based Method for the identification of lungs Cancer using CT scans. VFAST Trans Softw Eng. 2023;11(2):80–93.

Nigudgi S. Lung cancer CT image classification using hybrid-SVM transfer learning approach. Soft Comput. 2023;27(14):9845–59.

Diwakar M, Singh P, Shankar A. Multi-modal medical image fusion framework using co-occurrence filter and local extrema in NSST domain. Biomed Signal Process Control. 2021;68:102788. https://doi.org/10.1016/j.bspc.2021.102788 .

Das M, Gupta D, Bakde A. An end-to-end content-aware generative adversarial network-based method for multimodal medical image fusion. Data Analytics Intell Sys. 2024;7(1):7–10. https://doi.org/10.1088/978-0-7503-5417-2ch7 .

Jie Y, Xu Y, Li X, Tan H. (2024). TSJNet: A Multi-modality Target and Semantic Awareness Joint-driven Image Fusion Network. arXiv preprint arXiv:2402.01212 .

Dhaundiyal R, Tripathi A, Joshi K, Diwakar M, Singh P. Clustering based multi-modality medical image fusion. In: Journal of Physics: Conference Series. 2020 (Vol. 1478, No. 1, p. 012024). IOP Publishing.

Diwakar M, Singh P, Shankar A, Nayak RS, Nayak J, Vimal S, Sisodia D. Directive clustering contrast-based multi-modality medical image fusion for smart healthcare system. Netw Model Anal Health Inf Bioinf. 2022;11(1):15.

Download references

Acknowledgements

Not applicable.

This research received no external funding.

Author information

Authors and affiliations.

Al-Ameen Engineering College (Autonomous), Erode, Tamil Nadu, India

M. Mohamed Musthafa

Department of Computer science and Engineering, East Point College of Engineering & Technology, Bangalore, India

I. Manimozhi

Department of Computer Science and Engineering, JAIN (Deemed-to-be University), Bengaluru, 562112, India

T. R. Mahesh

Adama Science and Technology University, Adama, 302120, Ethiopia

Suresh Guluwadi

You can also search for this author in PubMed   Google Scholar

Contributions

M.M.M took care of the review of literature and methodology. M.T.R has done the formal analysis, data collection and investigation. I.M has done the initial drafting and statistical analysis. S.G has supervised the overall project. All the authors of the article have read and approved the final article.

Corresponding author

Correspondence to Suresh Guluwadi .

Ethics declarations

Ethics approval and consent to participate.

Not applicable. 

Consent for publication

Not applicable as the work is carried out on publicly available dataset.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Musthafa, M.M., Manimozhi, I., Mahesh, T.R. et al. Optimizing double-layered convolutional neural networks for efficient lung cancer classification through hyperparameter optimization and advanced image pre-processing techniques. BMC Med Inform Decis Mak 24 , 142 (2024). https://doi.org/10.1186/s12911-024-02553-9

Download citation

Received : 16 April 2024

Accepted : 22 May 2024

Published : 27 May 2024

DOI : https://doi.org/10.1186/s12911-024-02553-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Lung cancer
• Machine learning
• Classification
• Diagnostic accuracy

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

• General enquiries: [email protected]

• Open access
• Published: 30 May 2024

CRISPR-Cas and CRISPR-based screening system for precise gene editing and targeted cancer therapy

• Mingming Qin 1 , 2   na1 ,
• Chunhao Deng 3   na1 ,
• Liewei Wen 4 ,
• Guoqun Luo 1 &
• Ya Meng 4  

Journal of Translational Medicine volume  22 , Article number:  516 ( 2024 ) Cite this article

Metrics details

Target cancer therapy has been developed for clinical cancer treatment based on the discovery of CRISPR (clustered regularly interspaced short palindromic repeat) -Cas system. This forefront and cutting-edge scientific technique improves the cancer research into molecular level and is currently widely utilized in genetic investigation and clinical precision cancer therapy. In this review, we summarized the genetic modification by CRISPR/Cas and CRISPR screening system, discussed key components for successful CRISPR screening, including Cas enzymes, guide RNA (gRNA) libraries, target cells or organs. Furthermore, we focused on the application for CAR-T cell therapy, drug target, drug screening, or drug selection in both ex vivo and in vivo with CRISPR screening system. In addition, we elucidated the advantages and potential obstacles of CRISPR system in precision clinical medicine and described the prospects for future genetic therapy.

In summary, we provide a comprehensive and practical perspective on the development of CRISPR/Cas and CRISPR screening system for the treatment of cancer defects, aiming to further improve the precision and accuracy for clinical treatment and individualized gene therapy.

Introduction

Cancer therapy has been developed from the very initial surgical removal in the ancient to currently precision minimally invasive surgery; from the chemotherapy, radiotherapy to the targeted therapy and precision individualized immunotherapy, under the progress of precise and granular molecular characterization at present [ 1 ]. The newly discovered genome editing tool CRISPR (clustered regularly interspaced short palindromic repeat) /Cas system provides a powerful method for the investigation of cancer therapy [ 2 , 3 , 4 ]. It was described initially in bacteria as a primitive immune system to fight against viral infections and was universally recognized as a genomic modification system in the past decade [ 5 , 6 ]. In Prokaryotes, the short DNA repeats CRISPR exist between regular spacing units, and are recognized as intervening sequences derived from preexisting fragment of bacteriophages and conjugative plasmids, contributing to bacteria immune system [ 7 ]. The genetic sequences of the viral invaders or plasmid challengers are captured and aligned as spacer segments in the CRISPR region in bacteria or archaea [ 8 , 9 ], comprising the CRISPR-mediated adaptive immunity system [ 10 ]. Two classes of CRISPR-Cas systems have been described in prokaryotes based on their effector modules [ 11 , 12 , 13 , 14 ], characterized into 6 types, and 33 subtypes described in 2020 [ 15 ]. The Class 2 CRISPR-Cas system composed only 10% percentage but has expanded biotechnology toolbox for genome editing with 190,000 shares worldwide from 640 labs [ 16 , 17 ]. It consists of three types of effectors: type II, type V and type VI, with several widely recognized genetic editing enzymes, being Cas9 in type II, Cas12a (Cpf1), Cas12b (C2c1), Cas12c (C2c3), Cas14 subgroup in type V [ 18 ], Cas13a (C2c2), Cas13b (C2c6) and Cas13c (C2c7) in type VI [ 14 , 19 ]. Schematic representation of two classes of CRISPR/Cas systems were depicted in Fig.  1 .

Schematic representative of CRISPR/Cas loci in Class 1 and Class 2 system. Class 1 system show multi-component effectors, while the Class 2 system have one effector. Three subgroups of Class 2 CRISPR systems are presented. Representative Type II-A CRISPR protein contains: Streptococcus pyogenes Cas9 (SpCas9), Staphylococcus aureus Cas9 (SaCas9) and Streptococcus thermophilus Cas9 (StrCas9), all of which have the tracrRNA sequences. Type V CRISPR, which comprises Cas12a, Cas12b and Cas12c, exhibits distinct genome structures. Cas12b has the tracrRNA structure, while Cas12c only has one assistant protein cas1 for genome editing. Cas14 subgroup is not depicted in this figure. Type VI CRISPR systems show few assistant proteins to identify RNA virus, however,  type VI-B has csx27 and csx28 proteins to regulate nuclease activity. Illustrated according to Ref [ 14 , 16 , 20 , 21 ].

CRISPR/Cas system has been utilized for cellular genetic modification [ 22 , 23 ] and the generation of animal models for cancer research [ 24 , 25 ]. Furthermore, the CRISPR/Cas-based genetic screening system was developed for cellular investigation [ 26 , 27 , 28 ], as well as in tumor studies [ 25 , 29 ]. In addition, high throughput gRNA libraries have been established to enable efficient genetic screening, specially facilitating personalized treatment strategies for cancer patients individually [ 30 ]. In this review, we provide a comprehensive overview of the CRISPR/Cas system and essential elements for successful CRISPR screening system, including gRNA libraries, gRNA validation, and clinical application for cancer research. Furthermore, we explored the application of the CRISPR screening system in cancer therapy from both ex vivo and in vivo investigation, aiming to elucidate the inherent advantages and potential obstacles for clinical precision medicine.

The application of Class 2 CRISPR-Cas effectors and genome modification in cancer therapy

Type ii effector cas9 in cancer research.

Both Streptococcus pyogenes Cas9 (SpCas9) and Staphylococcus aureus Cas9 (SaCas9), classified as the type II-A effectors, showed comparable genome editing efficiency for in vitro and in vivo study [ 21 , 31 , 32 , 33 ]. These effectors enable rapid modification of cellular or animal models for transcriptional modulation via CRISPR knockout/knockin or high throughput genomic screening [ 23 , 34 ]. The compact size of SaCas9 renders it an optimal enzyme for in vivo AAV application. However, SpCas9, one of the pioneering Cas9 proteins, has been extensively investigated and utilized in CRISPR gene editing. Three variants of SpCas9 have been developed, the wild-type Cas9, nickase Cas9 (nCas9), and dead Cas9 (dCas9).

Cas9 mediated DNA cleavage with the two distinct active sites RuvC and HNH, under the assistance of CRISPR RNA (crRNA) and trans-activating crRNA (tracrRNA) ribonucleoprotein complex [ 8 ]. The dual-tracrRNA: crRNA chimera single guide RNA (sgRNA) was created and directed Cas9 nuclease to the potential target loci for site-specific DNA cleavage, initiating the genome editing system in vitro [ 35 ]. The binding of Cas9 to the adjacent sequence of three nucleotides, known as protospacer adjacent motif (PAM), triggers DNA cleavage by inducing double-strand breaks with its scissor-like activity [ 36 ]. The recently used Cas9-gRNA ribonucleoprotein (RNP) complexes remarkably increase fidelity and efficacy for double-strand DNA breaks with minimized cell mortality [ 37 ]. It also combined with repair donor to achieve site-specific correction of cystic fibrosis transmembrane conductance regulator (CFTR) gene mutations in epithelial organoids [ 38 ]. Cre-dependent Cas9 knockin mouse was generated, and KRAS, p53 , and LKB1 depletion resulted in carcinoma formation in these transgenic mice, providing a robust cancer model for research [ 24 ].

One mutation in D10A of Cas9 protein makes a nCas9, which improves genome editing specificity [ 39 ]. The combination of sgRNA pairs with nCas9 significantly enhances cutting specificity by 50-1000 folds in cell lines and mouse zygotes [ 40 ]. CRISPR-Cas base editing using nCas9 enables precise incorporation of point mutations in genomic DNA without inducing double-strand breaks, demonstrating its potential in treating genetic diseases caused by base-pair alterations through adenine base editors (ABEs) or cytosine base editors (CBEs) [ 41 ]. In addition, DNA base editors combining with the leading platform adeno-associated virus (AAV) vector for viral delivery expanded the CRISPR-base-edit toolkit for Prime-editing (PE) [ 42 ]. Meanwhile, the recently developed genome editing technique known as NICER utilizes Cas9 D10A nickase to correct heterozygous mutations. It generates multiple DNA nicks and triggers gene correction via interhomolog homologous recombination (IH-HR) which rarely induces genomic alterations, making it a precise strategy to restore genetic diseases or single nucleotide mutations [ 43 ]. Except the precise single nucleotide restoration, cancer translocations were generated by double strand breaks and paired nicks with either Cas9 or nCas9, creating endogenous chromosomal translocations cell model for investigating tumor driving genes [ 44 ].

Catalytically inactive Cas9, a ‘dead’ protein (dCas9) with both mutations in D10A and H840A of RuvC and HNH domains, showed its popularity in gene regulation with inhibition, activation, and cell imaging and labeling [ 45 ]. Genome-scale screenings utilizing CRISPR inhibition (CRISPRi) and CRISPR activation (CRISPRa) have been employed to identify both known and novel genes involved in controlling cell growth and sensitivity to toxins [ 46 ]. Precise inducible gene knockdown or overexpression can be supported using dCas9-KRAB (Krüppel-associated box) or Cas9 combined with Tetracycline Inducible Expression promoter (TetO) [ 47 ]. Firstly, the fusion of dCas9 with transcriptional repressor produces the CRISPRi genetic tool [ 48 ]. The dCas9-BFP-KRAB repressor domain enables the suppression of gene expression [ 49 ]. Second, fusing dCas9 with RNA polymerase (RNAP) omega subunit upregulates gene expression [ 50 ], and dCas9-VP64 was used for transcriptional activation [ 51 ]. In addition, dCas9 protein serves as a valuable tool for labeling of endogenous genomic loci in living cells. By employing an optimized sgRNA fused with EGFP-tagged dCas9, repetitive elements in telomeres and various other regions can be robustly labeled [ 52 ]. A double-color CRISPR labeling method was established by incorporating MS2 or PP7 RNA aptamers into the sgRNA, fused with the catalytically inactive Cas9 (dCas9) for direct visualization [ 53 ]. Finally, dCas9 can be employed for in vivo imaging of chromosomal dynamics and genome organization dimensions [ 47 ], allowing systematic fluorescent labeling of up to 10 proteins [ 48 ]. Summary of the type II Cas9 enzymes was depicted in Fig.  2 .

Summary of Cas9 proteins and modified nCas9 and dCas9 genome editing tools. (A) PAM for SpCas9 is NGG, while PAM for SaCas9 is NNGRRT with the ability to cut DNA double helix. (B) Mutation of D10A leads to the formation of nCas9 while both mutations generate dCas9 protein. (C) nCas9 can be applied for base editing such as CBE and ABE, also for Base editor and developed as NICER to repair heterogenous mutation. (D) dCas9 was modified to generate CRISPRi, CRISPRa and CRISPR labeling tools. dCas9: dead Cas9. nCas9: nickase Cas9. CBE: Cytosine Base Editor, ABE: Adenine Base Editor. RT: reverse transcriptase. pegRNA: prime editing guide RNA.

Type V and type VI effectors in cancer research

Mainly three subtypes of type V effectors were investigated for gene editing, named as type V-A, V-B and V-C. The type V-A effector Cpf1 (CRISPR from Prevotella and Francisella 1 ), exhibits enhanced genome editing specificity attributed to a T-rich PAM (-5′TTTV) [ 54 ], resulting in a staggered DNA double stranded break [ 55 ]. Two candidate Cpf1 (Cas12a) enzymes, AsCpf1 from Acidominococcus sp. BV3L6 and LbCpf1 from Lachnospiraceae bacterium ND2006 , show a robust genome editing ability in human cells compared to that of Cas9 [ 56 ]. Furthermore, successful generation of gene knockout transgenic mice was achieved using both AsCpf1 (40.7%) and LbCpf1 (28.6%), providing a wonderful animal model for research [ 57 , 58 ]. Multiplex genome editing was conducted using Cpf1 from Aspergillus aculeatus strain TBRC277 [ 59 ] and AsCpf1 was engineered with adeno-associated viral vectors (AAVs) for multiplex genome editing of mouse brain in vivo [ 60 ]. One-step generation of homology-directed repair (HDR) and checkpoint knockout CAR-T (KIKO CAR-T) was achieved with the adeno-associated virus and CRISPR/Cpf1 system, establishing an efficient AAV-Cpf1 double knockin system and opening new possibilities for cancer research [ 61 ]. The type V-B CRISPR effector Cas12b (C2c1) discovered in Bacillus hisashii (BhCas12b) showed a nickase effect at 37 °C for human gene editing, while BhCas12b v4, containing K846R/S893R/E837G mutants, demonstrated strong genome editing ability in human cells comparable to SpCas9 [ 62 ]. While the type V-C CRISPR effector Cas12c (C2c3) is a site-specific ribonuclease generating mature crRNAs for DNA targeting, crRNAs direct DNA binding by Cas12c without DNA cutting, providing a DNase-free pathway for transient antiviral immunity [ 63 ].

While both type II and type V are effective for DNA targeting in the genome level, the type VI effector Cas13 exhibits efficacy in treating genetic diseases and rescuing diseased sequences at the RNA level. They provide valuable genetic tools for diagnosis and degradation of viruses such as HIV and HPV [ 64 , 65 ]. Several Cas13 proteins were characterized, such as Cas13a, Cas13b, Cas13bt and Cas13d, showed the efficiency to cleave single stranded RNAs [ 66 , 67 , 68 ]. Of which Cas13a based SHERLOCK (Specific High-Sensitivity Enzymatic Reporter UnLOCKing) system can detect Zika or Dengue Virus as well as somatic mutations in cell free DNA (cfDNA) samples such as serine/threonine kinase (BRAF) V600E cancer mutation [ 69 ]. Shortened detection time and high sensitivity were applied for virus detection via SHERLPCKv2 system [ 70 , 71 ].

SHERLOCK enables to identify EGFR-T790M mutation in patient DNA with high efficiency by detecting 0.6% mutant ratio samples [ 72 ], this system was also used for DNA and RNA detection with single-base specificity and attomolar sensitivity in cancer patients samples [ 73 ]. Cas13b was used to fight RNA viruses such as porcine reproductive and respiratory syndrome virus (PRRSV) [ 74 ], chikungunya (CHIKV) and dengue in mosquito cells [ 75 ] as well as SARS-CoV-2 resistance [ 76 , 77 ]. Since Cas13b targets RNA without interfering genome sequence of the targeted gene, it provides a potential safer alternative to Cas9 enzymes. Catalytically inactive Cas13b (dCas13b) was engineered to direct adenosine-to-inosine deaminase for precise base editing, enabling the Programmable A to I Replacement (REPAIR) RNA editing platform. This platform can be utilized in transcriptome engineering of advanced leukemias, as well as head, liver, and breast cancers, thereby demonstrating a feasible strategy for investigating gene function in cancer at the RNA level [ 78 , 79 ]. The RNA-targeting CRISPR-Cas13 system showed promising roles in cancer diagnosis, therapy, and research; with the ability for early detection of cancer markers in liquid biopsy samples, degradation and manipulation of cancer-related mutant transcripts, as well as identification of novel therapeutic drug targets described in the recent review [ 80 ].

Altogether, the class 2 effectors expanded the current CRISPR/Cas toolkit. Cas9 possesses recognition ability of specific target sequences, and has the genomic editing ability for precision cancer treatment and mutation detection [ 2 ]. Meanwhile, the recently discovered Cas12 and Cas13 expand RNA editing tool, providing novel genetic methods for cancer diagnosis and molecular examination of cancer research [ 3 ].

The application of CRISPR screening system in cancer

The development of CRISPR/Cas system and high-throughput sequencing makes genetic screening easily accessible in basic biology, drug discovery, and personalized medicine for cancer therapy [ 3 ]. Cas9 nuclease is a preferred choice for genetic screening, and has been used for genomic modification in multiple researches [ 26 , 28 , 81 , 82 ]. One-step generation of multiplex genome mutations via CRISPR/Cas9 system was successfully achieved in mice, facilitating in vivo functional analysis of redundant genes [ 83 ]. CRISPR screening system was developed based on CRISPR/Cas combined with thousands of gRNAs integrated into viral vectors [ 81 , 84 ]. These libraries harbor gRNAs targeting various genes, and have received up to 1000 annual requests globally, enabling unbiased, phenotypic forward genetic screening [ 17 ]. The first whole genomic gRNA libraries for both mouse and human were generated with mouse lentiviral gRNA library containing 87,897 gRNAs for 19,150 coding genes, naming as (GeCKOv1), and was established to screen out unknown genes for Clostridium septicum alpha-toxin or 6-thioguanine (6TG) drug resistance [ 81 ]. However, low viral titer of the lentiviral delivery systems in GeCKOv1 limited the usage for biological screening, and genome-scale CRISPR knockout v2 (GeCKOv2), contained 123,411 unique sgRNAs targeting 19,050 annotated protein-coding genes and 1000 control sgRNAs (sg-NTCs), resulting in a 10-fold increase for viral generation [ 84 ]. Optimized mouse gRNA libraries targeting 20,611 genes with 130,209 gRNAs were also established with 100-fold increase of functional viral titer [ 84 ]. Innovative strategies of CRISPR-Cas9 system have been developed for large-scale genome knockout and transcriptional activation [ 85 ], as well as combinatorial genetic screening [ 27 ]. Processes for gRNA library generation and amplification were illustrated as depicted in the following Fig.  3 .

Schematic representation of gRNA library construction and virus production. (A) Oligoes synthesis and vector construction for gRNA library. (B) Amplification of gRNA library by bacterial culture, collection, and plasmid extraction. (C) PCR examination and sequence confirmation for library coverage. (D) Plasmids transfection and virus production with a certain gRNA library.

gRNA libraries for cancer research

Various of genome-scale gRNA libraries were established for CRISPR screening, and some gRNA libraries for specific selected genes were also established with small capacity. Established gRNA libraries of genome wide and specific selected targets for cancer research were summarized in the Table  1 .

Human lentiviral GeCKOv1 library (lentiCRISPRv1) was established for high throughput gene targeting of 18,080 genes, with 64,751 unique gRNAs total, and was used for cell viability-related gene screening in cancer. It was also examined for resistance to a therapeutic RAF inhibitor, vemurafenib, in a A375 melanoma model, leading to the discovery of novel genes sensitive to drug treatment [ 28 ]. GeCKOv2 library was also used to identify responsible genes related to EGF-induced apoptosis [ 86 ]. Genome-wide sgRNA library (mGeCKOa) transfection in non-metastatic mouse non-small cell lung cancer with 67,405 sgRNAs targeting 20,611 protein-coding genes. Cells were treated and transplanted into immunocompromised Nu/Nu mice, and tumor growth and migration were evaluated in vivo [ 25 ]. The pooled lentiviral sgRNA library with 73,151 gRNAs targeting 7114 gene and 100 non-targeting controls were used to screen the resistant genes for nucleotide analog 6TG treatment in human leukemic cell lines, screening resistance genes toward chemotherapeutic etoposide [ 26 ]. Patient-derived glioblastoma cell line (GBM), retinal epithelial cells (RPE1), colorectal carcinoma (HCT116 and DLD1), cervical carcinoma (Hela) and melanoma (A375) cells were subjected into genetic screening with the “90k library” containing 17,232 targeting genes and 91,320 gRNA sequences. Subsequentially, the supplemental library naming 176,500 TKO (Toronto KnockOut) library targeting 17,661 protein-coding genes were used to identify fitness genes in cancer cell lines [ 87 ]. Lentiviral vectors with genome-scale sgRNA library consisting of 70,290 guides (3 sgRNAs for each transcription start site (TSS)) were used for synthetic activation mediator (SAM)-based screening to target 200 bp upstream of the TSS and confer resistance to a BRAF inhibitor in melanoma cell line A375 and patient derived samples [ 51 ].

Although genome scale gRNA libraries are widely used in cancer research, its complexity and transcript isoform variance as well as difficulty in viral vectors cloning limited its usage. Other specific gRNA libraries for certain signal pathways or gene functions were established according to screening purpose for modulating endogenous genes. Total 5920 candidate enhancers were perturbed by the dCas9-KRAB enzyme, establishing the multiplex, expression quantitative trait locus (eQTL) framework, and total 664 cis enhancer-gene pairs were identified and enriched based on 254,974 single-cell transcriptomes in K562 derived from a chronic myologenous leukemia patient [ 49 ]. Undescribed immunotherapy targets for transplantable melanoma tumors in mice were explored with the 9992 sgRNAs targeting 2368 genes selected from transduced cells, establishing the in vivo genetic screen tumor models [ 88 ]. Recurrently mutated genes derived from pan-cancer The Cancer Genome Atlas datasets were recognized as well-known tumor suppressors genes (TSGs) or oncogenes. Total 49 orthologs of human TSGs were found in mouse genome, and the mouse TSG library containing 280 sgRNAs targeting 56 different genes (7 housekeeping genes) were used for tumor metastasis analysis [ 89 ].

The improvements of specificity and validation methods for gRNA Library

The procedure to perform pooled genome-editing experiments was clearly described, and successful CRISPR/Cas9 screening needs the specific and efficient gRNA sequence with proper quality and low off-target effect [ 91 ]. Off-target predictions calculated by algorithms indicating false positives and quantified error rates were developed by Bowtie and BWA sequencing methods, or considered by MIT-Broad score and the CFD score as summarized in previous reviews [ 92 ]. Computational tools for sgRNA designing with low off-target and high on-target efficacy and specificity have been developed and summarized in 2018 [ 93 ]. Several methods have been built for eliminating off-target results such as the utilization of high-efficiency delivery RNP tool, modification of the gRNA sequence, and improvement the specificity of Cas9 Enzymes [ 94 ]. The computational tool CRISPOR established high-quality gRNA libraries by selection according to off-target and on-target predictions, it also helps with vector cloning, gRNA validation and expression with primer designing and restriction enzymes depiction [ 95 ]. Optimized on-target efficiency prediction model was generated to illustrate the cleavage ability of gRNA sequence ( http://crispor.org ) [ 96 ]. Meanwhile, CRISPResso provides a robust and user-friendly computational pipeline to evaluate effects of coding and noncoding sequences and select off-target sites [ 97 ]. For precise gene selection analysis, the Model-based Analysis of Genome-wide CRISPR/Cas9 Knockout (MAGeCK) is the optimized method for both positive and negative selection, which offers high sensitivity and low FDR regardless of sequencing depth or sgRNA numbers for a single gene [ 98 ]. Besides that, intergration deficient lentiviral (IDLV) capture [ 99 ], and high-throughput genome-wide translocation sequencing (HTGTS) [ 100 ] are other methods for off-target detection.

Analysis of gRNAs abundance in pooled libraries plays an important role in targeting efficiency and screening accuracy and specificity. PCR products of gRNA library vectors can be sequenced on Hiseq 2500 and aligned to sgRNAs by Bowtie, an ultrafast, efficient program for aligning short DNA sequence to large genomes [ 101 ]. Rigorous analytical methods mitigate the false discovery rates generated by CRISPR screens via a Bayesian classifier of gene essentiality [ 102 ]. Sequence quality control can also be carried out under the guide of GPP Pooled Screen Analysis ( https://portals.broadinstitute.org/gpp/broad/ ), and statistical enrichment and gene depletion were calculated by hit calling algorithm STARS ( http://www.broadinstitute.org/rnai/public/software/index ) based on normalized fold changes [ 103 ]. High-content downstream gRNA library sequence validation in tumor immunology were summarized in the recent review [ 29 ]. Generally speaking, breaks labeling, enrichment on streptavidin and next-generation sequencing (BLESS) [ 104 ], genome-wide unbiased identification of DSBs enabled by sequencing (GUIDE-seq) [ 105 , 106 ] and discovery of in situ Cas off-targets and verification by sequencing (DISCOVER-seq) [ 107 , 108 , 109 ] were used as cell based methods with direct sequencing. More sensitive biochemical methods such as digested genome sequencing (Digenome-seq) [ 110 , 111 , 112 ], selective enrichment and identification of adapter-tagged DNA ends by sequencing (SITE-Seq) [ 113 ], circularization for in vitro reporting of cleavage effects by sequencing (CIRCLE-seq) [ 114 , 115 ] and circularization for high-throughput analysis of nuclease genome-wide effects by sequencing (CHANGE-seq) [ 116 ] were developed for accurate sequence confirmation.

CRISPR screening application in cancer therapy

The application of the CRISPR/Cas system for cancer therapy has been investigated using viral vectors including lentivirus, adenovirus, and AAV vectors, as well as non-viral vectors such as polymer nanoparticles, golden nanoparticles, or lipid nanoparticles in both ex vivo and in vivo circumstances as described in recent reviews [ 117 , 118 ]. Various cancer cell lines [ 2 , 4 , 87 , 119 , 120 ], T-cells via chimeric antigen receptor (CAR) integration or CAR-T system [ 90 , 121 , 122 ], and organoids derived from patient samples [ 123 ] have been explored for cancer therapy research. However, because of manipulation limitations in highly differentiated cells, in vivo clinical precision therapy involving modified cells with AAV vector delivery for the CRISPR modification system is widely used for a broad range of human diseases [ 118 ]. In this part, we mainly focus on the application of CRISPR screening system for cancer therapy, including ex vivo and in vivo approaches. Schematic representation of CRISPR screening applications for cancer research is summarized in Fig.  4 .

CRISPR screening and its applications in ex vivo and in vivo for cancer therapy. (A) CRISPR screening application in cultured cells. (B) CRISPR screening in vivo application in mouse with direct injection to organs and indirect injection in abdominal and tail vein. (C) Schematic representation of CRISPR screening applications for human cancers; Created with BioRender.com

CRISPR screening in vitro for cancer therapy

CRISPR screening has several potential applications in cancer therapy, including modified T cells and Chimeric antigen receptor CAR-T cancer treatment, novel target identification, drug resistance, drug selection exploration and so on [ 4 , 29 ]. The CRISPR screening system has been employed to investigate various cancer cell types originating from diverse organs including lymphatic system, esophagus, stomach, intestines, lungs, nervous system, skin, liver, blood cells as well as reproductive organs. CRISPR screening applications in Cancer therapy were summarized in Table  2 .

Modified T cell and CAR-T therapy for cancer therapy

Immune system is the most important defender to fight off cancer. Immunotherapy strategy is to make better immune cells such as tumor-infiltrating lymphocytes (TIL) or CAR-T cells to attack cancer via T-cell transfer. TIL therapy uses patient’s own lymphocytes to kill tumor, whereas CAR-T means modified T cells with specific proteins from surface of cancer cells, thus having the ability to attack tumors. In addition to Cas9 utilization, conjugated Cas12 (cCas12a) can be used for CAR-T cell generation. Using an AAV vector, Cas12a-crRNA complex showed robust efficiency to generate site-specific and precisely targeted CAR-T cells [ 149 ].

Recent review showed the importance of gamma retroviral or lentiviral vectors for CAR-T cell generation to target B-cell lymphomas and leukemias, although with complex manufacturing procedure, providing a promising “off-the-shelf” products for cancer treatment [ 150 ]. Whole-genome CRISPR/Cas9 screening was performed in CAR-T cells and co-cultured with Glioblastoma (GBM) stem cells (GSCs) to explore the PD-1 dependence genes such as TLE4 and IKZF2 for cancer treatment. Meanwhile, transduced GSCs were subjected to CAR-T challenge in order to identify enriched and depleted genes for cancer cell apoptosis [ 124 ]. Until 2021, total 3 FDA approved CAR-T therapies have been described as tisagenlecleucel, axicabtagene ciloleurel , and brexucabtagene autoleucel based one CD19-mediated CAR-T cells [ 151 ]. Although CAR-T is efficient in blood cancers, its efficiency loss impedes the treatment efficiency. To overcome refractory of B-cell malignancies, genome-scale CRISPR-Cas9 loss-of-function screens were performed, and revealed the crucial role of FADD and TNFRSF10B (TRAIL-R2) in mediating CAR-T cell cytotoxicity [ 125 ]. Except for precision CAR-T treatment, multiplexed CRISPR-Cas9 editing applications have been used to generate universal CAR-T products, with the aim of enhancing antitumor efficacy and improving safety of cell-based therapies [ 152 ].

Novel targets identification using CRISPR/Cas9 screening in cancer research

The invasion and metastasis of cancers make it more difficult to treat, and new targets should be identified for complete cure. Using genome-wide CRISPR/Cas9 screening, key drivers for invasion and metastasis of esophageal squamous cell carcinoma (ESCC) were identified by gain- and loss-of-function experiments, demonstrating that high expression of Mesoderm Specific Transcript (MEST), interacting with purine rich element binding Protein A, is associated with poor patient survival via activating SRCIN1/RASAL1-ERK-snail signaling [ 126 ]. Synergistical effect of genetic deletion and pharmacologic inhibition to increase cytotoxicity of MEK signaling inhibitors in pancreatic ductal adenocarcinoma cells was also investigated by CRISPR knockout screening [ 127 ]. Genome wide CRISPR/Cas9 knockout screening identified Zinc finger protein (ZNF) family member ZNF319 as a potent suppressor responsible for metastasis of breast cancer in an orthotopic murine model [ 153 ]. In hepatocellular carcinoma (HCC), CRISPR/Cas9 knockout library screening revealed the crucial role of pyruvate metabolism in HCC treatment, particularly when combined with a glutamine-deficient diet, showing the targetable metabolic vulnerabilities of pyruvate dehydrogenase α(PDHA), pyruvate dehydrogenase β(PDHB), and pyruvate carboxylase (PC) [ 128 ]. CRISPR-Cas9 knockout mutagenesis to exons encoding functional protein domains was performed to screen drug targets and dependencies, providing a comprehensive identification of protein domains for cancer cell sustainment [ 120 ]. In epithelial ovarian cancer (EOC), CRISPR-Cas9 screening combined with olaparib treatment successfully identified five genes, ATM, NBN, MUS81, RAD51B, and BRCA2, as predictive markers for olaparib response. Additionally, CDK12 emerged as a promising therapeutic target for EOC without compromising the efficacy of Olaparib response [ 129 ]. The whole-genome CRISPR screening in Guanine nucleotide-binding protein G(q) subunit alpha (GNAQ) mutant uveal melanoma (UM) cells showed that a member of Gα protein family Gαq promoted PI3K/AKT signaling pathway through focal adhesion kinase (FAK) for cell growth and survival [ 130 ].

Combinatorial CRISPR screening with scRNA-seq showed that driver gene alterations influenced TSGs, and triggered tumorigenesis in human mammary epithelial cells, indicating the impact of transcriptional epistasis on oncogenic mediators and potential therapeutic targets, including CDK4, SRPK1, and DNMT1 [ 131 ]. By analyzing the CRISPR-Cas9 screening data from Depmap (Cancer Dependency Map) and TCGA data of differentially expressed genes, the cell cycle pathway was identified as a key pathway of cell viability regulation in breast cancer patients [ 154 ]. The CRISPR/Cas9 screening in chemo-resistant small-cell lung cancer (SCLC) identified serine/threonine kinase cell division cycle 7 (CDC7) as a potential synergistic target. Combination of CDC7 inhibitor XL413 and chemotherapy led to apoptosis of chemo-sensitive SCLC in xenograft tumor [ 132 ]. Acute Myeloid Leukemia (AML) cell lines such as MOLM-13, MV4-11, HL-60, OCI-AML2, OCI-AML3 were examined for therapeutic targets via genome-wide CRISPR screening, indicating KAT2A inhibition as a therapeutic strategy in AML [ 133 ].

Hepatocellular carcinoma (HCC) was examined via CRISPRa for growth and metastasis driver genes. High MYADML2 protein level reduced sensitivity to chemotherapeutic drugs and led to worse survival [ 134 ]. Essential single nucleotide polymorphisms (SNPs) for PrCa proliferation were explored via dCas9-KRAB negative screening with 2166 candidate SNP sites in 9133 gRNAs. RIGOR program analysis identified 117 SNPs which tended to reside near 5 kb flanking the transcription start sites. SNP (rs60464856) site targeting in stable dCas9 expressing cell line showed significant down regulation of RUVBL1 gene, and further validation showed that RUVBL1 was associated with tumorigenesis [ 135 ]. dCas9-KRAB perturbation genome screening identified 470 high-confidence cis enhancer-gene pairs in 5920 enhancers in chronic myelogenous leukemia cell K562, facilitating the large-scale mapping of enhancer-gene regulatory interaction network [ 49 ].

The utilization of CRISPR-Cas9 in investigating drug resistance against tumors

Resistance to nucleotide analog 6- thioguanine was examined by genome-scale knockout screen in two human cell lines, identified DNA mismatch repair pathway, DNA topoisomerase II (TOP2A) and cyclin-dependent kinase 6, (CDK) for DNA topoisomerase II (TOP2A) poison etoposide, demonstrating Cas9/ sgRNA screens as a powerful tool for systematic genetic analysis in mammalian cells [ 26 ]. CRISPR knockout screening in human A549 lung adenocarcinoma cells identified 5 EGF-resistance genes, and further RNAi validation showed DUSP1 increased survival of EGF treated cells, providing a novel target for EGFR-overexpressing cancers [ 86 ]. Genome-wide knockout screening using CRISPR-Cas9 was also carried out in respiratory cancers, including Nasopharyngeal carcinoma (NPC) and lung cancer (LC). Nine genes were found to be associated with radiosensitivity of NPC cells (C666-1R, 6-8FR). Fanconi anemia pathway and the TGF-β signaling pathway were reported to be important contributors for radiosensitivity [ 136 ]. In the nervous system, neuroblastoma tumorigenesis was investigated via CRISPR genome-wide knockout screening, showed that ubiquitin-specific proteases (USPs) stabilize and increase half-life of repressor element-1 silencing transcription factor (REST), indicating its critical role in neuroblastoma generation [ 137 ]. As for reproductive cancers, drug resistance genes as well as lethal genes for cancer cell were identified. Genome-scale screening in ovarian cancer cell lines with the GeCKO library identified one previously validated gene SULF1 and a novel gene ZNF587B responsible for cisplatin resistance [ 138 ]. Cervical cancer cell lines such as Hela and Siha were incubated with cisplatin or paclitaxel, respectively, and screened by genome-scale CRISPR/Cas knockout library and ninety-seven genes were identified to be associated with drug resistance [ 139 ]. Prostate cancer (PrCa) is one of the most lethal causes of cancer-related death in males. Resistance to Enzalutamide, docetaxel, and Cabazitaxel in metastatic castration-resistant prostate cancer (mCRPC) is a big obstacle for cancer treatment of male patients. Whole-genome CRISPR/Cas9 knockout screening in mCRPC cell line C4 dissected the potential genes responsible for drug resistance. Two genes (IP6K2, XPO4) were validated after the screening process via bioinformatic prediction, highlighting the necessity to perform individualized validation [ 140 ].

Phase III clinical trial for Aurora-A (AURKA) inhibitor alisertib (MLN8237) in breast cancer failed to prolong patients’ survival. Rational drug combinations for better therapeutic outcome were carried out based on CRISPR/Cas9 knockout screening of 507 kinases, identifying synthetic lethality interactions with MLN8237 and Haspin (GSG2). The combination of MLN8237 and Haspin inhibitor CHR-6494 reduced tumor growth both in vitro and in vivo [ 141 ]. CRISPR screening for 656 E3 ubiquitin ligases in PrCa cells identified 51 genes as tumor repressors. The novel oncodriver Ring Finger Protein 19 A (RNF19A) was frequently amplified and highly expressed in PrCa. It correlated with castration resistance and ubiquitylated Thyroid Hormone Receptor Interactor 13 (TRIP13) and was activated by androgen receptor (AR), and Hypoxia Inducible Factor 1 Subunit Alpha (HIF1A), indicating AR/HIF1A-RNF19A-TRIP13 signaling axis for PrCa therapy [ 142 ].

Colorectal cancer (CRC) was examined for drug resistance to oxaliplatin and screened by CRISPR/Cas9 genome-wide library knockdown system. It found that low expression of mitochondrial elongation factor 2 (MIEF2) contributed to oxaliplatin drug resistance by reducing mitochondrial stability and inhibiting apoptosis via decreased cytochrome C release [ 143 ]. The CRISPRa system was employed to investigate genes associated with resistance to lymphoma radiotherapy, and a total of 8 genes were screened and subsequently validated, demonstrating a significant correlation with radiotherapy resistance [ 144 ]. Patients with Cisplatin-resistant Testicular Germ Cell Tumors (TGCTs) have poor prognosis, and developments of novel therapeutic strategies are critical. CRISPRa system revealed that NEDD8-activating enzyme E1 (NAE1) was highly expressed in drug-resistant colonies of TGCT cells, and indicated that neddylation inhibitor (MLN4924) combined with cisplatin as a novel treatment option for TGCTs [ 145 ].

Utilizing CRISPR/Cas9 screening for personalized drug selection through patient-derived organoids

Organoids derived from both healthy and diseased tissues offer a valuable resource for biological or pathological investigations. Although CRISPR screening showed powerful manipulation in cancer cells lines, it is also employed for tumor organoids derived from diverse cancer patients for personalized drug selection. Suspension culture increases efficiency of culturing cancer organoids for genome-wide CRISPR-Cas9 screening and large-scale perturbation screens [ 146 ]. Human fetal hepatocyte organoids were generated to model nonalcoholic fatty liver disease (NAFLD), and CRISPR screening was utilized to identify steatosis modulators in APOB −/− and MTTP −/− organoids [ 147 ]. CRISPR-Cas9 genetic intervention and high-throughput drug screening have been applied in digestive organoids for personalized disease modeling and therapy [ 155 ]. Human Pancreatic cancer organoid biobank established from 31 distinct tumor lines was used for CRISPR/Cas9 genome editing and drug screening, indicated increased sensitivity of kinase inhibitors dasatinib and VE-821 with driver gene ARID1A mutation [ 148 ]. Drug response evaluation by in vivo CRISPR screening (DREBIC) method was used in pancreatic ductal adenocarcinoma organoid [ 127 ].

CRISPR screening in vivo for cancer therapy

In 2022, FDA approved a total of five CAR-T cell products for the treatment of B cell acute lymphoblastic leukemia or high-grade lymphomas, as well as multiple myeloma using lentiviral or γ-retroviral approaches [ 156 ]. Notably, two clinical trials (NCT05143307/NCT03872479) employed AAV as the delivery method in their studies on cancer therapy in vivo based on CAR-T cells and CRISPR/Cas system [ 117 ]. CRISPR screening system provides a robust genetic tool for in vivo elucidation of CAR-T resistance mechanisms. Loss-of-function genetic screens in an immunocompetent murine model with B-cell acute lymphoblastic leukemia (B-ALL) identified the IFNR/JAK/STAT signaling and antigen processing and presentation pathway as key factors for CAR-T resistance in vivo. In addition, natural killer (NK) cells also engage in the resistance progress [ 157 ]. Gain-of-function CRISPR activation screen in primary CD8 + T cells identified a key factor PRODH2 for improving the in vivo efficacy of CAR-T based cell killing. Augmentation of PRODH2 enhanced metabolic function of CAR-T cells as an immune booster [ 158 ].

CRISPR screening was also utilized for in vivo investigation to elucidate gene function within a whole organism or the context of complex biological systems, using lentiviral or AAV mediated sgRNA transfection in living organisms. AAV was the widely used vector for in vivo genetic therapy due to its low immunogenicity and non-pathogenic character [ 118 ]. The limitation of AAV’s vector capacity has been addressed through the recent development of a two-split intern vectors system [ 159 ], while smaller SpCas9 orthologues such as SaCas9 have demonstrated comparable editing efficiency to that of SpCas9, rendering them suitable for AAV-SaCas9 mediated in vivo genome editing [ 21 ]. Additionally, Cre-dependent and constitutive Cas9 expressing transgenic mice were established with EGFP labeling, which provides an animal model for genome-wide targeting and contributes to in vivo investigation [ 24 ].

In vivo screens were performed in mouse brain, liver, pancreases, lung and so on. The application of SpCas9 and gRNAs using AAV vectors enabled multiple gene modifications in the adult mouse brain, demonstrating its potential for genetic regulation [ 33 ]. Gliomagenesis suppressors were investigated by in vivo stereotaxic injection of AAV carrier sgRNA library in conditional-Cas9 mouse brain [ 160 ]. Autochthonous invasion of AAV-mTSGs library in Cre-inducible Cas9 mice liver led to cancer development in situ, and the mice died within 4 months [ 89 ]. NIT1 cells (a non-obese-diabetic-derived mouse beta cell line) mutated with GeCKO-v2 were subcutaneously transplanted into type 1 diabetes mouse model to identify genes contributing to autoimmune killing resistance [ 161 ]. With the AAV9-LPL gene delivery into the lung, multiple mutations of KRAS G12D , p53 and LKB1 were obtained to induce macroscopic tumors. In vivo screening for lung cancer TSGs through CRISPR/Cas9 genome-wide knockout showed that ZNF24 contributed to P65 suppression via NF-κB pathway. Combinational inhibition of KRAS, NF-κB, and PD-1 effectively shrank autochthonous Kras G12D /ZNF24 −/− lung cancers in mouse [ 162 ].

Examination of immunotherapy-treated normal and Tcra -/- mice in vivo by CRISPR screening showed the loss of CD47 caused resistance to immunotherapy. Deletion of protein tyrosine phosphatase (PTPN2) increased immunotherapy efficacy [ 88 ]. CRISPR screening identified PD-1, Tim-3, and RNA helicase Dhx37 as regulators of tumor infiltration and degranulation. Depletion of Dhx37 improved CD8 T cells efficacy towards triple-negative breast cancer in vivo, and the NF-kB signal pathway was involved in the process [ 163 ]. In vivo applications of CRISPR screening system were summarized in the following Table  3 .

Limitations and prospection

The advances of CRISPR/Cas technology and screening strategies have revolutionized genetic identification, enabling the dissection of functional genes in specific biological processes and diseases, facilitating drug selection and individualized therapy. CRISPR screening has demonstrated great potential in cancer therapy by offering methods to combat drug resistance and aggressive behaviors, as well as identifying possible gene targets for novel approaches to treat cancers. However, there are still several obstacles for CRISPR/Cas application in clinical cancer treatment, including delivery of CRISPR/Cas9 system, Off-target effect, PAM limitation, as well as multiple gene-editing [ 117 ]. In this part, we paid more attention on limitations of CRISPR screening system and CAR-T cell therapy for cancers.

Limitations of CRISPR screening system

CRISPR screening delivery primarily relies on lentiviral and AAV vectors, which are crucial tools for either ex vivo or in vivo investigation. Of which, AAV vector has the advantages with mildly immunogenic and long-term transgene expression in post-mitotic cells, making it a leading platform for in vivo cancer therapy [ 164 ]. However, AAV vector showed some drawbacks in manufacturing, packaging size limitation, vector quality control and editing specificity, as described in the recent review [ 118 ].

Except the delivery limitations, the occurrence of off-target effects and unintended mutations induced by CRISPR technology are barriers to its application in clinical therapy. SpCas9 protein showed the ability to identify PAM sequence and cut specific DNA region in the CRISPR system. Due to the tolerance of gRNA recognition and nucleotide indels in the target region, even a single guide can generate thousands of off-targets as detected by sensitive high-throughput sequencing methods such as GUIDE-Seq and CIRCLE-seq [ 138 , 143 ]. This raises concerns regarding the application of CRISPR technology in gene therapy [ 165 ]. The reason of the off-target effect is the conformational states of HNH domain. The activated conformation of HNH increases DNA cleavage efficiency for DNA double-strand break formation, leading to both on- and off-target effects [ 166 ]. To minimize the probability of off-target mutagenesis, other high-fidelity nucleases such as SpCas9-HF1, eSpCas9 and HypaCas9 were developed [ 167 , 168 ]. In addition, PAM sequence limitation for Cas9 has been broadened by the identification of KKH SaCas9 variant, which exhibits robust genome editing activities with the PAM (NNNRRT) while maintaining comparable levels of off-target effects [ 169 ].

Anti-CRISPR is another obstacle to overcome because of the restriction of targeting specificity and activities. The VI-CRISPR inhibitors acrVIA1-7 from phage exhibit the ability to block Cas13a RNA targeting and dCas13a-mediated single nucleic acid editing. Specifically, AcrVIA1, 4, 5 and 6 bind to LwaCas13a, while AcrVIA2 and 3 interact with LwaCas13-crRNA complex [ 170 ].

Limitations of CAR-T cancer therapy

Although CAR-T showed success of B-cell malignance treatment, its usage in solid tumors still have some limitations such as T-cell exhaustion, lack of CAR-T cell persistence, and cytokine-related toxicities. To address these challenges, CRISPR technology has been used to generate safe and potent allogeneic universal CAR-T cell products for cancer immunotherapy [ 152 ]. However, hurdles remain for solid tumor CAR-T therapy due to target antigen heterogeneity, unable to pass through vascular endothelium to target tumor cells, and the immunosuppressive tumor microenvironments [ 171 ]. As viral vectors are commonly used for delivering CAR-T cells, safety concerns have arisen. To address this issue, virus-free CRISPR-CAR (VFC-CAR) T cells were generated [ 172 ]. Virus-free CAR-T cells (PD1-19bbz) were generated and a clinical trial was performed and registered at www.clinicaltrials.gov (NCT04213469) [ 173 ].

Future perspectives

Given the capacity of CRISPR to precisely modify the human genome in cells, ethical considerations have emerged as a pivotal factor for its application in genetic manipulation [ 174 , 175 , 176 ]. The challenges posed by off-target effects and unintended mutations serve as barriers to the clinical implementation of CRISPR technology. However, extensive efforts have been made to mitigate these concerns through the development of novel strategies, rendering CRISPR technologies indispensable tools for elucidating gene functions and noncoding elements involved in tumorigenesis, as well as facilitating the creation of next-generation cancer immunotherapies. In summary, CRISPR/Cas system continues to play an essential role in advancing human cancer research and clinical therapy.

Data availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Abbreviations

Clustered regularly interspaced short palindromic repeats

CRISPR-associated protein 9

Streptococcus pyogenes Cas9

Staphylococcus aureus Cas9

Streptococcus thermophilus Cas9

CRISPR from Prevotella and Francisella 1

Adeno-associated vector

Trans-activating crRNA

Protospacer adjacent motif

Ribonucleoprotein

Cystic fibrosis transmembrane conductance regulator

Cas9 nickase

Adenine base-editors

Cytosine base-editors

Interhomolog homologous recombination

CDD conserved protein domain family

Conserved Protein Domain Family HNH, His-Asn-His (HNH)

CRISPR inhibition

CRISPR activation

Krüppel-associated box

Tetracycline Inducible Expression promoter

RNA polymerase

Transcriptional activator consists of four copies of VP16

Cpf1 enzyme from Acidominococcus sp. BV3L6

Cpf1 enzyme from Lachnospiraceae bacterium ND2006

Homology-directed repair

Cas12b enzyme from Bacillus hisashii

Specific High-Sensitivity Enzymatic Reporter UnLOCKing

SARS-coronavirus-2

Open reading frame

Cell free DNA

porcine reproductive and respiratory syndrome virus

RNA viruses’ chikungunya

Catalytically inactive Cas13b

Programmable A to I Replacement

Toronto KnockOut

Genome-scale CRISPR-Cas9 knockout version 1

Model-based Analysis of Genome-wide CRISPR/Cas9 Knockout

Synthetic activation mediator

Transcription start site

Expression quantitative trait locus

The Cancer Genome Atlas

Genome-wide CRISPR/Cas9 Knockout

Integration deficient lentiviral

High-throughput genome-wide translocation sequencing

Breaks labeling, enrichment on streptavidin and next-generation sequencing

Genome-wide unbiased identification of DSBs enabled by sequencing

Discovery of in situ Cas off-targets and verification by sequencing

Digested genome sequencing

Selective enrichment and identification of adapter-tagged DNA ends by sequencing

Circularization for in vitro reporting of cleavage effects by sequencing

Circularization for high-throughput analysis of nuclease genome-wide ffects by sequencing

Chimeric antigen receptor

Tumor-infiltrating lymphocytes

Glioblastoma (GBM) stem cells

Programmed Cell Death Ligand 1

Esophageal squamous cell carcinoma

Mesoderm Specific Transcript

Zinc finger protein

Hepatocellular carcinoma

Pyruvate carboxylase

Epithelial ovarian cancer

Guanine nucleotide-binding protein G(q) subunit alpha

Uveal melanoma

Focal adhesion kinase

Tumor suppressor genes

Cancer Dependency Map

Chemo-resistant small-cell lung cancer

Serine/threonine kinase cell division cycle 7

Acute Myeloid Leukemia

Single nucleotide polymorphisms

DNA topoisomerase II

Cyclin-dependent kinase 6

Nasopharyngeal carcinoma

Lung cancer

Ubiquitin-specific proteases

Repressor element-1 silencing transcription factor

Prostate cancer

Colorectal cancer

Mitochondrial elongation factor 2

NEDD8-activating enzyme E1

Nonalcoholic fatty liver disease

B-cell acute lymphoblastic leukemia

Protein tyrosine phosphatase

Advancing Cancer Therapy. Nat Cancer. 2021;2:245–6. https://doi.org/10.1038/s43018-021-00192-x .

Article   Google Scholar  

Huang CH, Lee KC, Doudna JA. Applications of CRISPR-Cas enzymes in Cancer therapeutics and detection. Trends Cancer. 2018;4:499–512. https://doi.org/10.1016/j.trecan.2018.05.006 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Wang M, Chen M, Wu X, Huang X, Yu B. CRISPR applications in cancer diagnosis and treatment. Cell Mol Biol Lett. 2023;28:73. https://doi.org/10.1186/s11658-023-00483-4 .

Article   PubMed   PubMed Central   Google Scholar  

Yang X, Zhang B. A review on CRISPR/Cas: a versatile tool for cancer screening, diagnosis, and clinic treatment. Funct Integr Genomics. 2023;23:182. https://doi.org/10.1007/s10142-023-01117-w .

Article   CAS   PubMed   Google Scholar  

Haurwitz RE, Jinek M, Wiedenheft B, Zhou K, Doudna JA. Sequence- and structure-specific RNA processing by a CRISPR endonuclease. Science. 2010;329:1355–8. https://doi.org/10.1126/science.1192272 .

Wiedenheft B, Zhou K, Jinek M, Coyle SM, Ma W, Doudna JA. Structural basis for DNase activity of a conserved protein implicated in CRISPR-mediated genome defense. Cell. 2009;17:904–12. https://doi.org/10.1016/j.str.2009.03.019 .

Article   CAS   Google Scholar  

Mojica FJ, Diez-Villasenor C, Garcia-Martinez J, Soria E. Intervening sequences of regularly spaced prokaryotic repeats derive from foreign genetic elements. J Mol Evol. 2005;60:174–82. https://doi.org/10.1007/s00239-004-0046-3 .

Gasiunas G, Barrangou R, Horvath P, Siksnys V. Cas9-crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria. Proc Natl Acad Sci U S A. 2012;109:E2579–2586. https://doi.org/10.1073/pnas.1208507109 .

Jore MM, Lundgren M, van Duijn E, Bultema JB, Westra ER, Waghmare SP, Wiedenheft B, Pul U, Wurm R, Wagner R, et al. Structural basis for CRISPR RNA-guided DNA recognition by Cascade. Nat Struct Mol Biol. 2011;18:529–36. https://doi.org/10.1038/nsmb.2019 .

Sternberg SH, Haurwitz RE, Doudna JA. Mechanism of substrate selection by a highly specific CRISPR endoribonuclease. RNA. 2012;18:661–72. https://doi.org/10.1261/rna.030882.111 .

Makarova KS, Wolf YI, Alkhnbashi OS, Costa F, Shah SA, Saunders SJ, Barrangou R, Brouns SJ, Charpentier E, Haft DH, et al. An updated evolutionary classification of CRISPR-Cas systems. Nat Rev Microbiol. 2015;13:722–36. https://doi.org/10.1038/nrmicro3569 .

Koonin EV, Makarova KS, Zhang F. Diversity, classification and evolution of CRISPR-Cas systems. Curr Opin Microbiol. 2017;37:67–78. https://doi.org/10.1016/j.mib.2017.05.008 .

Makarova KS, Zhang F, Koonin EV. SnapShot: class 1 CRISPR-Cas systems. Cell. 2017;168:946–946. https://doi.org/10.1016/j.cell.2017.02.018 .

Makarova KS, Zhang F, Koonin EV. SnapShot: class 2 CRISPR-Cas systems. Cell. 2017;168:328–328. https://doi.org/10.1016/j.cell.2016.12.038 .

Makarova KS, Wolf YI, Iranzo J, Shmakov SA, Alkhnbashi OS, Brouns SJJ, Charpentier E, Cheng D, Haft DH, Horvath P, et al. Evolutionary classification of CRISPR-Cas systems: a burst of class 2 and derived variants. Nat Rev Microbiol. 2020;18:67–83. https://doi.org/10.1038/s41579-019-0299-x .

Tang Y, Fu Y. Class 2 CRISPR/Cas: an expanding biotechnology toolbox for and beyond genome editing. Cell Biosci. 2018;8:59. https://doi.org/10.1186/s13578-018-0255-x .

LaManna CM, Pyhtila B, Barrangou R. Sharing the CRISPR Toolbox with an Expanding Community. CRISPR J. 2020;3:248–52. https://doi.org/10.1089/crispr.2020.0075 .

Article   PubMed   Google Scholar  

Harrington LB, Burstein D, Chen JS, Paez-Espino D, Ma E, Witte IP, Cofsky JC, Kyrpides NC, Banfield JF, Doudna JA. Programmed DNA destruction by miniature CRISPR-Cas14 enzymes. Science. 2018;362:839–42. https://doi.org/10.1126/science.aav4294 .

Shmakov S, Smargon A, Scott D, Cox D, Pyzocha N, Yan W, Abudayyeh OO, Gootenberg JS, Makarova KS, Wolf YI, et al. Diversity and evolution of class 2 CRISPR-Cas systems. Nat Rev Microbiol. 2017;15:169–82. https://doi.org/10.1038/nrmicro.2016.184 .

Liu TY, Doudna JA. Chemistry of Class 1 CRISPR-Cas effectors: binding, editing, and regulation. J Biol Chem. 2020;295:14473–87. https://doi.org/10.1074/jbc.REV120.007034 .

Ran FA, Cong L, Yan WX, Scott DA, Gootenberg JS, Kriz AJ, Zetsche B, Shalem O, Wu X, Makarova KS, et al. In vivo genome editing using Staphylococcus aureus Cas9. Nature. 2015;520:186–91. https://doi.org/10.1038/nature14299 .

Cong L, Zhang F. Genome engineering using CRISPR-Cas9 system. Methods Mol Biol. 2015;1239:197–217. https://doi.org/10.1007/978-1-4939-1862-1_10 .

Zhang F, Wen Y, Guo X. CRISPR/Cas9 for genome editing: progress, implications and challenges. Hum Mol Genet. 2014;23:R40–46. https://doi.org/10.1093/hmg/ddu125 .

Platt RJ, Chen S, Zhou Y, Yim MJ, Swiech L, Kempton HR, Dahlman JE, Parnas O, Eisenhaure TM, Jovanovic M, et al. CRISPR-Cas9 knockin mice for genome editing and cancer modeling. Cell. 2014;159:440–55. https://doi.org/10.1016/j.cell.2014.09.014 .

Chen S, Sanjana NE, Zheng K, Shalem O, Lee K, Shi X, Scott DA, Song J, Pan JQ, Weissleder R, et al. Genome-wide CRISPR screen in a mouse model of tumor growth and metastasis. Cell. 2015;160:1246–60. https://doi.org/10.1016/j.cell.2015.02.038 .

Wang T, Wei JJ, Sabatini DM, Lander ES. Genetic screens in human cells using the CRISPR-Cas9 system. Science. 2014;343:80–4. https://doi.org/10.1126/science.1246981 .

Joung J, Konermann S, Gootenberg JS, Abudayyeh OO, Platt RJ, Brigham MD, Sanjana NE, Zhang F. Genome-scale CRISPR-Cas9 knockout and transcriptional activation screening. Nat Protoc. 2017;12:828–63. https://doi.org/10.1038/nprot.2017.016 .

Shalem O, Sanjana NE, Hartenian E, Shi X, Scott DA, Mikkelson T, Heckl D, Ebert BL, Root DE, Doench JG, Zhang F. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science. 2014;343:84–7. https://doi.org/10.1126/science.1247005 .

Holcomb EA, Pearson AN, Jungles KM, Tate A, James J, Jiang L, Huber AK, Green MD. High-content CRISPR screening in tumor immunology. Front Immunol. 2022;13:1041451. https://doi.org/10.3389/fimmu.2022.1041451 .

Henkel L, Rauscher B, Schmitt B, Winter J, Boutros M. Genome-scale CRISPR screening at high sensitivity with an empirically designed sgRNA library. BMC Biol. 2020;18:174. https://doi.org/10.1186/s12915-020-00905-1 .

Kumar N, Stanford W, de Solis C, Aradhana, Abraham ND, Dao TJ, Thaseen S, Sairavi A, Gonzalez CU, Ploski JE. The development of an AAV-Based CRISPR SaCas9 genome editing system that can be delivered to neurons in vivo and regulated via doxycycline and cre-recombinase. Front Mol Neurosci. 2018;11:413. https://doi.org/10.3389/fnmol.2018.00413 .

Liao HK, Hatanaka F, Araoka T, Reddy P, Wu MZ, Sui Y, Yamauchi T, Sakurai M, O’Keefe DD, Nunez-Delicado E et al. In vivo target gene activation via CRISPR/Cas9-Mediated trans-epigenetic modulation. Cell 2017, 171:1495–507 https://doi.org/10.1016/j.cell.2017.10.025 .

Swiech L, Heidenreich M, Banerjee A, Habib N, Li Y, Trombetta J, Sur M, Zhang F. In vivo interrogation of gene function in the mammalian brain using CRISPR-Cas9. Nat Biotechnol. 2015;33:102–6. https://doi.org/10.1038/nbt.3055 .

Hsu PD, Lander ES, Zhang F. Development and applications of CRISPR-Cas9 for genome engineering. Cell. 2014;157:1262–78. https://doi.org/10.1016/j.cell.2014.05.010 .

Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier E. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012;337:816–21. https://doi.org/10.1126/science.1225829 .

Sternberg SH, Redding S, Jinek M, Greene EC, Doudna JA. DNA interrogation by the CRISPR RNA-guided endonuclease Cas9. Nature. 2014;507:62–7. https://doi.org/10.1038/nature13011 .

Lin S, Staahl BT, Alla RK, Doudna JA. Enhanced homology-directed human genome engineering by controlled timing of CRISPR/Cas9 delivery. Elife. 2014;3:e04766. https://doi.org/10.7554/eLife.04766 .

Schwank G, Koo BK, Sasselli V, Dekkers JF, Heo I, Demircan T, Sasaki N, Boymans S, Cuppen E, van der Ent CK, et al. Functional repair of CFTR by CRISPR/Cas9 in intestinal stem cell organoids of cystic fibrosis patients. Cell Stem Cell. 2013;13:653–8. https://doi.org/10.1016/j.stem.2013.11.002 .

Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, Hsu PD, Wu X, Jiang W, Marraffini LA, Zhang F. Multiplex genome engineering using CRISPR/Cas systems. Science. 2013;339:819–23. https://doi.org/10.1126/science.1231143 .

Ran FA, Hsu PD, Lin CY, Gootenberg JS, Konermann S, Trevino AE, Scott DA, Inoue A, Matoba S, Zhang Y, Zhang F. Double nicking by RNA-guided CRISPR Cas9 for enhanced genome editing specificity. Cell. 2013;154:1380–9. https://doi.org/10.1016/j.cell.2013.08.021 .

Dickson KA, Field N, Blackman T, Ma Y, Xie T, Kurangil E, Idrees S, Rathnayake SNH, Mahbub RM, Faiz A, Marsh DJ. CRISPR single base-editing: in silico predictions to variant clonal cell lines. Hum Mol Genet. 2023;32:2704–16. https://doi.org/10.1093/hmg/ddad105 .

Kantor A, McClements ME, MacLaren RE. CRISPR-Cas9 DNA base-editing and prime-editing. Int J Mol Sci. 2020. https://doi.org/10.3390/ijms21176240 . 21.

Tomita A, Sasanuma H, Owa T, Nakazawa Y, Shimada M, Fukuoka T, Ogi T, Nakada S. Inducing multiple nicks promotes interhomolog homologous recombination to correct heterozygous mutations in somatic cells. Nat Commun. 2023;14:5607. https://doi.org/10.1038/s41467-023-41048-5 .

Renouf B, Piganeau M, Ghezraoui H, Jasin M, Brunet E. Creating cancer translocations in human cells using Cas9 DSBs and nCas9 paired nicks. Methods Enzymol. 2014;546:251–71. https://doi.org/10.1016/B978-0-12-801185-0.00012-X .

Martin Jinek KC, Ines Fonfara M, Hauer JA. Doudna,Emmanuelle Charpentier: a programmable dual RNA guided DNA endonuclease in adaptive bacterial immunity. Science. 2012. https://doi.org/10.1126/science.1225829 .

Gilbert LA, Horlbeck MA, Adamson B, Villalta JE, Chen Y, Whitehead EH, Guimaraes C, Panning B, Ploegh HL, Bassik MC, et al. Genome-scale CRISPR-Mediated control of gene repression and activation. Cell. 2014;159:647–61. https://doi.org/10.1016/j.cell.2014.09.029 .

Mandegar MA, Huebsch N, Frolov EB, Shin E, Truong A, Olvera MP, Chan AH, Miyaoka Y, Holmes K, Spencer CI, et al. CRISPR Interference efficiently induces specific and reversible gene silencing in human iPSCs. Cell Stem Cell. 2016;18:541–53. https://doi.org/10.1016/j.stem.2016.01.022 .

Gilbert LA, Larson MH, Morsut L, Liu Z, Brar GA, Torres SE, Stern-Ginossar N, Brandman O, Whitehead EH, Doudna JA, et al. CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes. Cell. 2013;154:442–51. https://doi.org/10.1016/j.cell.2013.06.044 .

Gasperini M, Hill AJ, McFaline-Figueroa JL, Martin B, Kim S, Zhang MD, Jackson D, Leith A, Schreiber J, Noble WS, et al. A genome-wide Framework for Mapping Gene Regulation via Cellular Genetic screens. Cell. 2019;176:377–90. https://doi.org/10.1016/j.cell.2018.11.029 .

Bikard D, Jiang W, Samai P, Hochschild A, Zhang F, Marraffini LA. Programmable repression and activation of bacterial gene expression using an engineered CRISPR-Cas system. Nucleic Acids Res. 2013;41:7429–37. https://doi.org/10.1093/nar/gkt520 .

Konermann S, Brigham MD, Trevino AE, Joung J, Abudayyeh OO, Barcena C, Hsu PD, Habib N, Gootenberg JS, Nishimasu H, et al. Genome-scale transcriptional activation by an engineered CRISPR-Cas9 complex. Nature. 2015;517:583–8. https://doi.org/10.1038/nature14136 .

Chen B, Gilbert LA, Cimini BA, Schnitzbauer J, Zhang W, Li GW, Park J, Blackburn EH, Weissman JS, Qi LS, Huang B. Dynamic imaging of genomic loci in living human cells by an optimized CRISPR/Cas system. Cell. 2013;155:1479–91. https://doi.org/10.1016/j.cell.2013.12.001 .

Wang S, Su JH, Zhang F, Zhuang X. An RNA-aptamer-based two-color CRISPR labeling system. Sci Rep. 2016;6:26857. https://doi.org/10.1038/srep26857 .

Kim HK, Song M, Lee J, Menon AV, Jung S, Kang YM, Choi JW, Woo E, Koh HC, Nam JW, Kim H. In vivo high-throughput profiling of CRISPR-Cpf1 activity. Nat Methods. 2017;14:153–9. https://doi.org/10.1038/nmeth.4104 .

Zetsche B, Gootenberg JS, Abudayyeh OO, Slaymaker IM, Makarova KS, Essletzbichler P, Volz SE, Joung J, van der Oost J, Regev A, et al. Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system. Cell. 2015;163:759–71. https://doi.org/10.1016/j.cell.2015.09.038 .

Kleinstiver BP, Tsai SQ, Prew MS, Nguyen NT, Welch MM, Lopez JM, McCaw ZR, Aryee MJ, Joung JK. Genome-wide specificities of CRISPR-Cas Cpf1 nucleases in human cells. Nat Biotechnol. 2016;34:869–74. https://doi.org/10.1038/nbt.3620 .

Kim Y, Cheong SA, Lee JG, Lee SW, Lee MS, Baek IJ, Sung YH. Generation of knockout mice by Cpf1-mediated gene targeting. Nat Biotechnol. 2016;34:808–10. https://doi.org/10.1038/nbt.3614 .

Hur JK, Kim K, Been KW, Baek G, Ye S, Hur JW, Ryu SM, Lee YS, Kim JS. Targeted mutagenesis in mice by electroporation of Cpf1 ribonucleoproteins. Nat Biotechnol. 2016;34:807–8. https://doi.org/10.1038/nbt.3596 .

Abdulrachman D, Champreda V, Eurwilaichitr L, Chantasingh D, Pootanakit K. Efficient multiplex CRISPR/Cpf1 (Cas12a) genome editing system in Aspergillus Aculeatus TBRC 277. J Biotechnol. 2022;355:53–64. https://doi.org/10.1016/j.jbiotec.2022.06.011 .

Zetsche B, Heidenreich M, Mohanraju P, Fedorova I, Kneppers J, DeGennaro EM, Winblad N, Choudhury SR, Abudayyeh OO, Gootenberg JS, et al. Multiplex gene editing by CRISPR-Cpf1 using a single crRNA array. Nat Biotechnol. 2017;35:31–4. https://doi.org/10.1038/nbt.3737 .

Dai X, Park JJ, Du Y, Kim HR, Wang G, Errami Y, Chen S. One-step generation of modular CAR-T cells with AAV-Cpf1. Nat Methods. 2019;16:247–54. https://doi.org/10.1038/s41592-019-0329-7 .

Strecker J, Jones S, Koopal B, Schmid-Burgk J, Zetsche B, Gao L, Makarova KS, Koonin EV, Zhang F. Engineering of CRISPR-Cas12b for human genome editing. Nat Commun. 2019;10:212. https://doi.org/10.1038/s41467-018-08224-4 .

Huang CJ, Adler BA, Doudna JA. A naturally DNase-free CRISPR-Cas12c enzyme silences gene expression. Mol Cell. 2022;82:2148–60. https://doi.org/10.1016/j.molcel.2022.04.020 .

Yin L, Zhao F, Sun H, Wang Z, Huang Y, Zhu W, Xu F, Mei S, Liu X, Zhang D, et al. CRISPR-Cas13a inhibits HIV-1 infection. Mol Ther Nucleic Acids. 2020;21:147–55. https://doi.org/10.1016/j.omtn.2020.05.030 .

Zheng X, Li Y, Yuan M, Shen Y, Chen S, Duan G. Rapid detection of HPV16/18 based on a CRISPR-Cas13a/Cas12a dual-channel system. Anal Methods. 2022;14:5065–75. https://doi.org/10.1039/d2ay01536f .

Xu C, Zhou Y, Xiao Q, He B, Geng G, Wang Z, Cao B, Dong X, Bai W, Wang Y, et al. Programmable RNA editing with compact CRISPR-Cas13 systems from uncultivated microbes. Nat Methods. 2021;18:499–506. https://doi.org/10.1038/s41592-021-01124-4 .

Zhang C, Konermann S, Brideau NJ, Lotfy P, Wu X, Novick SJ, Strutzenberg T, Griffin PR, Hsu PD, Lyumkis D. Structural basis for the RNA-Guided ribonuclease activity of CRISPR-Cas13d. Cell. 2018;175:212–23. https://doi.org/10.1016/j.cell.2018.09.001 . e217.

Tong H, Huang J, Xiao Q, He B, Dong X, Liu Y, Yang X, Han D, Wang Z, Wang X, et al. High-fidelity Cas13 variants for targeted RNA degradation with minimal collateral effects. Nat Biotechnol. 2023;41:108–19. https://doi.org/10.1038/s41587-022-01419-7 .

Gootenberg JS, Abudayyeh OO, Lee JW, Essletzbichler P, Dy AJ, Joung J, Verdine V, Donghia N, Daringer NM, Freije CA, et al. Nucleic acid detection with CRISPR-Cas13a/C2c2. Science. 2017;356:438–42. https://doi.org/10.1126/science.aam9321 .

Fozouni P, Son S, Diaz de Leon Derby M, Knott GJ, Gray CN, D’Ambrosio MV, Zhao C, Switz NA, Kumar GR, Stephens SI, et al. Amplification-free detection of SARS-CoV-2 with CRISPR-Cas13a and mobile phone microscopy. Cell. 2020. https://doi.org/10.1016/j.cell.2020.12.001 .

Yang L, Zhang Y, Yi W, Dong X, Niu M, Song Y, Han Y, Li H, Sun Y. A rapid and efficient platform for antiviral crRNA screening using CRISPR-Cas13a-based nucleic acid detection. Front Immunol. 2023;141116230. https://doi.org/10.3389/fimmu.2023.1116230 .

Gootenberg JS, Abudayyeh OO, Kellner MJ, Joung J, Collins JJ, Zhang F. Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and Csm6. Science. 2018;360:439–44. https://doi.org/10.1126/science.aaq0179 .

Kellner MJ, Koob JG, Gootenberg JS, Abudayyeh OO, Zhang F. SHERLOCK: nucleic acid detection with CRISPR nucleases. Nat Protoc. 2019;14:2986–3012. https://doi.org/10.1038/s41596-019-0210-2 .

Cui J, Techakriengkrai N, Nedumpun T, Suradhat S. Abrogation of PRRSV infectivity by CRISPR-Cas13b-mediated viral RNA cleavage in mammalian cells. Sci Rep. 2020;10:9617. https://doi.org/10.1038/s41598-020-66775-3 .

Tng PYL, Carabajal Paladino L, Verkuijl SAN, Purcell J, Merits A, Leftwich PT, Fragkoudis R, Noad R, Alphey L. Cas13b-dependent and Cas13b-independent RNA knockdown of viral sequences in mosquito cells following guide RNA expression. Commun Biol. 2020;3413. https://doi.org/10.1038/s42003-020-01142-6 .

Fareh M, Zhao W, Hu W, Casan JML, Kumar A, Symons J, Zerbato JM, Fong D, Voskoboinik I, Ekert PG, et al. Reprogrammed CRISPR-Cas13b suppresses SARS-CoV-2 replication and circumvents its mutational escape through mismatch tolerance. Nat Commun. 2021;12:4270. https://doi.org/10.1038/s41467-021-24577-9 .

Yu D, Han HJ, Yu J, Kim J, Lee GH, Yang JH, Song BM, Tark D, Choi BS, Kang SM, Heo WD. Pseudoknot-targeting Cas13b combats SARS-CoV-2 infection by suppressing viral replication. Mol Ther. 2023;31:1675–87. https://doi.org/10.1016/j.ymthe.2023.03.018 .

Cox DBT, Gootenberg JS, Abudayyeh OO, Franklin B, Kellner MJ, Joung J, Zhang F. RNA editing with CRISPR-Cas13. Science. 2017;358:1019–27. https://doi.org/10.1126/science.aaq0180 .

Granados-Riveron JT, Aquino-Jarquin G. CRISPR-Cas13 Precision Transcriptome Engineering in Cancer. Cancer Res. 2018;78:4107–13. https://doi.org/10.1158/0008-5472.CAN-18-0785 .

Palaz F, Kalkan AK, Can O, Demir AN, Tozluyurt A, Ozcan A, Ozsoz M. CRISPR-Cas13 System as a Promising and Versatile Tool for Cancer diagnosis, therapy, and Research. ACS Synth Biol. 2021;10:1245–67. https://doi.org/10.1021/acssynbio.1c00107 .

Koike-Yusa H, Li Y, Tan EP, Velasco-Herrera Mdel C, Yusa K. Genome-wide recessive genetic screening in mammalian cells with a lentiviral CRISPR-guide RNA library. Nat Biotechnol. 2014;32:267–73. https://doi.org/10.1038/nbt.2800 .

Shalem O, Sanjana NE, Zhang F. High-throughput functional genomics using CRISPR-Cas9. Nat Rev Genet. 2015;16:299–311. https://doi.org/10.1038/nrg3899 .

Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, Zhang F, Jaenisch R. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell. 2013;153:910–8. https://doi.org/10.1016/j.cell.2013.04.025 .

Sanjana NE, Shalem O, Zhang F. Improved vectors and genome-wide libraries for CRISPR screening. Nat Methods. 2014;11:783–4. https://doi.org/10.1038/nmeth.3047 .

Najm FJ, Strand C, Donovan KF, Hegde M, Sanson KR, Vaimberg EW, Sullender ME, Hartenian E, Kalani Z, Fusi N, et al. Orthologous CRISPR-Cas9 enzymes for combinatorial genetic screens. Nat Biotechnol. 2018;36:179–89. https://doi.org/10.1038/nbt.4048 .

Kim JS, Lee JH, Jeon SR, Kim Y, Jeon SH, Wu HG. Identification of genes involved in EGF-induced apoptosis using CRISPR/Cas9 knockout screening: implications for Novel therapeutic targets in EGFR-Overexpressing cancers. Cancer Res Treat. 2023. https://doi.org/10.4143/crt.2022.1414 .

Hart T, Chandrashekhar M, Aregger M, Steinhart Z, Brown KR, MacLeod G, Mis M, Zimmermann M, Fradet-Turcotte A, Sun S, et al. High-resolution CRISPR screens Reveal Fitness genes and genotype-specific Cancer liabilities. Cell. 2015;163:1515–26. https://doi.org/10.1016/j.cell.2015.11.015 .

Manguso RT, Pope HW, Zimmer MD, Brown FD, Yates KB, Miller BC, Collins NB, Bi K, LaFleur MW, Juneja VR, et al. In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target. Nature. 2017;547:413–8. https://doi.org/10.1038/nature23270 .

Wang G, Chow RD, Ye L, Guzman CD, Dai X, Dong MB, Zhang F, Sharp PA, Platt RJ, Chen S. Mapping a functional cancer genome atlas of tumor suppressors in mouse liver using AAV-CRISPR-mediated direct in vivo screening. Sci Adv. 2018;4:eaao5508. https://doi.org/10.1126/sciadv.aao5508 .

Parnas O, Jovanovic M, Eisenhaure TM, Herbst RH, Dixit A, Ye CJ, Przybylski D, Platt RJ, Tirosh I, Sanjana NE, et al. A genome-wide CRISPR screen in primary Immune cells to Dissect Regulatory Networks. Cell. 2015;162:675–86. https://doi.org/10.1016/j.cell.2015.06.059 .

Canver MC, Haeussler M, Bauer DE, Orkin SH, Sanjana NE, Shalem O, Yuan GC, Zhang F, Concordet JP, Pinello L. Integrated design, execution, and analysis of arrayed and pooled CRISPR genome-editing experiments. Nat Protoc. 2018;13:946–86. https://doi.org/10.1038/nprot.2018.005 .

Wilson LOW, O’Brien AR, Bauer DC. The current state and future of CRISPR-Cas9 gRNA Design Tools. Front Pharmacol. 2018;9:749. https://doi.org/10.3389/fphar.2018.00749 .

Cui Y, Xu J, Cheng M, Liao X, Peng S. Review of CRISPR/Cas9 sgRNA design tools. Interdiscip Sci. 2018;10:455–65. https://doi.org/10.1007/s12539-018-0298-z .

Vakulskas CA, Behlke MA. Evaluation and reduction of CRISPR off-target cleavage events. Nucleic Acid Ther. 2019;29:167–74. https://doi.org/10.1089/nat.2019.0790 .

Concordet JP, Haeussler M. CRISPOR: intuitive guide selection for CRISPR/Cas9 genome editing experiments and screens. Nucleic Acids Res. 2018;46:W242–5. https://doi.org/10.1093/nar/gky354 .

Haeussler M, Schonig K, Eckert H, Eschstruth A, Mianne J, Renaud JB, Schneider-Maunoury S, Shkumatava A, Teboul L, Kent J, et al. Evaluation of off-target and on-target scoring algorithms and integration into the guide RNA selection tool CRISPOR. Genome Biol. 2016;17148. https://doi.org/10.1186/s13059-016-1012-2 .

Pinello L, Canver MC, Hoban MD, Orkin SH, Kohn DB, Bauer DE, Yuan GC. Analyzing CRISPR genome-editing experiments with CRISPResso. Nat Biotechnol. 2016;34:695–7. https://doi.org/10.1038/nbt.3583 .

Li W, Xu H, Xiao T, Cong L, Love MI, Zhang F, Irizarry RA, Liu JS, Brown M, Liu XS. MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screens. Genome Biol. 2014;15:554. https://doi.org/10.1186/s13059-014-0554-4 .

Wang X, Wang Y, Wu X, Wang J, Wang Y, Qiu Z, Chang T, Huang H, Lin RJ, Yee JK. Unbiased detection of off-target cleavage by CRISPR-Cas9 and TALENs using integrase-defective lentiviral vectors. Nat Biotechnol. 2015;33:175–8. https://doi.org/10.1038/nbt.3127 .

Frock RL, Hu J, Meyers RM, Ho YJ, Kii E, Alt FW. Genome-wide detection of DNA double-stranded breaks induced by engineered nucleases. Nat Biotechnol. 2015;33:179–86. https://doi.org/10.1038/nbt.3101 .

Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009;10:R25. https://doi.org/10.1186/gb-2009-10-3-r25 .

Hart T, Brown KR, Sircoulomb F, Rottapel R, Moffat J. Measuring error rates in genomic perturbation screens: gold standards for human functional genomics. Mol Syst Biol. 2014;10:733. https://doi.org/10.15252/msb.20145216 .

Lane-Reticker SK, Kessler EA, Muscato AJ, Kim SY, Doench JG, Yates KB, Manguso RT, Dubrot J. Protocol for in vivo CRISPR screening using selective CRISPR antigen removal lentiviral vectors. STAR Protoc. 2023;4:102082. https://doi.org/10.1016/j.xpro.2023.102082 .

Crosetto N, Mitra A, Silva MJ, Bienko M, Dojer N, Wang Q, Karaca E, Chiarle R, Skrzypczak M, Ginalski K, et al. Nucleotide-resolution DNA double-strand break mapping by next-generation sequencing. Nat Methods. 2013;10:361–5. https://doi.org/10.1038/nmeth.2408 .

Tsai SQ, Zheng Z, Nguyen NT, Liebers M, Topkar VV, Thapar V, Wyvekens N, Khayter C, Iafrate AJ, Le LP, et al. GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases. Nat Biotechnol. 2015;33:187–97. https://doi.org/10.1038/nbt.3117 .

Malinin NL, Lee G, Lazzarotto CR, Li Y, Zheng Z, Nguyen NT, Liebers M, Topkar VV, Iafrate AJ, Le LP, et al. Defining genome-wide CRISPR-Cas genome-editing nuclease activity with GUIDE-seq. Nat Protoc. 2021;16:5592–615. https://doi.org/10.1038/s41596-021-00626-x .

Wienert B, Wyman SK, Richardson CD, Yeh CD, Akcakaya P, Porritt MJ, Morlock M, Vu JT, Kazane KR, Watry HL, et al. Unbiased detection of CRISPR off-targets in vivo using DISCOVER-Seq. Science. 2019;364:286–9. https://doi.org/10.1126/science.aav9023 .

Wienert B, Wyman SK, Yeh CD, Conklin BR, Corn JE. CRISPR off-target detection with DISCOVER-seq. Nat Protoc. 2020;15:1775–99. https://doi.org/10.1038/s41596-020-0309-5 .

Zou RS, Liu Y, Gaido OER, Konig MF, Mog BJ, Shen LL, Aviles-Vazquez F, Marin-Gonzalez A, Ha T. Improving the sensitivity of in vivo CRISPR off-target detection with DISCOVER-Seq. Nat Methods. 2023;20:706–13. https://doi.org/10.1038/s41592-023-01840-z .

Kim D, Bae S, Park J, Kim E, Kim S, Yu HR, Hwang J, Kim JI, Kim JS. Digenome-seq: genome-wide profiling of CRISPR-Cas9 off-target effects in human cells. Nat Methods. 2015;12:237–43. https://doi.org/10.1038/nmeth.3284 .

Kim D, Kang BC, Kim JS. Identifying genome-wide off-target sites of CRISPR RNA-guided nucleases and deaminases with Digenome-Seq. Nat Protoc. 2021;16:1170–92. https://doi.org/10.1038/s41596-020-00453-6 .

Park J, Childs L, Kim D, Hwang GH, Kim S, Kim ST, Kim JS, Bae S. Digenome-Seq web tool for profiling CRISPR specificity. Nat Methods. 2017;14:548–9. https://doi.org/10.1038/nmeth.4262 .

Cameron P, Fuller CK, Donohoue PD, Jones BN, Thompson MS, Carter MM, Gradia S, Vidal B, Garner E, Slorach EM, et al. Mapping the genomic landscape of CRISPR-Cas9 cleavage. Nat Methods. 2017;14:600–6. https://doi.org/10.1038/nmeth.4284 .

Lazzarotto CR, Nguyen NT, Tang X, Malagon-Lopez J, Guo JA, Aryee MJ, Joung JK, Tsai SQ. Defining CRISPR-Cas9 genome-wide nuclease activities with CIRCLE-seq. Nat Protoc. 2018;13:2615–42. https://doi.org/10.1038/s41596-018-0055-0 .

Tsai SQ, Nguyen NT, Malagon-Lopez J, Topkar VV, Aryee MJ, Joung JK. CIRCLE-seq: a highly sensitive in vitro screen for genome-wide CRISPR-Cas9 nuclease off-targets. Nat Methods. 2017;14:607–14. https://doi.org/10.1038/nmeth.4278 .

Lazzarotto CR, Malinin NL, Li Y, Zhang R, Yang Y, Lee G, Cowley E, He Y, Lan X, Jividen K, et al. CHANGE-seq reveals genetic and epigenetic effects on CRISPR-Cas9 genome-wide activity. Nat Biotechnol. 2020;38:1317–27. https://doi.org/10.1038/s41587-020-0555-7 .

Khoshandam M, Soltaninejad H, Mousazadeh M, Hamidieh AA, Hosseinkhani S. Clinical applications of the CRISPR/Cas9 genome-editing system: delivery options and challenges in precision medicine. Genes Dis. 2024;11:268–82. https://doi.org/10.1016/j.gendis.2023.02.027 .

Wang D, Zhang F, Gao G. CRISPR-Based therapeutic genome editing: strategies and in vivo delivery by AAV vectors. Cell. 2020;181:136–50. https://doi.org/10.1016/j.cell.2020.03.023 .

Proietti L, Manhart G, Heyes E, Troester S, Grebien F. Arrayed CRISPR/Cas9 screening for the functional validation of Cancer Genetic dependencies. Bio Protoc. 2022. https://doi.org/10.21769/BioProtoc.4577 .

Shi J, Wang E, Milazzo JP, Wang Z, Kinney JB, Vakoc CR. Discovery of cancer drug targets by CRISPR-Cas9 screening of protein domains. Nat Biotechnol. 2015;33:661–7. https://doi.org/10.1038/nbt.3235 .

Schmidt R, Steinhart Z, Layeghi M, Freimer JW, Bueno R, Nguyen VQ, Blaeschke F, Ye CJ, Marson A. CRISPR activation and interference screens decode stimulation responses in primary human T cells. Science. 2022;375:eabj4008. https://doi.org/10.1126/science.abj4008 .

Shifrut E, Carnevale J, Tobin V, Roth TL, Woo JM, Bui CT, Li PJ, Diolaiti ME, Ashworth A, Marson A. Genome-wide CRISPR Screens in Primary Human T Cells Reveal Key Regulators of Immune Function. Cell 2018, 175:1958–1971 e1915. https://doi.org/10.1016/j.cell.2018.10.024 .

Geurts MH, Clevers H. CRISPR engineering in organoids for gene repair and disease modelling. Nat Reviews Bioeng. 2023;1:32–45. https://doi.org/10.1038/s44222-022-00013-5 .

Wang D, Prager BC, Gimple RC, Aguilar B, Alizadeh D, Tang H, Lv D, Starr R, Brito A, Wu Q, et al. CRISPR Screening of CAR T cells and Cancer stem cells reveals critical dependencies for cell-based therapies. Cancer Discov. 2021;11:1192–211. https://doi.org/10.1158/2159-8290.CD-20-1243 .

Dufva O, Koski J, Maliniemi P, Ianevski A, Klievink J, Leitner J, Polonen P, Hohtari H, Saeed K, Hannunen T, et al. Integrated drug profiling and CRISPR screening identify essential pathways for CAR T-cell cytotoxicity. Blood. 2020;135:597–609. https://doi.org/10.1182/blood.2019002121 .

Xu WW, Liao L, Dai W, Zheng CC, Tan XP, He Y, Zhang QH, Huang ZH, Chen WY, Qin YR, et al. Genome-wide CRISPR/Cas9 screening identifies a targetable MEST-PURA interaction in cancer metastasis. EBioMedicine. 2023;92104587. https://doi.org/10.1016/j.ebiom.2023.104587 .

Szlachta K, Kuscu C, Tufan T, Adair SJ, Shang S, Michaels AD, Mullen MG, Fischer NL, Yang J, Liu L, et al. CRISPR knockout screening identifies combinatorial drug targets in pancreatic cancer and models cellular drug response. Nat Commun. 2018;94275. https://doi.org/10.1038/s41467-018-06676-2 .

Yang C, Lee D, Zhang MS, Tse AP, Wei L, Bao MH, Wong BP, Chan CY, Yuen VW, Chen Y, Wong CC. Genome-wide CRISPR/Cas9 Library Screening revealed Dietary Restriction of glutamine in combination with inhibition of pyruvate metabolism as effective Liver Cancer Treatment. Adv Sci (Weinh). 2022;9:e2202104. https://doi.org/10.1002/advs.202202104 .

Coelho R, Tozzi A, Disler M, Lombardo F, Fedier A, Lopez MN, Freuler F, Jacob F, Heinzelmann-Schwarz V. Overlapping gene dependencies for PARP inhibitors and carboplatin response identified by functional CRISPR-Cas9 screening in ovarian cancer. Cell Death Dis. 2022;13909. https://doi.org/10.1038/s41419-022-05347-x .

Arang N, Lubrano S, Rigiracciolo DC, Nachmanson D, Lippman SM, Mali P, Harismendy O, Gutkind JS. Whole-genome CRISPR screening identifies PI3K/AKT as a downstream component of the oncogenic GNAQ-focal adhesion kinase signaling circuitry. J Biol Chem. 2023;299102866. https://doi.org/10.1016/j.jbc.2022.102866 .

Zhao X, Li J, Liu Z, Powers S. Combinatorial CRISPR/Cas9 screening reveals epistatic networks of interacting tumor suppressor genes and therapeutic targets in human breast Cancer. Cancer Res. 2021;81:6090–105. https://doi.org/10.1158/0008-5472.CAN-21-2555 .

Deng L, Yang L, Zhu S, Li M, Wang Y, Cao X, Wang Q, Guo L. Identifying CDC7 as a synergistic target of chemotherapy in resistant small-cell lung cancer via CRISPR/Cas9 screening. Cell Death Discov. 2023;9:40. https://doi.org/10.1038/s41420-023-01315-2 .

Tzelepis K, Koike-Yusa H, De Braekeleer E, Li Y, Metzakopian E, Dovey OM, Mupo A, Grinkevich V, Li M, Mazan M, et al. A CRISPR dropout screen identifies genetic vulnerabilities and therapeutic targets in Acute myeloid leukemia. Cell Rep. 2016;17:1193–205. https://doi.org/10.1016/j.celrep.2016.09.079 .

Zhang B, Ren Z, Zheng H, Lin M, Chen G, Luo OJ, Zhu G. CRISPR activation screening in a mouse model for drivers of hepatocellular carcinoma growth and metastasis. iScience. 2023;26:106099. https://doi.org/10.1016/j.isci.2023.106099 .

Tian Y, Dong D, Wu L, Park JY, Wei GH, Wang L, consortium PE. Combined CRISPR and proteomics screening reveal a cohesin-CTCF-bound allele contributing to increased expression of RUVBL1 and prostate cancer progression. bioRxiv 2023. https://doi.org/10.1101/2023.01.18.524405 .

Zhou Z, Chen G, Shen M, Li J, Liu K, Liu M, Shi S, Yang D, Chen W, Chen S, et al. Genome-scale CRISPR-Cas9 knockout screening in nasopharyngeal carcinoma for radiosensitive and radioresistant genes. Transl Oncol. 2023;30101625. https://doi.org/10.1016/j.tranon.2023.101625 .

Karapurkar JK, Kim MS, Colaco JC, Suresh B, Sarodaya N, Kim DH, Park CH, Hong SH, Kim KS, Ramakrishna S. CRISPR/Cas9-based genome-wide screening of the deubiquitinase subfamily identifies USP3 as a protein stabilizer of REST blocking neuronal differentiation and promotes neuroblastoma tumorigenesis. J Exp Clin Cancer Res. 2023;42121. https://doi.org/10.1186/s13046-023-02694-1 .

Ouyang Q, Liu Y, Tan J, Li J, Yang D, Zeng F, Huang W, Kong Y, Liu Z, Zhou H, Liu Y. Loss of ZNF587B and SULF1 contributed to cisplatin resistance in ovarian cancer cell lines based on genome-scale CRISPR/Cas9 screening. Am J Cancer Res. 2019;9:988–98.

CAS   PubMed   PubMed Central   Google Scholar  

Cao C, Liu T, Zhang Q, Li R, Zeng Z, Cui Z, Wang X, Gong D, Tian X, Hu Z. Somatic mutations and CRISPR/Cas9 library screening integrated analysis identifies cervical cancer drug-resistant pathways. Clin Transl Med. 2021;11e632. https://doi.org/10.1002/ctm2.632 .

Haldrup J, Weiss S, Schmidt L, Sorensen KD. Investigation of enzalutamide, docetaxel, and cabazitaxel resistance in the castration resistant prostate cancer cell line C4 using genome-wide CRISPR/Cas9 screening. Sci Rep. 2023;13:9043. https://doi.org/10.1038/s41598-023-35950-7 .

Chen A, Wen S, Liu F, Zhang Z, Liu M, Wu Y, He B, Yan M, Kang T, Lam EW, et al. CRISPR/Cas9 screening identifies a kinetochore-microtubule dependent mechanism for Aurora-A inhibitor resistance in breast cancer. Cancer Commun (Lond). 2021;41:121–39. https://doi.org/10.1002/cac2.12125 .

Zhang N, Huang D, Ruan X, Ng AT, Tsu JH, Jiang G, Huang J, Zhan Y, Na R. CRISPR screening reveals gleason score and castration resistance related oncodriver ring finger protein 19 A (RNF19A) in prostate cancer. Drug Resist Updat. 2023;67100912. https://doi.org/10.1016/j.drup.2022.100912 .

Xie C, Li K, Li Y, Peng X, Teng B, He K, Jin A, Wang W, Wei Z. CRISPR-based knockout screening identifies the loss of MIEF2 to enhance oxaliplatin resistance in colorectal cancer through inhibiting the mitochondrial apoptosis pathway. Front Oncol. 2022;12881487. https://doi.org/10.3389/fonc.2022.881487 .

Luo BH, Huang JQ, Huang CY, Tian P, Chen AZ, Wu WH, Ma XM, Yuan YX, Yu L. Screening of Lymphoma Radiotherapy-resistant genes with CRISPR Activation Library. Pharmgenomics Pers Med. 2023;16:67–80. https://doi.org/10.2147/PGPM.S386085 .

Funke K, Einsfelder U, Hansen A, Arevalo L, Schneider S, Nettersheim D, Stein V, Schorle H. Genome-scale CRISPR screen reveals neddylation to contribute to cisplatin resistance of testicular germ cell tumours. Br J Cancer. 2023;128:2270–82. https://doi.org/10.1038/s41416-023-02247-5 .

Price S, Bhosle S, Goncalves E, Li X, McClurg DP, Barthorpe S, Beck A, Hall C, Lightfoot H, Farrow L, et al. A suspension technique for efficient large-scale cancer organoid culturing and perturbation screens. Sci Rep. 2022;12:5571. https://doi.org/10.1038/s41598-022-09508-y .

Hendriks D, Brouwers JF, Hamer K, Geurts MH, Luciana L, Massalini S, Lopez-Iglesias C, Peters PJ, Rodriguez-Colman MJ, Chuva de Sousa Lopes S, et al. Engineered human hepatocyte organoids enable CRISPR-based target discovery and drug screening for steatosis. Nat Biotechnol. 2023. https://doi.org/10.1038/s41587-023-01680-4 .

Hirt CK, Booij TH, Grob L, Simmler P, Toussaint NC, Keller D, Taube D, Ludwig V, Goryachkin A, Pauli C, et al. Drug screening and genome editing in human pancreatic cancer organoids identifies drug-gene interactions and candidates for off-label treatment. Cell Genom. 2022;2100095. https://doi.org/10.1016/j.xgen.2022.100095 .

Ling X, Chang L, Chen H, Liu T. Efficient generation of locus-specific human CAR-T cells with CRISPR/cCas12a. STAR Protoc. 2022;3:101321. https://doi.org/10.1016/j.xpro.2022.101321 .

Wagner DL, Koehl U, Chmielewski M, Scheid C, Stripecke R. Review: sustainable clinical development of CAR-T cells - switching from viral transduction towards CRISPR-Cas Gene Editing. Front Immunol. 2022;13:865424. https://doi.org/10.3389/fimmu.2022.865424 .

Sochacka-Cwikla A, Maczynski M, Regiec A. FDA-Approved drugs for hematological malignancies-the last Decade Review. Cancers (Basel). 2021. https://doi.org/10.3390/cancers14010087 . 14.

Dimitri A, Herbst F, Fraietta JA. Engineering the next-generation of CAR T-cells with CRISPR-Cas9 gene editing. Mol Cancer. 2022;21:78. https://doi.org/10.1186/s12943-022-01559-z .

Wang L, Zhou L, Li M, Zhao J, Liu Y, Chen Y, Qin X, Wang S, Chen H, Piao Y, et al. Genome-wide CRISPR/Cas9 knockout screening uncovers ZNF319 as a novel tumor suppressor critical for breast cancer metastasis. Biochem Biophys Res Commun. 2022;589:107–15. https://doi.org/10.1016/j.bbrc.2021.12.023 .

Sun X, Wang Z, Chen X, Shen K. CRISPR-cas9 screening identified Lethal genes enriched in cell cycle pathway and of prognosis significance in breast Cancer. Front Cell Dev Biol. 2021;9:646774. https://doi.org/10.3389/fcell.2021.646774 .

Wang Q, Guo F, Jin Y, Ma Y. Applications of human organoids in the personalized treatment for digestive diseases. Signal Transduct Target Ther. 2022;7:336. https://doi.org/10.1038/s41392-022-01194-6 .

Cortese MJ, Sauter C. A CRISPR non-viral manufacturing approach for CAR T cell therapies. Mol Ther. 2022;30:3338–40. https://doi.org/10.1016/j.ymthe.2022.09.022 .

Ramos A, Koch CE, Liu-Lupo Y, Hellinger RD, Kyung T, Abbott KL, Frose J, Goulet D, Gordon KS, Eidell KP, et al. Leukemia-intrinsic determinants of CAR-T response revealed by iterative in vivo genome-wide CRISPR screening. Nat Commun. 2023;148048. https://doi.org/10.1038/s41467-023-43790-2 .

Ye L, Park JJ, Peng L, Yang Q, Chow RD, Dong MB, Lam SZ, Guo J, Tang E, Zhang Y, et al. A genome-scale gain-of-function CRISPR screen in CD8 T cells identifies proline metabolism as a means to enhance CAR-T therapy. Cell Metab. 2022;34:595–614. https://doi.org/10.1016/j.cmet.2022.02.009 .

Madigan V, Zhang F, Dahlman JE. Drug delivery systems for CRISPR-based genome editors. Nat Rev Drug Discov. 2023;22:875–94. https://doi.org/10.1038/s41573-023-00762-x .

Chow RD, Guzman CD, Wang G, Schmidt F, Youngblood MW, Ye L, Errami Y, Dong MB, Martinez MA, Zhang S, et al. AAV-mediated direct in vivo CRISPR screen identifies functional suppressors in glioblastoma. Nat Neurosci. 2017;20:1329–41. https://doi.org/10.1038/nn.4620 .

Li J, Lee YC, Iessi IL, Wu C, Yi P, Cai EP. Protocol for genome-scale in vivo CRISPR screening to study protection of beta cells under autoimmunity in a type 1 diabetes mouse model. STAR Protoc. 2023;4:102155. https://doi.org/10.1016/j.xpro.2023.102155 .

Liu L, Lei Y, Chen W, Zhou Q, Zheng Z, Zeng G, Liu W, Feng P, Zhang Z, Yu L, Chen L. In vivo genome-wide CRISPR screening identifies ZNF24 as a negative NF-kappaB modulator in lung cancer. Cell Biosci. 2022;12193. https://doi.org/10.1186/s13578-022-00933-0 .

Dong MB, Wang G, Chow RD, Ye L, Zhu L, Dai X, Park JJ, Kim HR, Errami Y, Guzman CD, et al. Systematic Immunotherapy Target Discovery using genome-scale in vivo CRISPR screens in CD8 T cells. Cell. 2019;178:1189–204. https://doi.org/10.1016/j.cell.2019.07.044 .

Wang D, Tai PWL, Gao G. Adeno-associated virus vector as a platform for gene therapy delivery. Nat Rev Drug Discov. 2019;18:358–78. https://doi.org/10.1038/s41573-019-0012-9 .

Haeussler M. CRISPR off-targets: a question of context. Cell Biol Toxicol. 2020;36:5–9. https://doi.org/10.1007/s10565-019-09497-1 .

Sternberg SH, LaFrance B, Kaplan M, Doudna JA. Conformational control of DNA target cleavage by CRISPR-Cas9. Nature. 2015;527:110–3. https://doi.org/10.1038/nature15544 .

Slaymaker IM, Gao L, Zetsche B, Scott DA, Yan WX, Zhang F. Rationally engineered Cas9 nucleases with improved specificity. Science. 2016;351:84–8. https://doi.org/10.1126/science.aad5227 .

Chen JS, Dagdas YS, Kleinstiver BP, Welch MM, Sousa AA, Harrington LB, Sternberg SH, Joung JK, Yildiz A, Doudna JA. Enhanced proofreading governs CRISPR-Cas9 targeting accuracy. Nature. 2017;550:407–10. https://doi.org/10.1038/nature24268 .

Kleinstiver BP, Prew MS, Tsai SQ, Nguyen NT, Topkar VV, Zheng Z, Joung JK. Broadening the targeting range of Staphylococcus aureus CRISPR-Cas9 by modifying PAM recognition. Nat Biotechnol. 2015;33:1293–8. https://doi.org/10.1038/nbt.3404 .

Lin P, Qin S, Pu Q, Wang Z, Wu Q, Gao P, Schettler J, Guo K, Li R, Li G, et al. CRISPR-Cas13 inhibitors Block RNA editing in Bacteria and mammalian cells. Mol Cell. 2020;78:850–61. https://doi.org/10.1016/j.molcel.2020.03.033 .

Safarzadeh Kozani P, Safarzadeh Kozani P, Ahmadi Najafabadi M, Yousefi F, Mirarefin SMJ, Rahbarizadeh F. Recent advances in solid tumor CAR-T cell therapy: driving Tumor cells from Hero to zero? Front Immunol. 2022;13:795164. https://doi.org/10.3389/fimmu.2022.795164 .

Mueller KP, Piscopo NJ, Forsberg MH, Saraspe LA, Das A, Russell B, Smerchansky M, Cappabianca D, Shi L, Shankar K, et al. Production and characterization of virus-free, CRISPR-CAR T cells capable of inducing solid tumor regression. J Immunother Cancer. 2022. https://doi.org/10.1136/jitc-2021-004446 .

Hu Y, Zu C, Zhang M, Wei G, Li W, Fu S, Hong R, Zhou L, Wu W, Cui J, et al. Safety and efficacy of CRISPR-based non-viral PD1 locus specifically integrated anti-CD19 CAR-T cells in patients with relapsed or refractory Non-hodgkin’s lymphoma: a first-in-human phase I study. EClinicalMedicine. 2023;60:102010. https://doi.org/10.1016/j.eclinm.2023.102010 .

Johnson WG, Bowman DM. Inherited regulation for advanced ARTs: comparing jurisdictions’ applications of existing governance regimes to emerging reproductive technologies. J Law Biosci. 2022;9:lsab034. https://doi.org/10.1093/jlb/lsab034 .

Cyranoski D. China set to introduce gene-editing regulation following CRISPR-baby furore. Nature. 2019. https://doi.org/10.1038/d41586-019-01580-1 .

Scott CT, Selin C. What to expect when expecting CRISPR Baby Number Four. Am J Bioeth. 2019;19:7–9. https://doi.org/10.1080/15265161.2018.1562793 .

Download references

Acknowledgements

The authors would like to thank Prof. Guokai Chen and Dr. Weiwei Liu from University of Macau, Dr. Carlos Godoy-Parejo from Icahn School of Medicine at Mount Sinai, Ms Qinru Li from University of Toronto for critical comments and wonderful advises on the manuscript. We also thank other laboratory members for helpful discussions of our review.

This work was funded by Guangdong Basic and Applied Basic Research Foundation (File No.2020A1515110045), by Zhuhai High-level Health Personnel Team Project (File No. Zhuhai HLHPTP201702), and by Zhuhai People’s Hospital (File No. 2020XSYC-12).

Author information

Mingming Qin and Chunhao Deng contributed equally to this paper.

Authors and Affiliations

Reproductive Medical Center, Affiliated Foshan Maternity & Child Healthcare Hospital, Southern Medical University (Foshan Women and Children Hospital), Foshan, Guangdong, 528000, China

Mingming Qin & Guoqun Luo

Department of Developmental Biology, School of Basic Medical Sciences, Southern Medical University, Guangzhou, Guangdong, 510515, China

Mingming Qin

Chinese Medicine and Translational Medicine R&D center, Zhuhai UM Science & Technology Research Institute, Zhuhai, Guangdong, 519031, China

Chunhao Deng

Guangdong Provincial Key Laboratory of Tumor Interventional Diagnosis and Treatment, Zhuhai People’s Hospital, Zhuhai Clinical Medical College of Jinan University, Zhuhai, Guangdong, 519000, China

Liewei Wen & Ya Meng

You can also search for this author in PubMed   Google Scholar

Contributions

MQ and CD are responsible for the manuscript drafting and revision. LW is contributed to manuscript revision. GL is responsible for manuscript revision and financial support. YM is contributed to manuscript revision. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Guoqun Luo or Ya Meng .

Ethics declarations

Ethical approval.

Not applicable.

Consent for publication

Competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Qin, M., Deng, C., Wen, L. et al. CRISPR-Cas and CRISPR-based screening system for precise gene editing and targeted cancer therapy. J Transl Med 22 , 516 (2024). https://doi.org/10.1186/s12967-024-05235-2

Download citation

Received : 03 January 2024

Accepted : 24 April 2024

Published : 30 May 2024

DOI : https://doi.org/10.1186/s12967-024-05235-2

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• CRISPR-Cas system
• CRISPR screening
• Cancer therapy
• Precision medicine

Journal of Translational Medicine

ISSN: 1479-5876

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

A systematic literature review of empirical research on ChatGPT in education

• Open access
• Published: 26 May 2024
• Volume 3 , article number  60 , ( 2024 )

Cite this article

You have full access to this open access article

• Yazid Albadarin   ORCID: orcid.org/0009-0005-8068-8902 1 ,
• Mohammed Saqr 1 ,
• Nicolas Pope 1 &
• Markku Tukiainen 1  

104 Accesses

Explore all metrics

Over the last four decades, studies have investigated the incorporation of Artificial Intelligence (AI) into education. A recent prominent AI-powered technology that has impacted the education sector is ChatGPT. This article provides a systematic review of 14 empirical studies incorporating ChatGPT into various educational settings, published in 2022 and before the 10th of April 2023—the date of conducting the search process. It carefully followed the essential steps outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines, as well as Okoli’s (Okoli in Commun Assoc Inf Syst, 2015) steps for conducting a rigorous and transparent systematic review. In this review, we aimed to explore how students and teachers have utilized ChatGPT in various educational settings, as well as the primary findings of those studies. By employing Creswell’s (Creswell in Educational research: planning, conducting, and evaluating quantitative and qualitative research [Ebook], Pearson Education, London, 2015) coding techniques for data extraction and interpretation, we sought to gain insight into their initial attempts at ChatGPT incorporation into education. This approach also enabled us to extract insights and considerations that can facilitate its effective and responsible use in future educational contexts. The results of this review show that learners have utilized ChatGPT as a virtual intelligent assistant, where it offered instant feedback, on-demand answers, and explanations of complex topics. Additionally, learners have used it to enhance their writing and language skills by generating ideas, composing essays, summarizing, translating, paraphrasing texts, or checking grammar. Moreover, learners turned to it as an aiding tool to facilitate their directed and personalized learning by assisting in understanding concepts and homework, providing structured learning plans, and clarifying assignments and tasks. However, the results of specific studies (n = 3, 21.4%) show that overuse of ChatGPT may negatively impact innovative capacities and collaborative learning competencies among learners. Educators, on the other hand, have utilized ChatGPT to create lesson plans, generate quizzes, and provide additional resources, which helped them enhance their productivity and efficiency and promote different teaching methodologies. Despite these benefits, the majority of the reviewed studies recommend the importance of conducting structured training, support, and clear guidelines for both learners and educators to mitigate the drawbacks. This includes developing critical evaluation skills to assess the accuracy and relevance of information provided by ChatGPT, as well as strategies for integrating human interaction and collaboration into learning activities that involve AI tools. Furthermore, they also recommend ongoing research and proactive dialogue with policymakers, stakeholders, and educational practitioners to refine and enhance the use of AI in learning environments. This review could serve as an insightful resource for practitioners who seek to integrate ChatGPT into education and stimulate further research in the field.

Similar content being viewed by others

Empowering learners with ChatGPT: insights from a systematic literature exploration

Incorporating AI in foreign language education: An investigation into ChatGPT’s effect on foreign language learners

Large language models in education: A focus on the complementary relationship between human teachers and ChatGPT

Avoid common mistakes on your manuscript.

1 Introduction

Educational technology, a rapidly evolving field, plays a crucial role in reshaping the landscape of teaching and learning [ 82 ]. One of the most transformative technological innovations of our era that has influenced the field of education is Artificial Intelligence (AI) [ 50 ]. Over the last four decades, AI in education (AIEd) has gained remarkable attention for its potential to make significant advancements in learning, instructional methods, and administrative tasks within educational settings [ 11 ]. In particular, a large language model (LLM), a type of AI algorithm that applies artificial neural networks (ANNs) and uses massively large data sets to understand, summarize, generate, and predict new content that is almost difficult to differentiate from human creations [ 79 ], has opened up novel possibilities for enhancing various aspects of education, from content creation to personalized instruction [ 35 ]. Chatbots that leverage the capabilities of LLMs to understand and generate human-like responses have also presented the capacity to enhance student learning and educational outcomes by engaging students, offering timely support, and fostering interactive learning experiences [ 46 ].

The ongoing and remarkable technological advancements in chatbots have made their use more convenient, increasingly natural and effortless, and have expanded their potential for deployment across various domains [ 70 ]. One prominent example of chatbot applications is the Chat Generative Pre-Trained Transformer, known as ChatGPT, which was introduced by OpenAI, a leading AI research lab, on November 30th, 2022. ChatGPT employs a variety of deep learning techniques to generate human-like text, with a particular focus on recurrent neural networks (RNNs). Long short-term memory (LSTM) allows it to grasp the context of the text being processed and retain information from previous inputs. Also, the transformer architecture, a neural network architecture based on the self-attention mechanism, allows it to analyze specific parts of the input, thereby enabling it to produce more natural-sounding and coherent output. Additionally, the unsupervised generative pre-training and the fine-tuning methods allow ChatGPT to generate more relevant and accurate text for specific tasks [ 31 , 62 ]. Furthermore, reinforcement learning from human feedback (RLHF), a machine learning approach that combines reinforcement learning techniques with human-provided feedback, has helped improve ChatGPT’s model by accelerating the learning process and making it significantly more efficient.

This cutting-edge natural language processing (NLP) tool is widely recognized as one of today's most advanced LLMs-based chatbots [ 70 ], allowing users to ask questions and receive detailed, coherent, systematic, personalized, convincing, and informative human-like responses [ 55 ], even within complex and ambiguous contexts [ 63 , 77 ]. ChatGPT is considered the fastest-growing technology in history: in just three months following its public launch, it amassed an estimated 120 million monthly active users [ 16 ] with an estimated 13 million daily queries [ 49 ], surpassing all other applications [ 64 ]. This remarkable growth can be attributed to the unique features and user-friendly interface that ChatGPT offers. Its intuitive design allows users to interact seamlessly with the technology, making it accessible to a diverse range of individuals, regardless of their technical expertise [ 78 ]. Additionally, its exceptional performance results from a combination of advanced algorithms, continuous enhancements, and extensive training on a diverse dataset that includes various text sources such as books, articles, websites, and online forums [ 63 ], have contributed to a more engaging and satisfying user experience [ 62 ]. These factors collectively explain its remarkable global growth and set it apart from predecessors like Bard, Bing Chat, ERNIE, and others.

In this context, several studies have explored the technological advancements of chatbots. One noteworthy recent research effort, conducted by Schöbel et al. [ 70 ], stands out for its comprehensive analysis of more than 5,000 studies on communication agents. This study offered a comprehensive overview of the historical progression and future prospects of communication agents, including ChatGPT. Moreover, other studies have focused on making comparisons, particularly between ChatGPT and alternative chatbots like Bard, Bing Chat, ERNIE, LaMDA, BlenderBot, and various others. For example, O’Leary [ 53 ] compared two chatbots, LaMDA and BlenderBot, with ChatGPT and revealed that ChatGPT outperformed both. This superiority arises from ChatGPT’s capacity to handle a wider range of questions and generate slightly varied perspectives within specific contexts. Similarly, ChatGPT exhibited an impressive ability to formulate interpretable responses that were easily understood when compared with Google's feature snippet [ 34 ]. Additionally, ChatGPT was compared to other LLMs-based chatbots, including Bard and BERT, as well as ERNIE. The findings indicated that ChatGPT exhibited strong performance in the given tasks, often outperforming the other models [ 59 ].

Furthermore, in the education context, a comprehensive study systematically compared a range of the most promising chatbots, including Bard, Bing Chat, ChatGPT, and Ernie across a multidisciplinary test that required higher-order thinking. The study revealed that ChatGPT achieved the highest score, surpassing Bing Chat and Bard [ 64 ]. Similarly, a comparative analysis was conducted to compare ChatGPT with Bard in answering a set of 30 mathematical questions and logic problems, grouped into two question sets. Set (A) is unavailable online, while Set (B) is available online. The results revealed ChatGPT's superiority in Set (A) over Bard. Nevertheless, Bard's advantage emerged in Set (B) due to its capacity to access the internet directly and retrieve answers, a capability that ChatGPT does not possess [ 57 ]. However, through these varied assessments, ChatGPT consistently highlights its exceptional prowess compared to various alternatives in the ever-evolving chatbot technology.

The widespread adoption of chatbots, especially ChatGPT, by millions of students and educators, has sparked extensive discussions regarding its incorporation into the education sector [ 64 ]. Accordingly, many scholars have contributed to the discourse, expressing both optimism and pessimism regarding the incorporation of ChatGPT into education. For example, ChatGPT has been highlighted for its capabilities in enriching the learning and teaching experience through its ability to support different learning approaches, including adaptive learning, personalized learning, and self-directed learning [ 58 , 60 , 91 ]), deliver summative and formative feedback to students and provide real-time responses to questions, increase the accessibility of information [ 22 , 40 , 43 ], foster students’ performance, engagement and motivation [ 14 , 44 , 58 ], and enhance teaching practices [ 17 , 18 , 64 , 74 ].

On the other hand, concerns have been also raised regarding its potential negative effects on learning and teaching. These include the dissemination of false information and references [ 12 , 23 , 61 , 85 ], biased reinforcement [ 47 , 50 ], compromised academic integrity [ 18 , 40 , 66 , 74 ], and the potential decline in students' skills [ 43 , 61 , 64 , 74 ]. As a result, ChatGPT has been banned in multiple countries, including Russia, China, Venezuela, Belarus, and Iran, as well as in various educational institutions in India, Italy, Western Australia, France, and the United States [ 52 , 90 ].

Clearly, the advent of chatbots, especially ChatGPT, has provoked significant controversy due to their potential impact on learning and teaching. This indicates the necessity for further exploration to gain a deeper understanding of this technology and carefully evaluate its potential benefits, limitations, challenges, and threats to education [ 79 ]. Therefore, conducting a systematic literature review will provide valuable insights into the potential prospects and obstacles linked to its incorporation into education. This systematic literature review will primarily focus on ChatGPT, driven by the aforementioned key factors outlined above.

However, the existing literature lacks a systematic literature review of empirical studies. Thus, this systematic literature review aims to address this gap by synthesizing the existing empirical studies conducted on chatbots, particularly ChatGPT, in the field of education, highlighting how ChatGPT has been utilized in educational settings, and identifying any existing gaps. This review may be particularly useful for researchers in the field and educators who are contemplating the integration of ChatGPT or any chatbot into education. The following research questions will guide this study:

What are students' and teachers' initial attempts at utilizing ChatGPT in education?

What are the main findings derived from empirical studies that have incorporated ChatGPT into learning and teaching?

2 Methodology

To conduct this study, the authors followed the essential steps of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) and Okoli’s [ 54 ] steps for conducting a systematic review. These included identifying the study’s purpose, drafting a protocol, applying a practical screening process, searching the literature, extracting relevant data, evaluating the quality of the included studies, synthesizing the studies, and ultimately writing the review. The subsequent section provides an extensive explanation of how these steps were carried out in this study.

2.1 Identify the purpose

Given the widespread adoption of ChatGPT by students and teachers for various educational purposes, often without a thorough understanding of responsible and effective use or a clear recognition of its potential impact on learning and teaching, the authors recognized the need for further exploration of ChatGPT's impact on education in this early stage. Therefore, they have chosen to conduct a systematic literature review of existing empirical studies that incorporate ChatGPT into educational settings. Despite the limited number of empirical studies due to the novelty of the topic, their goal is to gain a deeper understanding of this technology and proactively evaluate its potential benefits, limitations, challenges, and threats to education. This effort could help to understand initial reactions and attempts at incorporating ChatGPT into education and bring out insights and considerations that can inform the future development of education.

2.2 Draft the protocol

The next step is formulating the protocol. This protocol serves to outline the study process in a rigorous and transparent manner, mitigating researcher bias in study selection and data extraction [ 88 ]. The protocol will include the following steps: generating the research question, predefining a literature search strategy, identifying search locations, establishing selection criteria, assessing the studies, developing a data extraction strategy, and creating a timeline.

2.3 Apply practical screen

The screening step aims to accurately filter the articles resulting from the searching step and select the empirical studies that have incorporated ChatGPT into educational contexts, which will guide us in answering the research questions and achieving the objectives of this study. To ensure the rigorous execution of this step, our inclusion and exclusion criteria were determined based on the authors' experience and informed by previous successful systematic reviews [ 21 ]. Table 1 summarizes the inclusion and exclusion criteria for study selection.

2.4 Literature search

We conducted a thorough literature search to identify articles that explored, examined, and addressed the use of ChatGPT in Educational contexts. We utilized two research databases: Dimensions.ai, which provides access to a large number of research publications, and lens.org, which offers access to over 300 million articles, patents, and other research outputs from diverse sources. Additionally, we included three databases, Scopus, Web of Knowledge, and ERIC, which contain relevant research on the topic that addresses our research questions. To browse and identify relevant articles, we used the following search formula: ("ChatGPT" AND "Education"), which included the Boolean operator "AND" to get more specific results. The subject area in the Scopus and ERIC databases were narrowed to "ChatGPT" and "Education" keywords, and in the WoS database was limited to the "Education" category. The search was conducted between the 3rd and 10th of April 2023, which resulted in 276 articles from all selected databases (111 articles from Dimensions.ai, 65 from Scopus, 28 from Web of Science, 14 from ERIC, and 58 from Lens.org). These articles were imported into the Rayyan web-based system for analysis. The duplicates were identified automatically by the system. Subsequently, the first author manually reviewed the duplicated articles ensured that they had the same content, and then removed them, leaving us with 135 unique articles. Afterward, the titles, abstracts, and keywords of the first 40 manuscripts were scanned and reviewed by the first author and were discussed with the second and third authors to resolve any disagreements. Subsequently, the first author proceeded with the filtering process for all articles and carefully applied the inclusion and exclusion criteria as presented in Table  1 . Articles that met any one of the exclusion criteria were eliminated, resulting in 26 articles. Afterward, the authors met to carefully scan and discuss them. The authors agreed to eliminate any empirical studies solely focused on checking ChatGPT capabilities, as these studies do not guide us in addressing the research questions and achieving the study's objectives. This resulted in 14 articles eligible for analysis.

2.5 Quality appraisal

The examination and evaluation of the quality of the extracted articles is a vital step [ 9 ]. Therefore, the extracted articles were carefully evaluated for quality using Fink’s [ 24 ] standards, which emphasize the necessity for detailed descriptions of methodology, results, conclusions, strengths, and limitations. The process began with a thorough assessment of each study's design, data collection, and analysis methods to ensure their appropriateness and comprehensive execution. The clarity, consistency, and logical progression from data to results and conclusions were also critically examined. Potential biases and recognized limitations within the studies were also scrutinized. Ultimately, two articles were excluded for failing to meet Fink’s criteria, particularly in providing sufficient detail on methodology, results, conclusions, strengths, or limitations. The review process is illustrated in Fig.  1 .

The study selection process

2.6 Data extraction

The next step is data extraction, the process of capturing the key information and categories from the included studies. To improve efficiency, reduce variation among authors, and minimize errors in data analysis, the coding categories were constructed using Creswell's [ 15 ] coding techniques for data extraction and interpretation. The coding process involves three sequential steps. The initial stage encompasses open coding , where the researcher examines the data, generates codes to describe and categorize it, and gains a deeper understanding without preconceived ideas. Following open coding is axial coding , where the interrelationships between codes from open coding are analyzed to establish more comprehensive categories or themes. The process concludes with selective coding , refining and integrating categories or themes to identify core concepts emerging from the data. The first coder performed the coding process, then engaged in discussions with the second and third authors to finalize the coding categories for the first five articles. The first coder then proceeded to code all studies and engaged again in discussions with the other authors to ensure the finalization of the coding process. After a comprehensive analysis and capturing of the key information from the included studies, the data extraction and interpretation process yielded several themes. These themes have been categorized and are presented in Table  2 . It is important to note that open coding results were removed from Table  2 for aesthetic reasons, as it included many generic aspects, such as words, short phrases, or sentences mentioned in the studies.

2.7 Synthesize studies

In this stage, we will gather, discuss, and analyze the key findings that emerged from the selected studies. The synthesis stage is considered a transition from an author-centric to a concept-centric focus, enabling us to map all the provided information to achieve the most effective evaluation of the data [ 87 ]. Initially, the authors extracted data that included general information about the selected studies, including the author(s)' names, study titles, years of publication, educational levels, research methodologies, sample sizes, participants, main aims or objectives, raw data sources, and analysis methods. Following that, all key information and significant results from the selected studies were compiled using Creswell’s [ 15 ] coding techniques for data extraction and interpretation to identify core concepts and themes emerging from the data, focusing on those that directly contributed to our research questions and objectives, such as the initial utilization of ChatGPT in learning and teaching, learners' and educators' familiarity with ChatGPT, and the main findings of each study. Finally, the data related to each selected study were extracted into an Excel spreadsheet for data processing. The Excel spreadsheet was reviewed by the authors, including a series of discussions to ensure the finalization of this process and prepare it for further analysis. Afterward, the final result being analyzed and presented in various types of charts and graphs. Table 4 presents the extracted data from the selected studies, with each study labeled with a capital 'S' followed by a number.

This section consists of two main parts. The first part provides a descriptive analysis of the data compiled from the reviewed studies. The second part presents the answers to the research questions and the main findings of these studies.

3.1 Part 1: descriptive analysis

This section will provide a descriptive analysis of the reviewed studies, including educational levels and fields, participants distribution, country contribution, research methodologies, study sample size, study population, publication year, list of journals, familiarity with ChatGPT, source of data, and the main aims and objectives of the studies. Table 4 presents a comprehensive overview of the extracted data from the selected studies.

3.1.1 The number of the reviewed studies and publication years

The total number of the reviewed studies was 14. All studies were empirical studies and published in different journals focusing on Education and Technology. One study was published in 2022 [S1], while the remaining were published in 2023 [S2]-[S14]. Table 3 illustrates the year of publication, the names of the journals, and the number of reviewed studies published in each journal for the studies reviewed.

3.1.2 Educational levels and fields

The majority of the reviewed studies, 11 studies, were conducted in higher education institutions [S1]-[S10] and [S13]. Two studies did not specify the educational level of the population [S12] and [S14], while one study focused on elementary education [S11]. However, the reviewed studies covered various fields of education. Three studies focused on Arts and Humanities Education [S8], [S11], and [S14], specifically English Education. Two studies focused on Engineering Education, with one in Computer Engineering [S2] and the other in Construction Education [S3]. Two studies focused on Mathematics Education [S5] and [S12]. One study focused on Social Science Education [S13]. One study focused on Early Education [S4]. One study focused on Journalism Education [S9]. Finally, three studies did not specify the field of education [S1], [S6], and [S7]. Figure  2 represents the educational levels in the reviewed studies, while Fig.  3 represents the context of the reviewed studies.

Educational levels in the reviewed studies

Context of the reviewed studies

3.1.3 Participants distribution and countries contribution

The reviewed studies have been conducted across different geographic regions, providing a diverse representation of the studies. The majority of the studies, 10 in total, [S1]-[S3], [S5]-[S9], [S11], and [S14], primarily focused on participants from single countries such as Pakistan, the United Arab Emirates, China, Indonesia, Poland, Saudi Arabia, South Korea, Spain, Tajikistan, and the United States. In contrast, four studies, [S4], [S10], [S12], and [S13], involved participants from multiple countries, including China and the United States [S4], China, the United Kingdom, and the United States [S10], the United Arab Emirates, Oman, Saudi Arabia, and Jordan [S12], Turkey, Sweden, Canada, and Australia [ 13 ]. Figures  4 and 5 illustrate the distribution of participants, whether from single or multiple countries, and the contribution of each country in the reviewed studies, respectively.

The reviewed studies conducted in single or multiple countries

The Contribution of each country in the studies

3.1.4 Study population and sample size

Four study populations were included: university students, university teachers, university teachers and students, and elementary school teachers. Six studies involved university students [S2], [S3], [S5] and [S6]-[S8]. Three studies focused on university teachers [S1], [S4], and [S6], while one study specifically targeted elementary school teachers [S11]. Additionally, four studies included both university teachers and students [S10] and [ 12 , 13 , 14 ], and among them, study [S13] specifically included postgraduate students. In terms of the sample size of the reviewed studies, nine studies included a small sample size of less than 50 participants [S1], [S3], [S6], [S8], and [S10]-[S13]. Three studies had 50–100 participants [S2], [S9], and [S14]. Only one study had more than 100 participants [S7]. It is worth mentioning that study [S4] adopted a mixed methods approach, including 10 participants for qualitative analysis and 110 participants for quantitative analysis.

3.1.5 Participants’ familiarity with using ChatGPT

The reviewed studies recruited a diverse range of participants with varying levels of familiarity with ChatGPT. Five studies [S2], [S4], [S6], [S8], and [S12] involved participants already familiar with ChatGPT, while eight studies [S1], [S3], [S5], [S7], [S9], [S10], [S13] and [S14] included individuals with differing levels of familiarity. Notably, one study [S11] had participants who were entirely unfamiliar with ChatGPT. It is important to note that four studies [S3], [S5], [S9], and [S11] provided training or guidance to their participants before conducting their studies, while ten studies [S1], [S2], [S4], [S6]-[S8], [S10], and [S12]-[S14] did not provide training due to the participants' existing familiarity with ChatGPT.

3.1.6 Research methodology approaches and source(S) of data

The reviewed studies adopted various research methodology approaches. Seven studies adopted qualitative research methodology [S1], [S4], [S6], [S8], [S10], [S11], and [S12], while three studies adopted quantitative research methodology [S3], [S7], and [S14], and four studies employed mixed-methods, which involved a combination of both the strengths of qualitative and quantitative methods [S2], [S5], [S9], and [S13].

In terms of the source(s) of data, the reviewed studies obtained their data from various sources, such as interviews, questionnaires, and pre-and post-tests. Six studies relied on interviews as their primary source of data collection [S1], [S4], [S6], [S10], [S11], and [S12], four studies relied on questionnaires [S2], [S7], [S13], and [S14], two studies combined the use of pre-and post-tests and questionnaires for data collection [S3] and [S9], while two studies combined the use of questionnaires and interviews to obtain the data [S5] and [S8]. It is important to note that six of the reviewed studies were quasi-experimental [S3], [S5], [S8], [S9], [S12], and [S14], while the remaining ones were experimental studies [S1], [S2], [S4], [S6], [S7], [S10], [S11], and [S13]. Figures  6 and 7 illustrate the research methodologies and the source (s) of data used in the reviewed studies, respectively.

Research methodologies in the reviewed studies

Source of data in the reviewed studies

3.1.7 The aim and objectives of the studies

The reviewed studies encompassed a diverse set of aims, with several of them incorporating multiple primary objectives. Six studies [S3], [S6], [S7], [S8], [S11], and [S12] examined the integration of ChatGPT in educational contexts, and four studies [S4], [S5], [S13], and [S14] investigated the various implications of its use in education, while three studies [S2], [S9], and [S10] aimed to explore both its integration and implications in education. Additionally, seven studies explicitly explored attitudes and perceptions of students [S2] and [S3], educators [S1] and [S6], or both [S10], [S12], and [S13] regarding the utilization of ChatGPT in educational settings.

3.2 Part 2: research questions and main findings of the reviewed studies

This part will present the answers to the research questions and the main findings of the reviewed studies, classified into two main categories (learning and teaching) according to AI Education classification by [ 36 ]. Figure  8 summarizes the main findings of the reviewed studies in a visually informative diagram. Table 4 provides a detailed list of the key information extracted from the selected studies that led to generating these themes.

The main findings in the reviewed studies

4 Students' initial attempts at utilizing ChatGPT in learning and main findings from students' perspective

4.1 virtual intelligent assistant.

Nine studies demonstrated that ChatGPT has been utilized by students as an intelligent assistant to enhance and support their learning. Students employed it for various purposes, such as answering on-demand questions [S2]-[S5], [S8], [S10], and [S12], providing valuable information and learning resources [S2]-[S5], [S6], and [S8], as well as receiving immediate feedback [S2], [S4], [S9], [S10], and [S12]. In this regard, students generally were confident in the accuracy of ChatGPT's responses, considering them relevant, reliable, and detailed [S3], [S4], [S5], and [S8]. However, some students indicated the need for improvement, as they found that answers are not always accurate [S2], and that misleading information may have been provided or that it may not always align with their expectations [S6] and [S10]. It was also observed by the students that the accuracy of ChatGPT is dependent on several factors, including the quality and specificity of the user's input, the complexity of the question or topic, and the scope and relevance of its training data [S12]. Many students felt that ChatGPT's answers were not always accurate and most of them believed that it requires good background knowledge to work with.

4.2 Writing and language proficiency assistant

Six of the reviewed studies highlighted that ChatGPT has been utilized by students as a valuable assistant tool to improve their academic writing skills and language proficiency. Among these studies, three mainly focused on English education, demonstrating that students showed sufficient mastery in using ChatGPT for generating ideas, summarizing, paraphrasing texts, and completing writing essays [S8], [S11], and [S14]. Furthermore, ChatGPT helped them in writing by making students active investigators rather than passive knowledge recipients and facilitated the development of their writing skills [S11] and [S14]. Similarly, ChatGPT allowed students to generate unique ideas and perspectives, leading to deeper analysis and reflection on their journalism writing [S9]. In terms of language proficiency, ChatGPT allowed participants to translate content into their home languages, making it more accessible and relevant to their context [S4]. It also enabled them to request changes in linguistic tones or flavors [S8]. Moreover, participants used it to check grammar or as a dictionary [S11].

4.3 Valuable resource for learning approaches

Five studies demonstrated that students used ChatGPT as a valuable complementary resource for self-directed learning. It provided learning resources and guidance on diverse educational topics and created a supportive home learning environment [S2] and [S4]. Moreover, it offered step-by-step guidance to grasp concepts at their own pace and enhance their understanding [S5], streamlined task and project completion carried out independently [S7], provided comprehensive and easy-to-understand explanations on various subjects [S10], and assisted in studying geometry operations, thereby empowering them to explore geometry operations at their own pace [S12]. Three studies showed that students used ChatGPT as a valuable learning resource for personalized learning. It delivered age-appropriate conversations and tailored teaching based on a child's interests [S4], acted as a personalized learning assistant, adapted to their needs and pace, which assisted them in understanding mathematical concepts [S12], and enabled personalized learning experiences in social sciences by adapting to students' needs and learning styles [S13]. On the other hand, it is important to note that, according to one study [S5], students suggested that using ChatGPT may negatively affect collaborative learning competencies between students.

4.4 Enhancing students' competencies

Six of the reviewed studies have shown that ChatGPT is a valuable tool for improving a wide range of skills among students. Two studies have provided evidence that ChatGPT led to improvements in students' critical thinking, reasoning skills, and hazard recognition competencies through engaging them in interactive conversations or activities and providing responses related to their disciplines in journalism [S5] and construction education [S9]. Furthermore, two studies focused on mathematical education have shown the positive impact of ChatGPT on students' problem-solving abilities in unraveling problem-solving questions [S12] and enhancing the students' understanding of the problem-solving process [S5]. Lastly, one study indicated that ChatGPT effectively contributed to the enhancement of conversational social skills [S4].

4.5 Supporting students' academic success

Seven of the reviewed studies highlighted that students found ChatGPT to be beneficial for learning as it enhanced learning efficiency and improved the learning experience. It has been observed to improve students' efficiency in computer engineering studies by providing well-structured responses and good explanations [S2]. Additionally, students found it extremely useful for hazard reporting [S3], and it also enhanced their efficiency in solving mathematics problems and capabilities [S5] and [S12]. Furthermore, by finding information, generating ideas, translating texts, and providing alternative questions, ChatGPT aided students in deepening their understanding of various subjects [S6]. It contributed to an increase in students' overall productivity [S7] and improved efficiency in composing written tasks [S8]. Regarding learning experiences, ChatGPT was instrumental in assisting students in identifying hazards that they might have otherwise overlooked [S3]. It also improved students' learning experiences in solving mathematics problems and developing abilities [S5] and [S12]. Moreover, it increased students' successful completion of important tasks in their studies [S7], particularly those involving average difficulty writing tasks [S8]. Additionally, ChatGPT increased the chances of educational success by providing students with baseline knowledge on various topics [S10].

5 Teachers' initial attempts at utilizing ChatGPT in teaching and main findings from teachers' perspective

5.1 valuable resource for teaching.

The reviewed studies showed that teachers have employed ChatGPT to recommend, modify, and generate diverse, creative, organized, and engaging educational contents, teaching materials, and testing resources more rapidly [S4], [S6], [S10] and [S11]. Additionally, teachers experienced increased productivity as ChatGPT facilitated quick and accurate responses to questions, fact-checking, and information searches [S1]. It also proved valuable in constructing new knowledge [S6] and providing timely answers to students' questions in classrooms [S11]. Moreover, ChatGPT enhanced teachers' efficiency by generating new ideas for activities and preplanning activities for their students [S4] and [S6], including interactive language game partners [S11].

5.2 Improving productivity and efficiency

The reviewed studies showed that participants' productivity and work efficiency have been significantly enhanced by using ChatGPT as it enabled them to allocate more time to other tasks and reduce their overall workloads [S6], [S10], [S11], [S13], and [S14]. However, three studies [S1], [S4], and [S11], indicated a negative perception and attitude among teachers toward using ChatGPT. This negativity stemmed from a lack of necessary skills to use it effectively [S1], a limited familiarity with it [S4], and occasional inaccuracies in the content provided by it [S10].

5.3 Catalyzing new teaching methodologies

Five of the reviewed studies highlighted that educators found the necessity of redefining their teaching profession with the assistance of ChatGPT [S11], developing new effective learning strategies [S4], and adapting teaching strategies and methodologies to ensure the development of essential skills for future engineers [S5]. They also emphasized the importance of adopting new educational philosophies and approaches that can evolve with the introduction of ChatGPT into the classroom [S12]. Furthermore, updating curricula to focus on improving human-specific features, such as emotional intelligence, creativity, and philosophical perspectives [S13], was found to be essential.

5.4 Effective utilization of CHATGPT in teaching

According to the reviewed studies, effective utilization of ChatGPT in education requires providing teachers with well-structured training, support, and adequate background on how to use ChatGPT responsibly [S1], [S3], [S11], and [S12]. Establishing clear rules and regulations regarding its usage is essential to ensure it positively impacts the teaching and learning processes, including students' skills [S1], [S4], [S5], [S8], [S9], and [S11]-[S14]. Moreover, conducting further research and engaging in discussions with policymakers and stakeholders is indeed crucial for the successful integration of ChatGPT in education and to maximize the benefits for both educators and students [S1], [S6]-[S10], and [S12]-[S14].

6 Discussion

The purpose of this review is to conduct a systematic review of empirical studies that have explored the utilization of ChatGPT, one of today’s most advanced LLM-based chatbots, in education. The findings of the reviewed studies showed several ways of ChatGPT utilization in different learning and teaching practices as well as it provided insights and considerations that can facilitate its effective and responsible use in future educational contexts. The results of the reviewed studies came from diverse fields of education, which helped us avoid a biased review that is limited to a specific field. Similarly, the reviewed studies have been conducted across different geographic regions. This kind of variety in geographic representation enriched the findings of this review.

In response to RQ1 , "What are students' and teachers' initial attempts at utilizing ChatGPT in education?", the findings from this review provide comprehensive insights. Chatbots, including ChatGPT, play a crucial role in supporting student learning, enhancing their learning experiences, and facilitating diverse learning approaches [ 42 , 43 ]. This review found that this tool, ChatGPT, has been instrumental in enhancing students' learning experiences by serving as a virtual intelligent assistant, providing immediate feedback, on-demand answers, and engaging in educational conversations. Additionally, students have benefited from ChatGPT’s ability to generate ideas, compose essays, and perform tasks like summarizing, translating, paraphrasing texts, or checking grammar, thereby enhancing their writing and language competencies. Furthermore, students have turned to ChatGPT for assistance in understanding concepts and homework, providing structured learning plans, and clarifying assignments and tasks, which fosters a supportive home learning environment, allowing them to take responsibility for their own learning and cultivate the skills and approaches essential for supportive home learning environment [ 26 , 27 , 28 ]. This finding aligns with the study of Saqr et al. [ 68 , 69 ] who highlighted that, when students actively engage in their own learning process, it yields additional advantages, such as heightened motivation, enhanced achievement, and the cultivation of enthusiasm, turning them into advocates for their own learning.

Moreover, students have utilized ChatGPT for tailored teaching and step-by-step guidance on diverse educational topics, streamlining task and project completion, and generating and recommending educational content. This personalization enhances the learning environment, leading to increased academic success. This finding aligns with other recent studies [ 26 , 27 , 28 , 60 , 66 ] which revealed that ChatGPT has the potential to offer personalized learning experiences and support an effective learning process by providing students with customized feedback and explanations tailored to their needs and abilities. Ultimately, fostering students' performance, engagement, and motivation, leading to increase students' academic success [ 14 , 44 , 58 ]. This ultimate outcome is in line with the findings of Saqr et al. [ 68 , 69 ], which emphasized that learning strategies are important catalysts of students' learning, as students who utilize effective learning strategies are more likely to have better academic achievement.

Teachers, too, have capitalized on ChatGPT's capabilities to enhance productivity and efficiency, using it for creating lesson plans, generating quizzes, providing additional resources, generating and preplanning new ideas for activities, and aiding in answering students’ questions. This adoption of technology introduces new opportunities to support teaching and learning practices, enhancing teacher productivity. This finding aligns with those of Day [ 17 ], De Castro [ 18 ], and Su and Yang [ 74 ] as well as with those of Valtonen et al. [ 82 ], who revealed that emerging technological advancements have opened up novel opportunities and means to support teaching and learning practices, and enhance teachers’ productivity.

In response to RQ2 , "What are the main findings derived from empirical studies that have incorporated ChatGPT into learning and teaching?", the findings from this review provide profound insights and raise significant concerns. Starting with the insights, chatbots, including ChatGPT, have demonstrated the potential to reshape and revolutionize education, creating new, novel opportunities for enhancing the learning process and outcomes [ 83 ], facilitating different learning approaches, and offering a range of pedagogical benefits [ 19 , 43 , 72 ]. In this context, this review found that ChatGPT could open avenues for educators to adopt or develop new effective learning and teaching strategies that can evolve with the introduction of ChatGPT into the classroom. Nonetheless, there is an evident lack of research understanding regarding the potential impact of generative machine learning models within diverse educational settings [ 83 ]. This necessitates teachers to attain a high level of proficiency in incorporating chatbots, such as ChatGPT, into their classrooms to create inventive, well-structured, and captivating learning strategies. In the same vein, the review also found that teachers without the requisite skills to utilize ChatGPT realized that it did not contribute positively to their work and could potentially have adverse effects [ 37 ]. This concern could lead to inequity of access to the benefits of chatbots, including ChatGPT, as individuals who lack the necessary expertise may not be able to harness their full potential, resulting in disparities in educational outcomes and opportunities. Therefore, immediate action is needed to address these potential issues. A potential solution is offering training, support, and competency development for teachers to ensure that all of them can leverage chatbots, including ChatGPT, effectively and equitably in their educational practices [ 5 , 28 , 80 ], which could enhance accessibility and inclusivity, and potentially result in innovative outcomes [ 82 , 83 ].

Additionally, chatbots, including ChatGPT, have the potential to significantly impact students' thinking abilities, including retention, reasoning, analysis skills [ 19 , 45 ], and foster innovation and creativity capabilities [ 83 ]. This review found that ChatGPT could contribute to improving a wide range of skills among students. However, it found that frequent use of ChatGPT may result in a decrease in innovative capacities, collaborative skills and cognitive capacities, and students' motivation to attend classes, as well as could lead to reduced higher-order thinking skills among students [ 22 , 29 ]. Therefore, immediate action is needed to carefully examine the long-term impact of chatbots such as ChatGPT, on learning outcomes as well as to explore its incorporation into educational settings as a supportive tool without compromising students' cognitive development and critical thinking abilities. In the same vein, the review also found that it is challenging to draw a consistent conclusion regarding the potential of ChatGPT to aid self-directed learning approach. This finding aligns with the recent study of Baskara [ 8 ]. Therefore, further research is needed to explore the potential of ChatGPT for self-directed learning. One potential solution involves utilizing learning analytics as a novel approach to examine various aspects of students' learning and support them in their individual endeavors [ 32 ]. This approach can bridge this gap by facilitating an in-depth analysis of how learners engage with ChatGPT, identifying trends in self-directed learning behavior, and assessing its influence on their outcomes.

Turning to the significant concerns, on the other hand, a fundamental challenge with LLM-based chatbots, including ChatGPT, is the accuracy and quality of the provided information and responses, as they provide false information as truth—a phenomenon often referred to as "hallucination" [ 3 , 49 ]. In this context, this review found that the provided information was not entirely satisfactory. Consequently, the utilization of chatbots presents potential concerns, such as generating and providing inaccurate or misleading information, especially for students who utilize it to support their learning. This finding aligns with other findings [ 6 , 30 , 35 , 40 ] which revealed that incorporating chatbots such as ChatGPT, into education presents challenges related to its accuracy and reliability due to its training on a large corpus of data, which may contain inaccuracies and the way users formulate or ask ChatGPT. Therefore, immediate action is needed to address these potential issues. One possible solution is to equip students with the necessary skills and competencies, which include a background understanding of how to use it effectively and the ability to assess and evaluate the information it generates, as the accuracy and the quality of the provided information depend on the input, its complexity, the topic, and the relevance of its training data [ 28 , 49 , 86 ]. However, it's also essential to examine how learners can be educated about how these models operate, the data used in their training, and how to recognize their limitations, challenges, and issues [ 79 ].

Furthermore, chatbots present a substantial challenge concerning maintaining academic integrity [ 20 , 56 ] and copyright violations [ 83 ], which are significant concerns in education. The review found that the potential misuse of ChatGPT might foster cheating, facilitate plagiarism, and threaten academic integrity. This issue is also affirmed by the research conducted by Basic et al. [ 7 ], who presented evidence that students who utilized ChatGPT in their writing assignments had more plagiarism cases than those who did not. These findings align with the conclusions drawn by Cotton et al. [ 13 ], Hisan and Amri [ 33 ] and Sullivan et al. [ 75 ], who revealed that the integration of chatbots such as ChatGPT into education poses a significant challenge to the preservation of academic integrity. Moreover, chatbots, including ChatGPT, have increased the difficulty in identifying plagiarism [ 47 , 67 , 76 ]. The findings from previous studies [ 1 , 84 ] indicate that AI-generated text often went undetected by plagiarism software, such as Turnitin. However, Turnitin and other similar plagiarism detection tools, such as ZeroGPT, GPTZero, and Copyleaks, have since evolved, incorporating enhanced techniques to detect AI-generated text, despite the possibility of false positives, as noted in different studies that have found these tools still not yet fully ready to accurately and reliably identify AI-generated text [ 10 , 51 ], and new novel detection methods may need to be created and implemented for AI-generated text detection [ 4 ]. This potential issue could lead to another concern, which is the difficulty of accurately evaluating student performance when they utilize chatbots such as ChatGPT assistance in their assignments. Consequently, the most LLM-driven chatbots present a substantial challenge to traditional assessments [ 64 ]. The findings from previous studies indicate the importance of rethinking, improving, and redesigning innovative assessment methods in the era of chatbots [ 14 , 20 , 64 , 75 ]. These methods should prioritize the process of evaluating students' ability to apply knowledge to complex cases and demonstrate comprehension, rather than solely focusing on the final product for assessment. Therefore, immediate action is needed to address these potential issues. One possible solution would be the development of clear guidelines, regulatory policies, and pedagogical guidance. These measures would help regulate the proper and ethical utilization of chatbots, such as ChatGPT, and must be established before their introduction to students [ 35 , 38 , 39 , 41 , 89 ].

In summary, our review has delved into the utilization of ChatGPT, a prominent example of chatbots, in education, addressing the question of how ChatGPT has been utilized in education. However, there remain significant gaps, which necessitate further research to shed light on this area.

7 Conclusions

This systematic review has shed light on the varied initial attempts at incorporating ChatGPT into education by both learners and educators, while also offering insights and considerations that can facilitate its effective and responsible use in future educational contexts. From the analysis of 14 selected studies, the review revealed the dual-edged impact of ChatGPT in educational settings. On the positive side, ChatGPT significantly aided the learning process in various ways. Learners have used it as a virtual intelligent assistant, benefiting from its ability to provide immediate feedback, on-demand answers, and easy access to educational resources. Additionally, it was clear that learners have used it to enhance their writing and language skills, engaging in practices such as generating ideas, composing essays, and performing tasks like summarizing, translating, paraphrasing texts, or checking grammar. Importantly, other learners have utilized it in supporting and facilitating their directed and personalized learning on a broad range of educational topics, assisting in understanding concepts and homework, providing structured learning plans, and clarifying assignments and tasks. Educators, on the other hand, found ChatGPT beneficial for enhancing productivity and efficiency. They used it for creating lesson plans, generating quizzes, providing additional resources, and answers learners' questions, which saved time and allowed for more dynamic and engaging teaching strategies and methodologies.

However, the review also pointed out negative impacts. The results revealed that overuse of ChatGPT could decrease innovative capacities and collaborative learning among learners. Specifically, relying too much on ChatGPT for quick answers can inhibit learners' critical thinking and problem-solving skills. Learners might not engage deeply with the material or consider multiple solutions to a problem. This tendency was particularly evident in group projects, where learners preferred consulting ChatGPT individually for solutions over brainstorming and collaborating with peers, which negatively affected their teamwork abilities. On a broader level, integrating ChatGPT into education has also raised several concerns, including the potential for providing inaccurate or misleading information, issues of inequity in access, challenges related to academic integrity, and the possibility of misusing the technology.

Accordingly, this review emphasizes the urgency of developing clear rules, policies, and regulations to ensure ChatGPT's effective and responsible use in educational settings, alongside other chatbots, by both learners and educators. This requires providing well-structured training to educate them on responsible usage and understanding its limitations, along with offering sufficient background information. Moreover, it highlights the importance of rethinking, improving, and redesigning innovative teaching and assessment methods in the era of ChatGPT. Furthermore, conducting further research and engaging in discussions with policymakers and stakeholders are essential steps to maximize the benefits for both educators and learners and ensure academic integrity.

It is important to acknowledge that this review has certain limitations. Firstly, the limited inclusion of reviewed studies can be attributed to several reasons, including the novelty of the technology, as new technologies often face initial skepticism and cautious adoption; the lack of clear guidelines or best practices for leveraging this technology for educational purposes; and institutional or governmental policies affecting the utilization of this technology in educational contexts. These factors, in turn, have affected the number of studies available for review. Secondly, the utilization of the original version of ChatGPT, based on GPT-3 or GPT-3.5, implies that new studies utilizing the updated version, GPT-4 may lead to different findings. Therefore, conducting follow-up systematic reviews is essential once more empirical studies on ChatGPT are published. Additionally, long-term studies are necessary to thoroughly examine and assess the impact of ChatGPT on various educational practices.

Despite these limitations, this systematic review has highlighted the transformative potential of ChatGPT in education, revealing its diverse utilization by learners and educators alike and summarized the benefits of incorporating it into education, as well as the forefront critical concerns and challenges that must be addressed to facilitate its effective and responsible use in future educational contexts. This review could serve as an insightful resource for practitioners who seek to integrate ChatGPT into education and stimulate further research in the field.

Data availability

The data supporting our findings are available upon request.

Abbreviations

• Artificial intelligence

AI in education

Large language model

Artificial neural networks

Chat Generative Pre-Trained Transformer

Recurrent neural networks

Long short-term memory

Reinforcement learning from human feedback

Natural language processing

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

AlAfnan MA, Dishari S, Jovic M, Lomidze K. ChatGPT as an educational tool: opportunities, challenges, and recommendations for communication, business writing, and composition courses. J Artif Intell Technol. 2023. https://doi.org/10.37965/jait.2023.0184 .

Article   Google Scholar  

Ali JKM, Shamsan MAA, Hezam TA, Mohammed AAQ. Impact of ChatGPT on learning motivation. J Engl Stud Arabia Felix. 2023;2(1):41–9. https://doi.org/10.56540/jesaf.v2i1.51 .

Alkaissi H, McFarlane SI. Artificial hallucinations in ChatGPT: implications in scientific writing. Cureus. 2023. https://doi.org/10.7759/cureus.35179 .

Anderson N, Belavý DL, Perle SM, Hendricks S, Hespanhol L, Verhagen E, Memon AR. AI did not write this manuscript, or did it? Can we trick the AI text detector into generated texts? The potential future of ChatGPT and AI in sports & exercise medicine manuscript generation. BMJ Open Sport Exerc Med. 2023;9(1): e001568. https://doi.org/10.1136/bmjsem-2023-001568 .

Ausat AMA, Massang B, Efendi M, Nofirman N, Riady Y. Can chat GPT replace the role of the teacher in the classroom: a fundamental analysis. J Educ. 2023;5(4):16100–6.

Google Scholar  

Baidoo-Anu D, Ansah L. Education in the Era of generative artificial intelligence (AI): understanding the potential benefits of ChatGPT in promoting teaching and learning. Soc Sci Res Netw. 2023. https://doi.org/10.2139/ssrn.4337484 .

Basic Z, Banovac A, Kruzic I, Jerkovic I. Better by you, better than me, chatgpt3 as writing assistance in students essays. 2023. arXiv preprint arXiv:2302.04536 .‏

Baskara FR. The promises and pitfalls of using chat GPT for self-determined learning in higher education: an argumentative review. Prosiding Seminar Nasional Fakultas Tarbiyah dan Ilmu Keguruan IAIM Sinjai. 2023;2:95–101. https://doi.org/10.47435/sentikjar.v2i0.1825 .

Behera RK, Bala PK, Dhir A. The emerging role of cognitive computing in healthcare: a systematic literature review. Int J Med Inform. 2019;129:154–66. https://doi.org/10.1016/j.ijmedinf.2019.04.024 .

Chaka C. Detecting AI content in responses generated by ChatGPT, YouChat, and Chatsonic: the case of five AI content detection tools. J Appl Learn Teach. 2023. https://doi.org/10.37074/jalt.2023.6.2.12 .

Chiu TKF, Xia Q, Zhou X, Chai CS, Cheng M. Systematic literature review on opportunities, challenges, and future research recommendations of artificial intelligence in education. Comput Educ Artif Intell. 2023;4:100118. https://doi.org/10.1016/j.caeai.2022.100118 .

Choi EPH, Lee JJ, Ho M, Kwok JYY, Lok KYW. Chatting or cheating? The impacts of ChatGPT and other artificial intelligence language models on nurse education. Nurse Educ Today. 2023;125:105796. https://doi.org/10.1016/j.nedt.2023.105796 .

Cotton D, Cotton PA, Shipway JR. Chatting and cheating: ensuring academic integrity in the era of ChatGPT. Innov Educ Teach Int. 2023. https://doi.org/10.1080/14703297.2023.2190148 .

Crawford J, Cowling M, Allen K. Leadership is needed for ethical ChatGPT: Character, assessment, and learning using artificial intelligence (AI). J Univ Teach Learn Pract. 2023. https://doi.org/10.53761/1.20.3.02 .

Creswell JW. Educational research: planning, conducting, and evaluating quantitative and qualitative research [Ebook]. 4th ed. London: Pearson Education; 2015.

Curry D. ChatGPT Revenue and Usage Statistics (2023)—Business of Apps. 2023. https://www.businessofapps.com/data/chatgpt-statistics/

Day T. A preliminary investigation of fake peer-reviewed citations and references generated by ChatGPT. Prof Geogr. 2023. https://doi.org/10.1080/00330124.2023.2190373 .

De Castro CA. A Discussion about the Impact of ChatGPT in education: benefits and concerns. J Bus Theor Pract. 2023;11(2):p28. https://doi.org/10.22158/jbtp.v11n2p28 .

Deng X, Yu Z. A meta-analysis and systematic review of the effect of Chatbot technology use in sustainable education. Sustainability. 2023;15(4):2940. https://doi.org/10.3390/su15042940 .

Eke DO. ChatGPT and the rise of generative AI: threat to academic integrity? J Responsib Technol. 2023;13:100060. https://doi.org/10.1016/j.jrt.2023.100060 .

Elmoazen R, Saqr M, Tedre M, Hirsto L. A systematic literature review of empirical research on epistemic network analysis in education. IEEE Access. 2022;10:17330–48. https://doi.org/10.1109/access.2022.3149812 .

Farrokhnia M, Banihashem SK, Noroozi O, Wals AEJ. A SWOT analysis of ChatGPT: implications for educational practice and research. Innov Educ Teach Int. 2023. https://doi.org/10.1080/14703297.2023.2195846 .

Fergus S, Botha M, Ostovar M. Evaluating academic answers generated using ChatGPT. J Chem Educ. 2023;100(4):1672–5. https://doi.org/10.1021/acs.jchemed.3c00087 .

Fink A. Conducting research literature reviews: from the Internet to Paper. Incorporated: SAGE Publications; 2010.

Firaina R, Sulisworo D. Exploring the usage of ChatGPT in higher education: frequency and impact on productivity. Buletin Edukasi Indonesia (BEI). 2023;2(01):39–46. https://doi.org/10.56741/bei.v2i01.310 .

Firat, M. (2023). How chat GPT can transform autodidactic experiences and open education.  Department of Distance Education, Open Education Faculty, Anadolu Unive .‏ https://orcid.org/0000-0001-8707-5918

Firat M. What ChatGPT means for universities: perceptions of scholars and students. J Appl Learn Teach. 2023. https://doi.org/10.37074/jalt.2023.6.1.22 .

Fuchs K. Exploring the opportunities and challenges of NLP models in higher education: is Chat GPT a blessing or a curse? Front Educ. 2023. https://doi.org/10.3389/feduc.2023.1166682 .

García-Peñalvo FJ. La percepción de la inteligencia artificial en contextos educativos tras el lanzamiento de ChatGPT: disrupción o pánico. Educ Knowl Soc. 2023;24: e31279. https://doi.org/10.14201/eks.31279 .

Gilson A, Safranek CW, Huang T, Socrates V, Chi L, Taylor A, Chartash D. How does ChatGPT perform on the United States medical Licensing examination? The implications of large language models for medical education and knowledge assessment. JMIR Med Educ. 2023;9: e45312. https://doi.org/10.2196/45312 .

Hashana AJ, Brundha P, Ayoobkhan MUA, Fazila S. Deep Learning in ChatGPT—A Survey. In   2023 7th international conference on trends in electronics and informatics (ICOEI) . 2023. (pp. 1001–1005). IEEE. https://doi.org/10.1109/icoei56765.2023.10125852

Hirsto L, Saqr M, López-Pernas S, Valtonen T. (2022). A systematic narrative review of learning analytics research in K-12 and schools.  Proceedings . https://ceur-ws.org/Vol-3383/FLAIEC22_paper_9536.pdf

Hisan UK, Amri MM. ChatGPT and medical education: a double-edged sword. J Pedag Educ Sci. 2023;2(01):71–89. https://doi.org/10.13140/RG.2.2.31280.23043/1 .

Hopkins AM, Logan JM, Kichenadasse G, Sorich MJ. Artificial intelligence chatbots will revolutionize how cancer patients access information: ChatGPT represents a paradigm-shift. JNCI Cancer Spectr. 2023. https://doi.org/10.1093/jncics/pkad010 .

Househ M, AlSaad R, Alhuwail D, Ahmed A, Healy MG, Latifi S, Sheikh J. Large Language models in medical education: opportunities, challenges, and future directions. JMIR Med Educ. 2023;9: e48291. https://doi.org/10.2196/48291 .

Ilkka T. The impact of artificial intelligence on learning, teaching, and education. Minist de Educ. 2018. https://doi.org/10.2760/12297 .

Iqbal N, Ahmed H, Azhar KA. Exploring teachers’ attitudes towards using CHATGPT. Globa J Manag Adm Sci. 2022;3(4):97–111. https://doi.org/10.46568/gjmas.v3i4.163 .

Irfan M, Murray L, Ali S. Integration of Artificial intelligence in academia: a case study of critical teaching and learning in Higher education. Globa Soc Sci Rev. 2023;8(1):352–64. https://doi.org/10.31703/gssr.2023(viii-i).32 .

Jeon JH, Lee S. Large language models in education: a focus on the complementary relationship between human teachers and ChatGPT. Educ Inf Technol. 2023. https://doi.org/10.1007/s10639-023-11834-1 .

Khan RA, Jawaid M, Khan AR, Sajjad M. ChatGPT—Reshaping medical education and clinical management. Pak J Med Sci. 2023. https://doi.org/10.12669/pjms.39.2.7653 .

King MR. A conversation on artificial intelligence, Chatbots, and plagiarism in higher education. Cell Mol Bioeng. 2023;16(1):1–2. https://doi.org/10.1007/s12195-022-00754-8 .

Kooli C. Chatbots in education and research: a critical examination of ethical implications and solutions. Sustainability. 2023;15(7):5614. https://doi.org/10.3390/su15075614 .

Kuhail MA, Alturki N, Alramlawi S, Alhejori K. Interacting with educational chatbots: a systematic review. Educ Inf Technol. 2022;28(1):973–1018. https://doi.org/10.1007/s10639-022-11177-3 .

Lee H. The rise of ChatGPT: exploring its potential in medical education. Anat Sci Educ. 2023. https://doi.org/10.1002/ase.2270 .

Li L, Subbareddy R, Raghavendra CG. AI intelligence Chatbot to improve students learning in the higher education platform. J Interconnect Netw. 2022. https://doi.org/10.1142/s0219265921430325 .

Limna P. A Review of Artificial Intelligence (AI) in Education during the Digital Era. 2022. https://ssrn.com/abstract=4160798

Lo CK. What is the impact of ChatGPT on education? A rapid review of the literature. Educ Sci. 2023;13(4):410. https://doi.org/10.3390/educsci13040410 .

Luo W, He H, Liu J, Berson IR, Berson MJ, Zhou Y, Li H. Aladdin’s genie or pandora’s box For early childhood education? Experts chat on the roles, challenges, and developments of ChatGPT. Early Educ Dev. 2023. https://doi.org/10.1080/10409289.2023.2214181 .

Meyer JG, Urbanowicz RJ, Martin P, O’Connor K, Li R, Peng P, Moore JH. ChatGPT and large language models in academia: opportunities and challenges. Biodata Min. 2023. https://doi.org/10.1186/s13040-023-00339-9 .

Mhlanga D. Open AI in education, the responsible and ethical use of ChatGPT towards lifelong learning. Soc Sci Res Netw. 2023. https://doi.org/10.2139/ssrn.4354422 .

Neumann, M., Rauschenberger, M., & Schön, E. M. (2023). “We Need To Talk About ChatGPT”: The Future of AI and Higher Education.‏ https://doi.org/10.1109/seeng59157.2023.00010

Nolan B. Here are the schools and colleges that have banned the use of ChatGPT over plagiarism and misinformation fears. Business Insider . 2023. https://www.businessinsider.com

O’Leary DE. An analysis of three chatbots: BlenderBot, ChatGPT and LaMDA. Int J Intell Syst Account, Financ Manag. 2023;30(1):41–54. https://doi.org/10.1002/isaf.1531 .

Okoli C. A guide to conducting a standalone systematic literature review. Commun Assoc Inf Syst. 2015. https://doi.org/10.17705/1cais.03743 .

OpenAI. (2023). https://openai.com/blog/chatgpt

Perkins M. Academic integrity considerations of AI large language models in the post-pandemic era: ChatGPT and beyond. J Univ Teach Learn Pract. 2023. https://doi.org/10.53761/1.20.02.07 .

Plevris V, Papazafeiropoulos G, Rios AJ. Chatbots put to the test in math and logic problems: A preliminary comparison and assessment of ChatGPT-3.5, ChatGPT-4, and Google Bard. arXiv (Cornell University) . 2023. https://doi.org/10.48550/arxiv.2305.18618

Rahman MM, Watanobe Y (2023) ChatGPT for education and research: opportunities, threats, and strategies. Appl Sci 13(9):5783. https://doi.org/10.3390/app13095783

Ram B, Verma P. Artificial intelligence AI-based Chatbot study of ChatGPT, google AI bard and baidu AI. World J Adv Eng Technol Sci. 2023;8(1):258–61. https://doi.org/10.30574/wjaets.2023.8.1.0045 .

Rasul T, Nair S, Kalendra D, Robin M, de Oliveira Santini F, Ladeira WJ, Heathcote L. The role of ChatGPT in higher education: benefits, challenges, and future research directions. J Appl Learn Teach. 2023. https://doi.org/10.37074/jalt.2023.6.1.29 .

Ratnam M, Sharm B, Tomer A. ChatGPT: educational artificial intelligence. Int J Adv Trends Comput Sci Eng. 2023;12(2):84–91. https://doi.org/10.30534/ijatcse/2023/091222023 .

Ray PP. ChatGPT: a comprehensive review on background, applications, key challenges, bias, ethics, limitations and future scope. Internet Things Cyber-Phys Syst. 2023;3:121–54. https://doi.org/10.1016/j.iotcps.2023.04.003 .

Roumeliotis KI, Tselikas ND. ChatGPT and Open-AI models: a preliminary review. Future Internet. 2023;15(6):192. https://doi.org/10.3390/fi15060192 .

Rudolph J, Tan S, Tan S. War of the chatbots: Bard, Bing Chat, ChatGPT, Ernie and beyond. The new AI gold rush and its impact on higher education. J Appl Learn Teach. 2023. https://doi.org/10.37074/jalt.2023.6.1.23 .

Ruiz LMS, Moll-López S, Nuñez-Pérez A, Moraño J, Vega-Fleitas E. ChatGPT challenges blended learning methodologies in engineering education: a case study in mathematics. Appl Sci. 2023;13(10):6039. https://doi.org/10.3390/app13106039 .

Sallam M, Salim NA, Barakat M, Al-Tammemi AB. ChatGPT applications in medical, dental, pharmacy, and public health education: a descriptive study highlighting the advantages and limitations. Narra J. 2023;3(1): e103. https://doi.org/10.52225/narra.v3i1.103 .

Salvagno M, Taccone FS, Gerli AG. Can artificial intelligence help for scientific writing? Crit Care. 2023. https://doi.org/10.1186/s13054-023-04380-2 .

Saqr M, López-Pernas S, Helske S, Hrastinski S. The longitudinal association between engagement and achievement varies by time, students’ profiles, and achievement state: a full program study. Comput Educ. 2023;199:104787. https://doi.org/10.1016/j.compedu.2023.104787 .

Saqr M, Matcha W, Uzir N, Jovanović J, Gašević D, López-Pernas S. Transferring effective learning strategies across learning contexts matters: a study in problem-based learning. Australas J Educ Technol. 2023;39(3):9.

Schöbel S, Schmitt A, Benner D, Saqr M, Janson A, Leimeister JM. Charting the evolution and future of conversational agents: a research agenda along five waves and new frontiers. Inf Syst Front. 2023. https://doi.org/10.1007/s10796-023-10375-9 .

Shoufan A. Exploring students’ perceptions of CHATGPT: thematic analysis and follow-up survey. IEEE Access. 2023. https://doi.org/10.1109/access.2023.3268224 .

Sonderegger S, Seufert S. Chatbot-mediated learning: conceptual framework for the design of Chatbot use cases in education. Gallen: Institute for Educational Management and Technologies, University of St; 2022. https://doi.org/10.5220/0010999200003182 .

Book   Google Scholar  

Strzelecki A. To use or not to use ChatGPT in higher education? A study of students’ acceptance and use of technology. Interact Learn Environ. 2023. https://doi.org/10.1080/10494820.2023.2209881 .

Su J, Yang W. Unlocking the power of ChatGPT: a framework for applying generative AI in education. ECNU Rev Educ. 2023. https://doi.org/10.1177/20965311231168423 .

Sullivan M, Kelly A, McLaughlan P. ChatGPT in higher education: Considerations for academic integrity and student learning. J ApplLearn Teach. 2023;6(1):1–10. https://doi.org/10.37074/jalt.2023.6.1.17 .

Szabo A. ChatGPT is a breakthrough in science and education but fails a test in sports and exercise psychology. Balt J Sport Health Sci. 2023;1(128):25–40. https://doi.org/10.33607/bjshs.v127i4.1233 .

Taecharungroj V. “What can ChatGPT do?” analyzing early reactions to the innovative AI chatbot on Twitter. Big Data Cognit Comput. 2023;7(1):35. https://doi.org/10.3390/bdcc7010035 .

Tam S, Said RB. User preferences for ChatGPT-powered conversational interfaces versus traditional methods. Biomed Eng Soc. 2023. https://doi.org/10.58496/mjcsc/2023/004 .

Tedre M, Kahila J, Vartiainen H. (2023). Exploration on how co-designing with AI facilitates critical evaluation of ethics of AI in craft education. In: Langran E, Christensen P, Sanson J (Eds).  Proceedings of Society for Information Technology and Teacher Education International Conference . 2023. pp. 2289–2296.

Tlili A, Shehata B, Adarkwah MA, Bozkurt A, Hickey DT, Huang R, Agyemang B. What if the devil is my guardian angel: ChatGPT as a case study of using chatbots in education. Smart Learn Environ. 2023. https://doi.org/10.1186/s40561-023-00237-x .

Uddin SMJ, Albert A, Ovid A, Alsharef A. Leveraging CHATGPT to aid construction hazard recognition and support safety education and training. Sustainability. 2023;15(9):7121. https://doi.org/10.3390/su15097121 .

Valtonen T, López-Pernas S, Saqr M, Vartiainen H, Sointu E, Tedre M. The nature and building blocks of educational technology research. Comput Hum Behav. 2022;128:107123. https://doi.org/10.1016/j.chb.2021.107123 .

Vartiainen H, Tedre M. Using artificial intelligence in craft education: crafting with text-to-image generative models. Digit Creat. 2023;34(1):1–21. https://doi.org/10.1080/14626268.2023.2174557 .

Ventayen RJM. OpenAI ChatGPT generated results: similarity index of artificial intelligence-based contents. Soc Sci Res Netw. 2023. https://doi.org/10.2139/ssrn.4332664 .

Wagner MW, Ertl-Wagner BB. Accuracy of information and references using ChatGPT-3 for retrieval of clinical radiological information. Can Assoc Radiol J. 2023. https://doi.org/10.1177/08465371231171125 .

Wardat Y, Tashtoush MA, AlAli R, Jarrah AM. ChatGPT: a revolutionary tool for teaching and learning mathematics. Eurasia J Math, Sci Technol Educ. 2023;19(7):em2286. https://doi.org/10.29333/ejmste/13272 .

Webster J, Watson RT. Analyzing the past to prepare for the future: writing a literature review. Manag Inf Syst Quart. 2002;26(2):3.

Xiao Y, Watson ME. Guidance on conducting a systematic literature review. J Plan Educ Res. 2017;39(1):93–112. https://doi.org/10.1177/0739456x17723971 .

Yan D. Impact of ChatGPT on learners in a L2 writing practicum: an exploratory investigation. Educ Inf Technol. 2023. https://doi.org/10.1007/s10639-023-11742-4 .

Yu H. Reflection on whether Chat GPT should be banned by academia from the perspective of education and teaching. Front Psychol. 2023;14:1181712. https://doi.org/10.3389/fpsyg.2023.1181712 .

Zhu C, Sun M, Luo J, Li T, Wang M. How to harness the potential of ChatGPT in education? Knowl Manag ELearn. 2023;15(2):133–52. https://doi.org/10.34105/j.kmel.2023.15.008 .

Download references

The paper is co-funded by the Academy of Finland (Suomen Akatemia) Research Council for Natural Sciences and Engineering for the project Towards precision education: Idiographic learning analytics (TOPEILA), Decision Number 350560.

Author information

Authors and affiliations.

School of Computing, University of Eastern Finland, 80100, Joensuu, Finland

Yazid Albadarin, Mohammed Saqr, Nicolas Pope & Markku Tukiainen

You can also search for this author in PubMed   Google Scholar

Contributions

YA contributed to the literature search, data analysis, discussion, and conclusion. Additionally, YA contributed to the manuscript’s writing, editing, and finalization. MS contributed to the study’s design, conceptualization, acquisition of funding, project administration, allocation of resources, supervision, validation, literature search, and analysis of results. Furthermore, MS contributed to the manuscript's writing, revising, and approving it in its finalized state. NP contributed to the results, and discussions, and provided supervision. NP also contributed to the writing process, revisions, and the final approval of the manuscript in its finalized state. MT contributed to the study's conceptualization, resource management, supervision, writing, revising the manuscript, and approving it.

Corresponding author

Correspondence to Yazid Albadarin .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

See Table  4

The process of synthesizing the data presented in Table  4 involved identifying the relevant studies through a search process of databases (ERIC, Scopus, Web of Knowledge, Dimensions.ai, and lens.org) using specific keywords "ChatGPT" and "education". Following this, inclusion/exclusion criteria were applied, and data extraction was performed using Creswell's [ 15 ] coding techniques to capture key information and identify common themes across the included studies.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Albadarin, Y., Saqr, M., Pope, N. et al. A systematic literature review of empirical research on ChatGPT in education. Discov Educ 3 , 60 (2024). https://doi.org/10.1007/s44217-024-00138-2

Download citation

Received : 22 October 2023

Accepted : 10 May 2024

Published : 26 May 2024

DOI : https://doi.org/10.1007/s44217-024-00138-2

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Large language models
• Educational technology
• Systematic review

Advertisement

• Find a journal
• Publish with us
• Track your research

BU Researcher Named a 2024 Hertz Fellow; Award Honors “Innovators with the Greatest Potential”

Hertz Fellow Emmy Blumenthal says the opportunity to conduct research as a BU undergraduate was fundamental to “forming my intellectual identity.” Photo courtesy of Blumenthal

Biophysicist and recent alum Emmy Blumenthal wins prestigious fellowship that comes with up to $250,000 in funding

Andrew thurston.

A Boston University researcher has been ranked among the nation’s “most promising innovators in science and technology.” Emmy Blumenthal has been named a 2024 Hertz Fellow —one of just 18 recipients of the prestigious honor this year.

Given by the Fannie and John Hertz Foundation, the award includes five years of funding, worth up to $250,000 , as well as mentorship and networking opportunities. According to the foundation, past winners have gone on to win Nobel Prizes and MacArthur “genius grants,” launch more than 375 companies, and secure over 3,000 patents.

Blumenthal, who is preparing to start a PhD program, specializes in biophysics, a field that uses quantitative techniques and tools from physics to figure out how complex biological and ecological systems work.

“Being named a Hertz Fellow is a significant validation of my and my collaborators’ work, and the scientific significance of theoretical biophysics as a whole,” says Blumenthal (CAS’23), a BU postbaccalaureate research assistant. “It is also incredibly encouraging that the Hertz Foundation views me as a promising future scientist. I hope my work in graduate school is impactful and I make good use of the freedom the fellowship provides.”

The foundation was launched in the 1950s by auto industry entrepreneur John D. Hertz—best known for the Hertz rental car company—who donated his fortune to a fund for undergraduate scholarships; the nonprofit later transitioned to supporting graduate students. The foundation says it backs “innovators with the greatest potential to create transformative solutions to the world’s most urgent challenges.”

Although Blumenthal’s work is formidably esoteric—“analyzing ecological models and other biophysical systems with a very large number of degrees of freedom interacting complexly, applying and expanding tools from statistical and thermal physics”—their goals are straightforward: come up with ways of providing “a deeper understanding of the natural world and facilitating future breakthroughs that might improve human well-being.”

Blumenthal’s former professors—and now colleagues—in BU’s physics department say the recognition from Hertz is well-deserved, and acknowledge Blumenthal not just for their talent, but also for their generosity in helping other students.

“Emmy was, from the very beginning, much more than a superb student,” says Shyamsunder Erramilli, a BU College of Arts & Sciences chair and professor of physics , who spoke to The Brink to share reflections on Blumenthal from across his department. “In class, Emmy would courageously offer speculative ideas to open-ended questions; out of the classroom, Emmy would dive deeper into a problem and uncover high-quality and relevant published material. In a lab class, Emmy creatively invented solutions, such as using machine vision software of their own creation.”

Blumenthal credits their time as an undergraduate in BU’s physics department with sparking a passion for science—and for helping them develop and grow as a researcher.

“The research opportunities provided by faculty members—particularly Professors Alex Sushkov and Pankaj Mehta—were fundamental to introducing me to physics research and forming my intellectual identity,” says Blumenthal, whose undergraduate contributions to research were published twice in Physical Review Letters . “The physics department at BU—especially the undergraduate community—has a uniquely supportive culture. I also feel exceptionally lucky to have been supported by BU’s Trustee Scholarship throughout my time as an undergraduate.”

In recent years, BU students have also been named Churchill, Fulbright, Goldwater, and Truman Scholars, placing Blumenthal in a long line of successful award recipients.

“They collectively demonstrate the strength of BU as a top research university, as well as BU’s serious commitment to providing undergraduates with opportunities to engage in meaningful research on campus,” says Jeff Berg , BU’s director of national and international scholarships. His office helps connect students to award opportunities, like the Hertz Fellowship. “For Emmy, whose deep passion for physics stretches back to high school, they arrived on campus and wasted no time in taking full advantage of the research landscape here. It has been so exciting and rewarding to see Emmy’s growth and achievements across their years at BU.”

This fall, Blumenthal will begin doctoral studies at Princeton University, but for now, they are digging into their research at BU.

“One of the things that makes the Hertz selection process especially selective and rigorous is the program’s interest in candidates who want to leverage their talents to have a positive impact on the world,” says Amie Grills , BU’s associate provost for undergraduate affairs. “Emmy not only demonstrates the exceptional research talent, but also the personal integrity that the Hertz Fellowship demands.

“I hope this inspires current undergraduates to explore how they might participate in research and other experiential learning opportunities aligned with their interests at BU.”

Explore Related Topics:

• Undergraduate Research
• Share this story
• 1 Comments Add

Editor, The Brink Twitter Profile

Andrew Thurston is originally from England, but has grown to appreciate the serial comma and the Red Sox, while keeping his accent (mostly) and love of West Ham United. He joined BU in 2007, and is the editor of the University’s research news site, The Brink ; he was formerly director of alumni publications. Before joining BU, he edited consumer and business magazines, including for corporations, nonprofits, and the UK government. His work has won awards from the Council for Advancement and Support of Education, the In-House Agency Forum, Folio: , and the British Association of Communicators in Business. Andrew has a bachelor’s degree in English and related literature from the University of York. Profile

Comments & Discussion

Boston University moderates comments to facilitate an informed, substantive, civil conversation. Abusive, profane, self-promotional, misleading, incoherent or off-topic comments will be rejected. Moderators are staffed during regular business hours (EST) and can only accept comments written in English. Statistics or facts must include a citation or a link to the citation.

Post a comment. Cancel reply

Your email address will not be published. Required fields are marked *

Latest from The Brink

Is ai biased against some groups and spreading misinformation and extreme views, 3d printing robot uses ai machine learning for us army research, permanent birth control—vasectomies and tubal ligations—up following supreme court abortion ruling, explain this how do planets form, bu’s innovator of the year has pioneered devices to advance astronomy, microscopy, eye exams, bu study shows a correlation between social media use and desire for cosmetic procedures, covid-19 photo contest winners capture moments of joy, sorrow, meaning in crisis, how high-level lawsuits are disrupting climate change policies, the h5n1 bird flu is a growing threat for farm animals and humans—how serious is it, why is a bu researcher so fascinated with the diets of dung beetles, we are underestimating the health harms of climate disasters, three bu researchers elected aaas fellows, should people be fined for sleeping outside, secrets of ancient egyptian nile valley settlements found in forgotten treasure, not having job flexibility or security can leave workers feeling depressed, anxious, and hopeless, bu electrical engineer vivek goyal named a 2024 guggenheim fellow, can the bias in algorithms help us see our own, do immigrants and immigration help the economy, how do people carry such heavy loads on their heads.

Suggestions or feedback?

MIT News | Massachusetts Institute of Technology

• Machine learning
• Social justice
• Black holes
• Classes and programs

Departments

• Aeronautics and Astronautics
• Brain and Cognitive Sciences
• Architecture
• Political Science
• Mechanical Engineering

Centers, Labs, & Programs

• Abdul Latif Jameel Poverty Action Lab (J-PAL)
• Picower Institute for Learning and Memory
• Lincoln Laboratory
• School of Architecture + Planning
• School of Engineering
• School of Humanities, Arts, and Social Sciences
• Sloan School of Management
• School of Science
• MIT Schwarzman College of Computing

Scientists identify mechanism behind drug resistance in malaria parasite

Press contact :.

Previous image Next image

Share this news article on:

Related links.

• Peter Dedon
• Antimicrobial Resistance Interdisciplinary Research Group
• Singapore-MIT Alliance for Research and Technology (SMART)
• Department of Biological Engineering

Related Topics

• Drug resistance
• Antibiotics
• Biological engineering
• Drug development
• International initiatives
• Collaboration

Related Articles

Satellite-based method measures carbon in peat bogs

Newly discovered bacterial communication system aids antimicrobial resistance

SMART launches research group to advance AI, automation, and the future of work

A novel combination therapy counters antibiotic-resistant Mycobacterium abscessus infections

Previous item Next item

More MIT News

Getting to systemic sustainability

Read full story →

New MIT-LUMA Lab created to address climate challenges in the Mediterranean region

MIT Press releases Direct to Open impact report

Modeling the threat of nuclear war

Modular, scalable hardware architecture for a quantum computer

Looking for a specific action in a video? This AI-based method can find it for you

• More news on MIT News homepage →

Massachusetts Institute of Technology 77 Massachusetts Avenue, Cambridge, MA, USA

• Map (opens in new window)
• Events (opens in new window)
• People (opens in new window)
• Careers (opens in new window)
• Accessibility
• Social Media Hub
• MIT on Facebook
• MIT on YouTube
• MIT on Instagram