paper note research

13.5 Research Process: Making Notes, Synthesizing Information, and Keeping a Research Log

Learning outcomes.

By the end of this section, you will be able to:

• Employ the methods and technologies commonly used for research and communication within various fields.
• Practice and apply strategies such as interpretation, synthesis, response, and critique to compose texts that integrate the writer’s ideas with those from appropriate sources.
• Analyze and make informed decisions about intellectual property based on the concepts that motivate them.
• Apply citation conventions systematically.

As you conduct research, you will work with a range of “texts” in various forms, including sources and documents from online databases as well as images, audio, and video files from the Internet. You may also work with archival materials and with transcribed and analyzed primary data. Additionally, you will be taking notes and recording quotations from secondary sources as you find materials that shape your understanding of your topic and, at the same time, provide you with facts and perspectives. You also may download articles as PDFs that you then annotate. Like many other students, you may find it challenging to keep so much material organized, accessible, and easy to work with while you write a major research paper. As it does for many of those students, a research log for your ideas and sources will help you keep track of the scope, purpose, and possibilities of any research project.

A research log is essentially a journal in which you collect information, ask questions, and monitor the results. Even if you are completing the annotated bibliography for Writing Process: Informing and Analyzing , keeping a research log is an effective organizational tool. Like Lily Tran’s research log entry, most entries have three parts: a part for notes on secondary sources, a part for connections to the thesis or main points, and a part for your own notes or questions. Record source notes by date, and allow room to add cross-references to other entries.

Summary of Assignment: Research Log

Your assignment is to create a research log similar to the student model. You will use it for the argumentative research project assigned in Writing Process: Integrating Research to record all secondary source information: your notes, complete publication data, relation to thesis, and other information as indicated in the right-hand column of the sample entry.

Another Lens. A somewhat different approach to maintaining a research log is to customize it to your needs or preferences. You can apply shading or color coding to headers, rows, and/or columns in the three-column format (for colors and shading). Or you can add columns to accommodate more information, analysis, synthesis, or commentary, formatting them as you wish. Consider adding a column for questions only or one for connections to other sources. Finally, consider a different visual format , such as one without columns. Another possibility is to record some of your comments and questions so that you have an aural rather than a written record of these.

Writing Center

At this point, or at any other point during the research and writing process, you may find that your school’s writing center can provide extensive assistance. If you are unfamiliar with the writing center, now is a good time to pay your first visit. Writing centers provide free peer tutoring for all types and phases of writing. Discussing your research with a trained writing center tutor can help you clarify, analyze, and connect ideas as well as provide feedback on works in progress.

Quick Launch: Beginning Questions

You may begin your research log with some open pages in which you freewrite, exploring answers to the following questions. Although you generally would do this at the beginning, it is a process to which you likely will return as you find more information about your topic and as your focus changes, as it may during the course of your research.

• What information have I found so far?
• What do I still need to find?
• Where am I most likely to find it?

These are beginning questions. Like Lily Tran, however, you will come across general questions or issues that a quick note or freewrite may help you resolve. The key to this section is to revisit it regularly. Written answers to these and other self-generated questions in your log clarify your tasks as you go along, helping you articulate ideas and examine supporting evidence critically. As you move further into the process, consider answering the following questions in your freewrite:

• What evidence looks as though it best supports my thesis?
• What evidence challenges my working thesis?
• How is my thesis changing from where it started?

Creating the Research Log

As you gather source material for your argumentative research paper, keep in mind that the research is intended to support original thinking. That is, you are not writing an informational report in which you simply supply facts to readers. Instead, you are writing to support a thesis that shows original thinking, and you are collecting and incorporating research into your paper to support that thinking. Therefore, a research log, whether digital or handwritten, is a great way to keep track of your thinking as well as your notes and bibliographic information.

In the model below, Lily Tran records the correct MLA bibliographic citation for the source. Then, she records a note and includes the in-text citation here to avoid having to retrieve this information later. Perhaps most important, Tran records why she noted this information—how it supports her thesis: The human race must turn to sustainable food systems that provide healthy diets with minimal environmental impact, starting now . Finally, she makes a note to herself about an additional visual to include in the final paper to reinforce the point regarding the current pressure on food systems. And she connects the information to other information she finds, thus cross-referencing and establishing a possible synthesis. Use a format similar to that in Table 13.4 to begin your own research log.

Types of Research Notes

Taking good notes will make the research process easier by enabling you to locate and remember sources and use them effectively. While some research projects requiring only a few sources may seem easily tracked, research projects requiring more than a few sources are more effectively managed when you take good bibliographic and informational notes. As you gather evidence for your argumentative research paper, follow the descriptions and the electronic model to record your notes. You can combine these with your research log, or you can use the research log for secondary sources and your own note-taking system for primary sources if a division of this kind is helpful. Either way, be sure to include all necessary information.

Bibliographic Notes

These identify the source you are using. When you locate a useful source, record the information necessary to find that source again. It is important to do this as you find each source, even before taking notes from it. If you create bibliographic notes as you go along, then you can easily arrange them in alphabetical order later to prepare the reference list required at the end of formal academic papers. If your instructor requires you to use MLA formatting for your essay, be sure to record the following information:

• Title of source
• Title of container (larger work in which source is included)
• Other contributors
• Publication date

When using MLA style with online sources, also record the following information:

• Date of original publication
• Date of access
• DOI (A DOI, or digital object identifier, is a series of digits and letters that leads to the location of an online source. Articles in journals are often assigned DOIs to ensure that the source can be located, even if the URL changes. If your source is listed with a DOI, use that instead of a URL.)

It is important to understand which documentation style your instructor will require you to use. Check the Handbook for MLA Documentation and Format and APA Documentation and Format styles . In addition, you can check the style guide information provided by the Purdue Online Writing Lab .

Informational Notes

These notes record the relevant information found in your sources. When writing your essay, you will work from these notes, so be sure they contain all the information you need from every source you intend to use. Also try to focus your notes on your research question so that their relevance is clear when you read them later. To avoid confusion, work with separate entries for each piece of information recorded. At the top of each entry, identify the source through brief bibliographic identification (author and title), and note the page numbers on which the information appears. Also helpful is to add personal notes, including ideas for possible use of the information or cross-references to other information. As noted in Writing Process: Integrating Research , you will be using a variety of formats when borrowing from sources. Below is a quick review of these formats in terms of note-taking processes. By clarifying whether you are quoting directly, paraphrasing, or summarizing during these stages, you can record information accurately and thus take steps to avoid plagiarism.

Direct Quotations, Paraphrases, and Summaries

A direct quotation is an exact duplication of the author’s words as they appear in the original source. In your notes, put quotation marks around direct quotations so that you remember these words are the author’s, not yours. One advantage of copying exact quotations is that it allows you to decide later whether to include a quotation, paraphrase, or summary. ln general, though, use direct quotations only when the author’s words are particularly lively or persuasive.

A paraphrase is a restatement of the author’s words in your own words. Paraphrase to simplify or clarify the original author’s point. In your notes, use paraphrases when you need to record details but not exact words.

A summary is a brief condensation or distillation of the main point and most important details of the original source. Write a summary in your own words, with facts and ideas accurately represented. A summary is useful when specific details in the source are unimportant or irrelevant to your research question. You may find you can summarize several paragraphs or even an entire article or chapter in just a few sentences without losing useful information. It is a good idea to note when your entry contains a summary to remind you later that it omits detailed information. See Writing Process Integrating Research for more detailed information and examples of quotations, paraphrases, and summaries and when to use them.

Other Systems for Organizing Research Logs and Digital Note-Taking

Students often become frustrated and at times overwhelmed by the quantity of materials to be managed in the research process. If this is your first time working with both primary and secondary sources, finding ways to keep all of the information in one place and well organized is essential.

Because gathering primary evidence may be a relatively new practice, this section is designed to help you navigate the process. As mentioned earlier, information gathered in fieldwork is not cataloged, organized, indexed, or shelved for your convenience. Obtaining it requires diligence, energy, and planning. Online resources can assist you with keeping a research log. Your college library may have subscriptions to tools such as Todoist or EndNote. Consult with a librarian to find out whether you have access to any of these. If not, use something like the template shown in Figure 13.8 , or another like it, as a template for creating your own research notes and organizational tool. You will need to have a record of all field research data as well as the research log for all secondary sources.

As an Amazon Associate we earn from qualifying purchases.

This book may not be used in the training of large language models or otherwise be ingested into large language models or generative AI offerings without OpenStax's permission.

Want to cite, share, or modify this book? This book uses the Creative Commons Attribution License and you must attribute OpenStax.

Access for free at https://openstax.org/books/writing-guide/pages/1-unit-introduction

• Authors: Michelle Bachelor Robinson, Maria Jerskey, featuring Toby Fulwiler
• Publisher/website: OpenStax
• Book title: Writing Guide with Handbook
• Publication date: Dec 21, 2021
• Location: Houston, Texas
• Book URL: https://openstax.org/books/writing-guide/pages/1-unit-introduction
• Section URL: https://openstax.org/books/writing-guide/pages/13-5-research-process-making-notes-synthesizing-information-and-keeping-a-research-log

© Dec 19, 2023 OpenStax. Textbook content produced by OpenStax is licensed under a Creative Commons Attribution License . The OpenStax name, OpenStax logo, OpenStax book covers, OpenStax CNX name, and OpenStax CNX logo are not subject to the Creative Commons license and may not be reproduced without the prior and express written consent of Rice University.

Office of Undergraduate Research

• Office of Undergraduate Research FAQ's
• URSA Engage
• Resources for Students
• Resources for Faculty
• Engaging in Research
• Spring Poster Symposium (SPS)
• Ecampus SPS Videos
• Earn Money by Participating in Research Studies
• Transcript Notation
• Student Publications

How to take Research Notes

How to take research notes.

Your research notebook is an important piece of information useful for future projects and presentations. Maintaining organized and legible notes allows your research notebook to be a valuable resource to you and your research group. It allows others and yourself to replicate experiments, and it also serves as a useful troubleshooting tool. Besides it being an important part of the research process, taking detailed notes of your research will help you stay organized and allow you to easily review your work.

Here are some common reasons to maintain organized notes:

• Keeps a record of your goals and thoughts during your research experiments.
• Keeps a record of what worked and what didn't in your research experiments.
• Enables others to use your notes as a guide for similar procedures and techniques.
• A helpful tool to reference when writing a paper, submitting a proposal, or giving a presentation.
• Assists you in answering experimental questions.
• Useful to efficiently share experimental approaches, data, and results with others.

Before taking notes:

• Ask your research professor what note-taking method they recommend or prefer.
• Consider what type of media you'll be using to take notes.
• Once you have decided on how you'll be taking notes, be sure to keep all of your notes in one place to remain organized.
• Plan on taking notes regularly (meetings, important dates, procedures, journal/manuscript revisions, etc.).
• This is useful when applying to programs or internships that ask about your research experience.

Note Taking Tips:

Taking notes by hand:.

• Research notebooks don’t belong to you so make sure your notes are legible for others.
• Use post-it notes or tabs to flag important sections.
• Start sorting your notes early so that you don't become backed up and disorganized.
• Only write with a pen as pencils aren’t permanent & sharpies can bleed through.
• Make it a habit to write in your notebook and not directly on sticky notes or paper towels. Rewriting notes can waste time and sometimes lead to inaccurate data or results.

Taking Notes Electronically

• Make sure your device is charged and backed up to store data.
• Invest in note-taking apps or E-Ink tablets
• Create shortcuts to your folders so you have easier access
• Create outlines.
• Keep your notes short and legible.

Note Taking Tips Continued:

Things to avoid.

• Avoid using pencils or markers that may bleed through.
• Avoid erasing entries. Instead, draw a straight line through any mistakes and write the date next to the crossed-out information.
• Avoid writing in cursive.
• Avoid delaying your entries so you don’t fall behind and forget information.

Formatting Tips

• Use bullet points to condense your notes to make them simpler to access or color-code them.
• Tracking your failures and mistakes can improve your work in the future.
• If possible, take notes as you’re experimenting or make time at the end of each workday to get it done.
• Record the date at the start of every day, including all dates spent on research.

Types of media to use when taking notes:

Traditional paper notebook.

• Pros: Able to take quick notes, convenient access to notes, cheaper option
• Cons: Requires a table of contents or tabs as it is not easily searchable, can get damaged easily, needs to be scanned if making a digital copy

Electronic notebook  

• Apple Notes  
• Pros: Easily searchable, note-taking apps available, easy to edit & customize
• Cons: Can be difficult to find notes if they are unorganized, not as easy to take quick notes, can be a more expensive option

Combination of both

Contact info.

618 Kerr Administration Building Corvallis, OR 97331

541-737-5105

Research Guides

Gould library, reading well and taking research notes.

• How to read for college
• How to take research notes
• How to use sources in your writing
• Tools for note taking and annotations
• Mobile apps for notes and annotations
• Assistive technology
• How to cite your sources

Be Prepared: Keep track of which notes are direct quotes, which are summary, and which are your own thoughts. For example, enclose direct quotes in quotation marks, and enclose your own thoughts in brackets. That way you'll never be confused when you're writing.

Be Clear: Make sure you have noted the source and page number!

Be Organized: Keep your notes organized but in a single place so that you can refer back to notes about other readings at the same time.

Be Consistent: You'll want to find specific notes later, and one way to do that is to be consistent in the way you describe things. If you use consistent terms or tags or keywords, you'll be able to find your way back more easily.

Recording what you find

Take full notes

Whether you take notes on cards, in a notebook, or on the computer, it's vital to record information accurately and completely. Otherwise, you won't be able to trust your own notes. Most importantly, distinguish between (1) direct quotation; (2) paraphrases and summaries of the text; and (3) your own thoughts. On a computer, you have many options for making these distinctions, such as parentheses, brackets, italic or bold text, etc.

Know when to quote, paraphrase, and summarize

• Summarize when you only need to remember the main point of the passage, chapter, etc.
• Paraphrase when you are able to able to clearly state a source's point or meaning in your own words.
• Quote exactly when you need the author's exact words or authority as evidience to back up your claim. You may also want to be sure and use the author's exact wording, either because they stated their point so well, or because you want to refute that point and need to demonstrate you aren't misrepresenting the author's words.

Get the context right

Don't just record the author's words or ideas; be sure and capture the context and meaning that surrounds those ideas as well. It can be easy to take a short quote from an author that completely misrepresents his or her actual intentions if you fail to take the context into account. You should also be sure to note when the author is paraphrasing or summarizing another author's point of view--don't accidentally represent those ideas as the ideas of the author.

Example of reading notes

Here is an example of reading notes taken in Evernote, with citation and page numbers noted as well as quotation marks for direct quotes and brackets around the reader's own thoughts.

• << Previous: How to read for college
• Next: How to use sources in your writing >>
• Last Updated: Feb 7, 2024 12:22 PM
• URL: https://gouldguides.carleton.edu/activereading

Questions? Contact [email protected]

Powered by Springshare.

Finally see how to stop getting stuck in a project management tool

20 min. personalized consultation with a project management expert

Smart Note-Taking for Research Paper Writing

How to organize research notes using the Zettelkasten Method when writing academic papers

With plenty of note-taking tips and apps available, online and in paper form, it’s become extremely easy to take note of information, ideas, or thoughts. As simple as it is to write down an idea or jot down a quote, the skill of academic research and writing for a thesis paper is on another level entirely. And keeping a record or an archive of all of the information you need can quickly require a very organized system.

The use of index cards seems old-fashioned considering that note-taking apps (psst! Hypernotes ) offer better functionality and are arguably more user-friendly. However, software is only there to help aid our individual workflow and thinking process. That’s why understanding and learning how to properly research, take notes and write academic papers is still a highly valuable skill.

Let’s Start Writing! But Where to Start…

Writing academic papers is a vital skill most students need to learn and practice. Academic papers are usually time-intensive pieces of written content that are a requirement throughout school or at University. Whether a topic is assigned or you have to choose your own, there’s little room for variation in how to begin.

Popular and purposeful in analyzing and evaluating the knowledge of the author as well as assessing if the learning objectives were met, research papers serve as information-packed content. Most of us may not end up working jobs in academic professions or be researchers at institutions, where writing research papers is also part of the job, but we often read such papers. 

Despite the fact that most research papers or dissertations aren’t often read in full, journalists, academics, and other professionals regularly use academic papers as a basis for further literary publications or blog articles. The standard of academic papers ensures the validity of the information and gives the content authority. 

There’s no-nonsense in research papers. To make sure to write convincing and correct content, the research stage is extremely important. And, naturally, when doing any kind of research, we take notes.

Why Take Notes?

There are particular standards defined for writing academic papers . In order to meet these standards, a specific amount of background information and researched literature is required. Taking notes helps keep track of read/consumed literary material as well as keeps a file of any information that may be of importance to the topic. 

The aim of writing isn’t merely to advertise fully formed opinions, but also serves the purpose of developing opinions worth sharing in the first place. 

What is Note-Taking?

Note-taking (sometimes written as notetaking or note-taking ) is the practice of recording information from different sources and platforms. For academic writing, note-taking is the process of obtaining and compiling information that answers and supports the research paper’s questions and topic. Notes can be in one of three forms: summary, paraphrase, or direct quotation.

Note-taking is an excellent process useful for anyone to turn individual thoughts and information into organized ideas ready to be communicated through writing. Notes are, however, only as valuable as the context. Since notes are also a byproduct of the information we consume daily, it’s important to categorize information, develop connections, and establish relationships between pieces of information. 

What Type of Notes Can I Take?

• Explanation of complex theories
• Background information on events or persons of interest
• Definitions of terms
• Quotations of significant value
• Illustrations or graphics

Note-Taking 101

Taking notes or doing research for academic papers shouldn’t be that difficult, considering we take notes all the time. Wrong. Note-taking for research papers isn’t the same as quickly noting down an interesting slogan or cool quote from a video, putting it on a sticky note, and slapping it onto your bedroom or office wall.

Note-taking for research papers requires focus and careful deliberation of which information is important to note down, keep on file, or use and reference in your own writing. Depending on the topic and requirements of your research paper from your University or institution, your notes might include explanations of complex theories, definitions, quotations, and graphics. 

Stages of Research Paper Writing

1. Preparation Stage

Before you start, it’s recommended to draft a plan or an outline of how you wish to begin preparing to write your research paper. Make note of the topic you will be writing on, as well as the stylistic and literary requirements for your paper.

2. Research Stage

In the research stage, finding good and useful literary material for background knowledge is vital. To find particular publications on a topic, you can use Google Scholar or access literary databases and institutions made available to you through your school, university, or institution. 

Make sure to write down the source location of the literary material you find. Always include the reference title, author, page number, and source destination. This saves you time when formatting your paper in the later stages and helps keep the information you collect organized and referenceable.

In the worst-case scenario, you’ll have to do a backward search to find the source of a quote you wrote down without reference to the original literary material. 

3. Writing Stage

When writing, an outline or paper structure is helpful to visually break up the piece into sections. Once you have defined the sections, you can begin writing and referencing the information you have collected in the research stage.

Clearly mark which text pieces and information where you relied on background knowledge, which texts are directly sourced, and which information you summarized or have written in your own words. This is where your paper starts to take shape.  

4. Draft Stage

After organizing all of your collected notes and starting to write your paper, you are already in the draft stage. In the draft stage, the background information collected and the text written in your own words come together. Every piece of information is structured by the subtopics and sections you defined in the previous stages. 

5. Final Stage

Success! Well… almost! In the final stage, you look over your whole paper and check for consistency and any irrelevancies. Read through the entire paper for clarity, grammatical errors , and peace of mind that you have included everything important. 

Make sure you use the correct formatting and referencing method requested by your University or institution for research papers. Don’t forget to save it and then send the paper on its way.

Best Practice Note-Taking Tips

• Find relevant and authoritative literary material through the search bar of literary databases and institutions.
• Practice citation repeatedly! Always keep a record of the reference book title, author, page number, and source location. At best, format the citation in the necessary format from the beginning. 
• Organize your notes according to topic or reference to easily find the information again when in the writing stage. Work invested in the early stages eases the writing and editing process of the later stages.
• Summarize research notes and write in your own words as much as possible. Cite direct quotes and clearly mark copied text in your notes to avoid plagiarism.  

Take Smart Notes

Taking smart notes isn’t as difficult as it seems. It’s simply a matter of principle, defined structure, and consistency. Whether you opt for a paper-based system or use a digital tool to write and organize your notes depends solely on your individual personality, needs, and workflow.

With various productivity apps promoting diverse techniques, a good note-taking system to take smart notes is the Zettelkasten Method . Invented by Niklas Luhmann, a german sociologist and researcher, the Zettelkasten Method is known as the smart note-taking method that popularized personalized knowledge management. 

As a strategic process for thinking and writing, the Zettelkasten Method helps you organize your knowledge while working, studying, or researching. Directly translated as a ‘note box’, Zettelkasten is simply a framework to help organize your ideas, thoughts, and information by relating pieces of knowledge and connecting pieces of information to each other.

Hypernotes is a note-taking app that can be used as a software-based Zettelkasten, with integrated features to make smart note-taking so much easier, such as auto-connecting related notes, and syncing to multiple devices. In each notebook, you can create an archive of your thoughts, ideas, and information. 

Using the tag system to connect like-minded ideas and information to one another and letting Hypernotes do its thing with bi-directional linking, you’ll soon create a web of knowledge about anything you’ve ever taken note of. This feature is extremely helpful to navigate through the enormous amounts of information you’ve written down. Another benefit is that it assists you in categorizing and making connections between your ideas, thoughts, and saved information in a single notebook. Navigate through your notes, ideas, and knowledge easily.

Ready, Set, Go!

Writing academic papers is no simple task. Depending on the requirements, resources available, and your personal research and writing style, techniques, apps, or practice help keep you organized and increase your productivity. 

Whether you use a particular note-taking app like Hypernotes for your research paper writing or opt for a paper-based system, make sure you follow a particular structure. Repeat the steps that help you find the information you need quicker and allow you to reproduce or create knowledge naturally.

Images from NeONBRAND , hana_k and Surface via Unsplash 

A well-written piece is made up of authoritative sources and uses the art of connecting ideas, thoughts, and information together. Good luck to all students and professionals working on research paper writing! We hope these tips help you in organizing the information and aid your workflow in your writing process.

Cheers, Jessica and the Zenkit Team

FREE 20 MIN. CONSULTATION WITH A PROJECT MANAGEMENT EXPERT

Wanna see how to simplify your workflow with Zenkit in less than a day?

• digital note app
• how to smart notes
• how to take notes
• hypernotes note app
• hypernotes take notes
• note archive software
• note taking app for students
• note taking tips
• note-taking
• note-taking app
• organize research paper
• reading notes
• research note taking
• research notes
• research paper writing
• smart notes
• taking notes zettelkasten method
• thesis writing
• writing a research paper
• writing a thesis paper
• zettelkasten method

More from Karen Bradford

10 Ways to Remember What You Study

More from Kelly Moser

How Hot Desking Elevates the Office Environment in 2024

More from Chris Harley

8 Productivity Tools for Successful Content Marketing

2 thoughts on “ Smart Note-Taking for Research Paper Writing ”

Thanks for sending really an exquisite text.

Great article thank you for sharing!

Leave a Reply 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.

Zenkit Comment Policy

At Zenkit, we strive to post helpful, informative, and timely content. We want you to feel welcome to comment with your own thoughts, feedback, and critiques, however we do not welcome inappropriate or rude comments. We reserve the right to delete comments or ban users from commenting as needed to keep our comments section relevant and respectful.

What we encourage:

• Smart, informed, and helpful comments that contribute to the topic. Funny commentary is also thoroughly encouraged.
• Constructive criticism, either of the article itself or the ideas contained in it.
• Found technical issues with the site? Send an email to [email protected] and specify the issue and what kind of device, operating system, and OS version you are using.
• Noticed spam or inappropriate behaviour that we haven’t yet sorted out? Flag the comment or report the offending post to [email protected] .

What we’d rather you avoid:

Rude or inappropriate comments.

We welcome heated discourse, and we’re aware that some topics cover things people feel passionately about. We encourage you to voice your opinions, however in order for discussions to remain constructive, we ask that you remember to criticize ideas, not people.

Please avoid:

• name-calling
• ad hominem attacks
• responding to a post’s tone instead of its actual content
• knee-jerk contradiction

Comments that we find to be hateful, inflammatory, threatening, or harassing may be removed. This includes racist, gendered, ableist, ageist, homophobic, and transphobic slurs and language of any sort. If you don’t have something nice to say about another user, don't say it. Treat others the way you’d like to be treated.

Trolling or generally unkind behaviour

If you’re just here to wreak havoc and have some fun, and you’re not contributing meaningfully to the discussions, we will take actions to remove you from the conversation. Please also avoid flagging or downvoting other users’ comments just because you disagree with them.

Every interpretation of spamming is prohibited. This means no unauthorized promotion of your own brand, product, or blog, unauthorized advertisements, links to any kind of online gambling, malicious sites, or otherwise inappropriate material.

Comments that are irrelevant or that show you haven’t read the article

We know that some comments can veer into different topics at times, but remain related to the original topic. Be polite, and please avoid promoting off-topic commentary. Ditto avoid complaining we failed to mention certain topics when they were clearly covered in the piece. Make sure you read through the whole piece before saying your piece!

Breaches of privacy

This should really go without saying, but please do not post personal information that may be used by others for malicious purposes. Please also do not impersonate authors of this blog, or other commenters (that’s just weird).

• Cambridge Libraries

Study Skills

Research skills.

• Searching the literature
• Note making for dissertations
• Research Data Management
• Copyright and licenses
• Publishing in journals
• Publishing academic books
• Depositing your thesis
• Research metrics
• Build your online profile
• Finding support

Note making for dissertations: First steps into writing

Note making (as opposed to note taking) is an active practice of recording relevant parts of reading for your research as well as your reflections and critiques of those studies. Note making, therefore, is a pre-writing exercise that helps you to organise your thoughts prior to writing. In this module, we will cover:

• The difference between note taking and note making
• Seven tips for good note making
• Strategies for structuring your notes and asking critical questions
• Different styles of note making

To complete this section, you will need:

• Approximately 20-30 minutes.
• Access to the internet. All the resources used here are available freely.
• Some equipment for jotting down your thoughts, a pen and paper will do, or your phone or another electronic device.

Note taking v note making

When you think about note taking, what comes to mind? Perhaps trying to record everything said in a lecture? Perhaps trying to write down everything included in readings required for a course?

• Note taking is a passive process. When you take notes, you are often trying to record everything that you are reading or listening to. However, you may have noticed that this takes a lot of effort and often results in too many notes to be useful.  
• Note making , on the other hand, is an active practice, based on the needs and priorities of your project. Note making is an opportunity for you to ask critical questions of your readings and to synthesise ideas as they pertain to your research questions. Making notes is a pre-writing exercise that develops your academic voice and makes writing significantly easier.

Seven tips for effective note making

Note making is an active process based on the needs of your research. This video contains seven tips to help you make brilliant notes from articles and books to make the most of the time you spend reading and writing.

• Transcript of Seven Tips for Effective Notemaking

Question prompts for strategic note making

You might consider structuring your notes to answer the following questions. Remember that note making is based on your needs, so not all of these questions will apply in all cases. You might try answering these questions using the note making styles discussed in the next section.

• Question prompts for strategic note making
• Background question prompts
• Critical question prompts
• Synthesis question prompts

Answer these six questions to frame your reading and provide context.

• What is the context in which the text was written? What came before it? Are there competing ideas?
• Who is the intended audience?
• What is the author’s purpose?
• How is the writing organised?
• What are the author’s methods?
• What is the author’s key argument and conclusions?

Answer these six questions to determine your critical perspectivess and develop your academic voice.

• What are the most interesting/compelling ideas (to you) in this study?
• Why do you find them interesting? How do they relate to your study?
• What questions do you have about the study?
• What could it cover better? How could it have defended its research better?
• What are the implications of the study? (Look not just to the conclusions but also to definitions and models)
• Are there any gaps in the study? (Look not just at conclusions but definitions, literature review, methodology)

Answer these five questions to compare aspects of various studies (such as for a literature review. 

• What are the similarities and differences in the literature?
• Critically analyse the strengths, limitations, debates and themes that emerg from the literature.
• What would you suggest for future research or practice?
• Where are the gaps in the literature? What is missing? Why?
• What new questions should be asked in this area of study?

Styles of note making

• Linear notes . Great for recording thoughts about your readings. [video]
• Mind mapping : Great for thinking through complex topics. [video]

Further sites that discuss techniques for note making:

• Note-taking techniques
• Common note-taking methods
• Strategies for effective note making  

Did you know?

How did you find this Research Skills module

Image Credits: Image #1: David Travis on Unsplash ; Image #2: Charles Deluvio on Unsplash

• << Previous: Searching the literature
• Next: Research Data Management >>
• Last Updated: Apr 11, 2024 9:35 AM
• URL: https://libguides.cam.ac.uk/research-skills

© Cambridge University Libraries | Accessibility | Privacy policy | Log into LibApps

Writing a Research Paper: 5. Taking Notes & Documenting Sources

• Getting Started
• 1. Topic Ideas
• 2. Thesis Statement & Outline
• 3. Appropriate Sources
• 4. Search Techniques
• 5. Taking Notes & Documenting Sources
• 6. Evaluating Sources
• 7. Citations & Plagiarism
• 8. Writing Your Research Paper

Taking Notes & Documenting Sources

How to take notes and document sources.

Note taking is a very important part of the research process.  It will help you:

• keep your ideas and sources organized
• effectively use the information you find
• avoid plagiarism

When you find good information to be used in your paper:

• Read the text critically, think how it is related to your argument, and decide how you are going to use it in your paper.
• Select the material that is relevant to your argument.
• Copy the original text for direct quotations or briefly summarize the content in your own words, and make note of how you will use it.
• Copy the citation or publication information of the source.
• << Previous: 4. Search Techniques
• Next: 6. Evaluating Sources >>
• Last Updated: Sep 26, 2023 5:26 PM
• URL: https://kenrick.libguides.com/writing-a-research-paper

How to Do Research: A Step-By-Step Guide: 4a. Take Notes

• Get Started
• 1a. Select a Topic
• 1b. Develop Research Questions
• 1c. Identify Keywords
• 1d. Find Background Information
• 1e. Refine a Topic
• 2a. Search Strategies
• 2d. Articles
• 2e. Videos & Images
• 2f. Databases
• 2g. Websites
• 2h. Grey Literature
• 2i. Open Access Materials
• 3a. Evaluate Sources
• 3b. Primary vs. Secondary
• 3c. Types of Periodicals
• 4a. Take Notes
• 4b. Outline the Paper
• 4c. Incorporate Source Material
• 5a. Avoid Plagiarism
• 5b. Zotero & MyBib
• 5c. MLA Formatting
• 5d. MLA Citation Examples
• 5e. APA Formatting
• 5f. APA Citation Examples
• 5g. Annotated Bibliographies

Note Taking in Bibliographic Management Tools

We encourage students to use bibliographic citation management tools (such as Zotero, EasyBib and RefWorks) to keep track of their research citations. Each service includes a note-taking function. Find more information about citation management tools here . Whether or not you're using one of these, the tips below will help you.

Tips for Taking Notes Electronically

• Try using a bibliographic citation management tool to keep track of your sources and to take notes.
• As you add sources, put them in the format you're using (MLA, APA, Chicago, etc.).
• Group sources by publication type (i.e., book, article, website).
• Number each source within the publication type group.
• For websites, include the URL information and the date you accessed each site.
• Next to each idea, include the source number from the Works Cited file and the page number from the source. See the examples below. Note that #A5 and #B2 refer to article source 5 and book source 2 from the Works Cited file.

#A5 p.35: 76.69% of the hyperlinks selected from homepage are for articles and the catalog #B2 p.76: online library guides evolved from the paper pathfinders of the 1960s

• When done taking notes, assign keywords or sub-topic headings to each idea, quote or summary.
• Use the copy and paste feature to group keywords or sub-topic ideas together.
• Back up your master list and note files frequently!

Tips for Taking Notes by Hand

• Use index cards to keep notes and track sources used in your paper.
• Include the citation (i.e., author, title, publisher, date, page numbers, etc.) in the format you're using. It will be easier to organize the sources alphabetically when creating the Works Cited page.
• Number the source cards.
• Use only one side to record a single idea, fact or quote from one source. It will be easier to rearrange them later when it comes time to organize your paper.
• Include a heading or key words at the top of the card. 
• Include the Work Cited source card number.
• Include the page number where you found the information.
• Use abbreviations, acronyms, or incomplete sentences to record information to speed up the notetaking process.
• Write down only the information that answers your research questions.
• Use symbols, diagrams, charts or drawings to simplify and visualize ideas.

Forms of Notetaking

Use one of these notetaking forms to capture information:

• Summarize : Capture the main ideas of the source succinctly by restating them in your own words.
• Paraphrase : Restate the author's ideas in your own words.
• Quote : Copy the quotation exactly as it appears in the original source. Put quotation marks around the text and note the name of the person you are quoting.

Example of a Work Cited Card

Example notecard.

• << Previous: Step 4: Write
• Next: 4b. Outline the Paper >>
• Last Updated: Feb 21, 2024 11:01 AM
• URL: https://libguides.elmira.edu/research

• Utility Menu

• ARC Scheduler
• Student Employment
• Note-taking

Think about how you take notes during class. Do you use a specific system? Do you feel that system is working for you? What could be improved? How might taking notes during a lecture, section, or seminar be different online versus in the classroom? 

Adjust how you take notes during synchronous vs. asynchronous learning (slightly) . 

First, let’s distinguish between  synchronous  and  asynchronous  instruction. Synchronous classes are live with the instructor and students together, and asynchronous instruction is material recorded by the professor for viewing by students at another time. Sometimes asynchronous instruction may include a recording of a live Zoom session with the instructor and students. 

With this distinction in mind,  here are some tips on how to take notes during both types of instruction:

Taking notes during live classes (synchronous instruction).

Taking notes when watching recorded classes (asynchronous instruction)., check in with yourself., if available, annotate lecture slides during lecture., consider writing notes by hand., review your notes., write down questions..

Below are some common and effective note-taking techniques: 

Cornell Notes

If you are looking for help with using some of the tips and techniques described above, come to the ARC’s note-taking workshop, offered several times every semester.

Register for ARC Workshops

Accordion style.

• Assessing Your Understanding
• Building Your Academic Support System
• Common Class Norms
• Effective Learning Practices
• First-Year Students
• How to Prepare for Class
• Interacting with Instructors
• Know and Honor Your Priorities
• Memory and Attention
• Minimizing Zoom Fatigue
• Office Hours
• Perfectionism
• Scheduling Time
• Senior Theses
• Study Groups
• Tackling STEM Courses
• Test Anxiety

• USC Libraries
• Research Guides

Organizing Your Social Sciences Research Paper

• Footnotes or Endnotes?
• 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
• Limitations of the Study
• Common Grammar Mistakes
• Writing Concisely
• Avoiding Plagiarism
• Further Readings
• Generative AI and Writing
• USC Libraries Tutorials and Other Guides
• Bibliography

Endnote Note citing a particular source or making a brief explanatory comment placed at the end of a research paper and arranged sequentially in relation to where the reference appears in the paper.

Footnote Note citing a particular source or making a brief explanatory comment placed at the bottom of a page corresponding to the item cited in the corresponding text above.

Fiske, Robert Hartwell. To the Point: A Dictionary of Concise Writing . New York: W.W. Norton and Company, 2014.

Structure and Writing Style

Advantages of Using Endnotes

• Endnotes are less distracting to the reader and allows the narrative to flow better.
• Endnotes don't clutter up the page.
• As a separate section of a research paper, endnotes allow the reader to read and contemplate all the notes at once.

Disadvantages of Using Endnotes

• If you want to look at the text of a particular endnote, you have to flip to the end of the research paper to find the information.
• Depending on how they are created [i.e., continuous numbering or numbers that start over for each chapter], you may have to remember the chapter number as well as the endnote number in order to find the correct one.
• Endnotes may carry a negative connotation much like the proverbial "fine print" or hidden disclaimers in advertising. A reader may believe you are trying to hide something by burying it in a hard-to-find endnote.

Advantages of Using Footnotes

• Readers interested in identifying the source or note can quickly glance down the page to find what they are looking for.
• It allows the reader to immediately link the footnote to the subject of the text without having to take the time to find the note at the back of the paper.
• Footnotes are automatically included when printing off specific pages.

Disadvantages of Using Footnotes

• Footnotes can clutter up the page and, thus, negatively impact the overall look of the page.
• If there are multiple columns, charts, or tables below only a small segment of text that includes a footnote, then you must decide where the footnotes should appear.
• If the footnotes are lengthy, there's a risk they could dominate the page, although this issue is considered acceptable in legal scholarship.
• Adding lengthy footnotes after the paper has been completed can alter the page where other sources are located [i.e., a long footnote can push text to the next page].
• It is more difficult learning how to insert footnotes using your word processing program than simply adding endnotes at the end of your paper.

Things to keep in mind when considering using either endnotes or footnotes in your research paper :

1.    Footnotes are numbered consecutively throughout a research paper, except for those notes accompanying special material (e.g., figures, tables, charts, etc.). Numbering of footnotes are "superscript"--Arabic numbers typed slightly above the line of text. Do not include periods, parentheses, or slashes. They can follow all punctuation marks except dashes. In general, to avoid interrupting the continuity of the text, footnote numbers are placed at the end of the sentence, clause, or phrase containing the quoted or paraphrased material. 2.    Depending on the writing style used in your class, endnotes may take the place of a list of resources cited in your paper or they may represent non-bibliographic items, such as comments or observations, followed by a separate list of references to the sources you cited and arranged alphabetically by the author's last name. If you are unsure about how to use endnotes, consult with your professor. 3.    In general, the use of footnotes in most academic writing is now considered a bit outdated and has been replaced by endnotes, which are much easier to place in your paper, even with the advent of word processing programs. However, some disciplines, such as law and history, still predominantly utilize footnotes. Consult with your professor about which form to use and always remember that, whichever style of citation you choose, apply it consistently throughout your paper.

NOTE:   Always think critically about the information you place in a footnote or endnote. Ask yourself, is this supplementary or tangential information that would otherwise disrupt the narrative flow of the text or is this essential information that I should integrate into the main text? If you are not sure, it's better to work it into the text. Too many notes implies a disorganized paper.

Cermak, Bonni and Jennifer Troxell. A Guide to Footnotes and Endnotes for NASA History Authors . NASA History Program. History Division; Hale, Ali. Should You Use Footnotes or Endnotes? DailyWritingTips.com; Tables, Appendices, Footnotes and Endnotes. The Writing Lab and The OWL. Purdue University; Lunsford, Andrea A. and Robert Connors. The St. Martin's Handbook . New York: St. Martin's Press, 1989; Saller, Carol. “Endnotes or Footnotes? Some Considerations.” The Chronicle of Higher Education 58 (January 6, 2012): http://chronicle.com/blogs/linguafranca/2012/01/06/endnotes-or-footnotes-some-considerations/.

• << Previous: Avoiding Plagiarism
• Next: Further Readings >>
• Last Updated: May 2, 2024 4:39 PM
• URL: https://libguides.usc.edu/writingguide

Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.

9 Organizing Research: Taking and Keeping Effective Notes

Once you’ve located the right primary and secondary sources, it’s time to glean all the information you can from them. In this chapter, you’ll first get some tips on taking and organizing notes. The second part addresses how to approach the sort of intermediary assignments (such as book reviews) that are often part of a history course.

Honing your own strategy for organizing your primary and secondary research is a pathway to less stress and better paper success. Moreover, if you can find the method that helps you best organize your notes, these methods can be applied to research you do for any of your classes.

Before the personal computing revolution, most historians labored through archives and primary documents and wrote down their notes on index cards, and then found innovative ways to organize them for their purposes. When doing secondary research, historians often utilized (and many still do) pen and paper for taking notes on secondary sources. With the advent of digital photography and useful note-taking tools like OneNote, some of these older methods have been phased out – though some persist. And, most importantly, once you start using some of the newer techniques below, you may find that you are a little “old school,” and might opt to integrate some of the older techniques with newer technology.

Whether you choose to use a low-tech method of taking and organizing your notes or an app that will help you organize your research, here are a few pointers for good note-taking.

Principles of note-taking

• If you are going low-tech, choose a method that prevents a loss of any notes. Perhaps use one spiral notebook, or an accordion folder, that will keep everything for your project in one space. If you end up taking notes away from your notebook or folder, replace them—or tape them onto blank pages if you are using a notebook—as soon as possible.
• If you are going high-tech, pick one application and stick with it. Using a cloud-based app, including one that you can download to your smart phone, will allow you to keep adding to your notes even if you find yourself with time to take notes unexpectedly.
• When taking notes, whether you’re using 3X5 note cards or using an app described below, write down the author and a shortened title for the publication, along with the page number on EVERY card. We can’t emphasize this point enough; writing down the bibliographic information the first time and repeatedly will save you loads of time later when you are writing your paper and must cite all key information.
• Include keywords or “tags” that capture why you thought to take down this information in a consistent place on each note card (and when using the apps described below). If you are writing a paper about why Martin Luther King, Jr., became a successful Civil Rights movement leader, for example, you may have a few theories as you read his speeches or how those around him described his leadership. Those theories—religious beliefs, choice of lieutenants, understanding of Gandhi—might become the tags you put on each note card.
• Note-taking applications can help organize tags for you, but if you are going low tech, a good idea is to put tags on the left side of a note card, and bibliographic info on the right side.

Organizing research- applications that can help

Using images in research.

• If you are in an archive: make your first picture one that includes the formal collection name, the box number, the folder name and call numbe r and anything else that would help you relocate this information if you or someone else needed to. Do this BEFORE you start taking photos of what is in the folder.
• If you are photographing a book or something you may need to return to the library: take a picture of all the front matter (the title page, the page behind the title with all the publication information, maybe even the table of contents).

Once you have recorded where you find it, resist the urge to rename these photographs. By renaming them, they may be re-ordered and you might forget where you found them. Instead, use tags for your own purposes, and carefully name and date the folder into which the photographs were automatically sorted. There is one free, open-source program, Tropy , which is designed to help organize photos taken in archives, as well as tag, annotate, and organize them. It was developed and is supported by the Roy Rosenzweig Center for History and New Media at George Mason University. It is free to download, and you can find it here: https://tropy.org/ ; it is not, however, cloud-based, so you should back up your photos. In other cases, if an archive doesn’t allow photography (this is highly unlikely if you’ve made the trip to the archive), you might have a laptop on hand so that you can transcribe crucial documents.

Using note or project-organizing apps

When you have the time to sit down and begin taking notes on your primary sources, you can annotate your photos in Tropy. Alternatively, OneNote, which is cloud-based, can serve as a way to organize your research. OneNote allows you to create separate “Notebooks” for various projects, but this doesn’t preclude you from searching for terms or tags across projects if the need ever arises. Within each project you can start new tabs, say, for each different collection that you have documents from, or you can start new tabs for different themes that you are investigating. Just as in Tropy, as you go through taking notes on your documents you can create your own “tags” and place them wherever you want in the notes.

Another powerful, free tool to help organize research, especially secondary research though not exclusively, is Zotero found @ https://www.zotero.org/ . Once downloaded, you can begin to save sources (and their URL) that you find on the internet to Zotero. You can create main folders for each major project that you have and then subfolders for various themes if you would like. Just like the other software mentioned, you can create notes and tags about each source, and Zotero can also be used to create bibliographies in the precise format that you will be using. Obviously, this function is super useful when doing a long-term, expansive project like a thesis or dissertation.

How History is Made: A Student’s Guide to Reading, Writing, and Thinking in the Discipline Copyright © 2022 by Stephanie Cole; Kimberly Breuer; Scott W. Palmer; and Brandon Blakeslee is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License , except where otherwise noted.

Share This Book

• Skip to primary navigation
• Skip to content

• Home – AI for Research

AI for Researchers

Effective research writing, reading & analysis with ai.

Avidnote enables you to write or read research papers faster, as well as analyze your research data with our AI templates.

Write faster

Write research papers, rewrite text, lit reviews, discussion etc

Effective reading

Summarize anything, proofread text, find research gaps etc

Analyze your data

Input your data and find correlations and insights.

and more...

Research methods, transcribe interviews, and more

AI for Reading Papers

Use the ai to engage with your papers. quickly get answers on your questions., upload a paper.

Begin by uploading a paper (or other document) that you want to read.

Use AI functions

Use our AI to ask question to your paper, summarize text, get explanations for texts, proofread and much more…

Save your notes

Easily save your notes. In Avidnote, you always have your notes and your paper side-by-side, no need to switch between windows.

AI for Writing Research

Describe your study (or paste text), and the AI generates text, rewrites, analyzes data and much more. Avidnote contains a long list of AI functionalities designed specifically for researchers.

Recommended by universities        

Quantitative research, ai for data analysis.

Analyze your research data and get insights on potential correlations or findings.

Qualitative research

Transcribe interviews automatically.

Transcribe and code/analyze interviews in minutes with AI, not hours. Save hours of work.

– No need to transcribe for hours on end. – Simply upload a file, and AI gives you a transcription within minutes. – Supports numerous languages. – Keeps your data private.

Supported by researchers across the world

Avidnote is developed by researchers, for researchers. It is used by researchers in both academia and enterprise.

Pricing Plans

Start out free. upgrade anytime., 5,000 ai generated words, ai research templates, ai reading (10 times), ai analysis, export as pdf/word, transcribe automatically, ai transcription (2 hours), ai code / analyze interviews, manage papers / docs, 300 mb storage, upload papers / docs, pdf / word / pptx / xlsx, take notes papers / docs, create research projects, create tags, create workbooks, manage references, find new papers, always included, basic support, full data privacy, cancel anytime, professional plan, the moderate plan, $ 19   month, 100,000 ai generated words / month, ai reading (unlimited), ai transcription (10 hours), 5,000 mb storage (5 gb), premium support, premium plan, the more premium plan, $ 59   month, 1 million ai generated words / month, ai transcription (100 hours), 20,000 mb storage (20 gb).

For group prices and enterprise. Please contact us . We give discounts to groups as well as students.

Questions & Answers

How do i access avidnote.

Once you register for free, you can login and start using it. Avidnote is entirely web-based, so do you do not need to download anything.

Who owns the data?

You own all of the data that you produce. Avidnote does not own any of the text that you produce. You can export all of your documents and data as Word files or PDFs.

Do you give group discounts?

Yes, please contact us for a demonstration for your research group or department. We give discounts to groups.

Can I cancel anytime?

Yes, you can cancel your plan anytime. No commitments.

Which plan should I get?

We recommend the Premium plan ($59/month) if your budget allows it. It gives you the most AI usage, transcription, storage and more.

Is my data used to train AI?

No, your data is private and is not used to train the AI algorithm. We have a data privacy agreement that ensures that your data stays private with you, and is not used to train the algorithm.

What if I am not happy?

You can cancel your plan any time.

If you have tried out Avidnote and are still not happy, we offer a full 30-day money back guarantee.

Is my data private?

Yes, all of the data that you input into Avidnote is private by default. The data is not shared with anyone else, and we secure it with industry level security and encryption measures.

What about plagiarism?

While Avidnote can be used to aid you in your research writing, it is not meant to replace the research process itself.

We do not recommend that you submit the outputted text as your own without rewriting it, adding to it, modifying it, and adding proper sources.

It is important to note that AI support tools like Avidnote are here to support you in the writing process, you as the researcher, are however responsible for utilizing the outputted text in way that is ethically sound.

AI writing tools provide us as educators an opportunity to teach students the importance of never submitting work that they have not produced themselves.

We need to encourage students to develop their own ideas and to use AI tools  as a supplement  to their research,  not  as a replacement for it.

In summary, we are committed to promoting ethical writing practices and ensuring that our platform is used for legitimate research purposes. We encourage our users to use Avidnote responsibly and to always give credit where credit is due.

Privacy Overview

Adding {{itemName}} to cart

Added {{itemName}} to cart

How To Make Notecards For Research Paper In Most Effective Way

Many supervisors, mentors, and teachers recommend their students and apprentices use research note cards while writing research papers. Notecards could be a great tool to organize your word and elements of research.

Note cards might seem like an old and outdated research method, but they still work. They do more than be a tool for you. Notecards help you organize your thoughts that are beneficial in your research and beyond. Let’s talk about some tips and tricks on how to make notecards for research papers.

Table of Contents

Why And How To Make Notecards For Research Paper?

With research note cards, it is easier to track your citations. When citing a source in your dissertation, you can write the source’s name on the note card and add the page number where you found the information. This way, you can quickly find the needed information.

Before writing notecards, look at all the information to write your research document. Once you know basic ideas, gather the main points of your research. Preferably, a 3″ x5″ note card would do your bidding.

Also, notecards look fantastic, and even if they’re scattered around the room, they would add an aesthetic touch to your room rather than making it look messy. Writing notecards will help you stay organized and  write a research paper fast .

Steps Towards Writing Notecards For Research

Here are steps to write perfect notecards for your research paper.

Get Yourself a Pack Of Fresh, Nice Smelling Notecards

When you think of how to make notecards for a research paper, the first thing that will pop up in your mind is: Where are the research note cards? For a dissertation, we will need a lot of them. Try to get some extra. That way, even if you grow short, you will have a new bundle to open and save time during your research process.

Gather More Ideas Than You Need

The more is always safe. It will be great to gather as many ideas and sources as possible when you have the  best research topic . It is the quality of a great writer to always  cite sources . It’s easier than ever to collect sources from the Internet as many as possible. The Internet is like an infinite library. When you have more data, sources, and ideas, you will have more choices to filter out the best. For example, you are  writing an outline for your dissertation  and adding critical points that you are about to discuss. You have twenty key points written on your notecards. When you reconsider and filter out the best, you will probably have half of them left, which is close to ten.

Shortlist The Sources

You have a lot of ideas and a lot of sources written on your notecards. Could you have a look at them again? Now you see that not all ideas sound impeccable anymore. You can take those notecards out, leaving you with the best of them. How easier was it with notecards? Imagine if you were doing this filtration process without notecards. You would have to write a whole new draft for this.

Use A Full Notecard For Each Idea

Remember we talked about getting extra notecards? Now you understand why. Every notecard must be devoted to a single idea. Using a separate note card for each citation, source, or quote would be best. Using one card for more than one idea will cause leaving out essential details. It will also confuse you and make you double-minded. Whatever the page number is, making index cards would always help. Whether you’re researching a 10-page research document or  writing a thesis for a research paper , every notecard must consist of a single idea, be it your own words or some text from a resource.

Write Down The Quotes

In the history of research, quoting and paraphrasing can be great tools to make your paper authentic and reliable. Please use separate notecards to include quotes. A direct statement in quotation marks or creating a bunch of them can make your research look more authentic. Note cards will help you remember where or when you will use them.

Label and Number The Note cards

Labeling and numbering note cards help you avoid trouble and confusion. Imagine the mess if your notecards suddenly fall out of your table and get scattered. It would be like having all your work wasted. You will need hours to reorganize them. Labeling and giving numbers will help you sort them and use them at the exact moment you are going to need them. If your note cards are all labeled and aligned, they can tell you a lot about  how to organize your research paper  as well.

Include Every Attribute / Aspect

A notecard must include every aspect of the source or citation you will use. Let’s create an outline of those factors. A notecard will typically include these necessary points:

• Name of The Author
• The Topic / Title of The Citation
• The Book / Paper that is cited
• Exact Number Of The Page
• Other contributors
• Editions / Versions / Volume
• Date of Publication

Let us discuss one trick that will help you beyond  writing research papers . It will help you in real life too. Whenever you do or say anything, ask yourself first:

Is it necessary?

The same goes for note cards. Only include what’s necessary.

Don’t Use Abbreviations Or Acronyms

When we are talking about how to make notecards for a research paper,  our writers  will disapprove of using abbreviations or acronyms. One abbreviation might have more than one meaning. The same goes for acronyms. This can lead to confusion. Staying accurate is the ultimate goal.

Now you can see that creating note cards for your dissertation is not rocket science if you have the right guide and  Academic writing service . We also learned that note cards are not as old as some might say, and they can help you get the best out of your research. However, if you still need clarification about how to make notecards for a research paper, wait to lose your heart. You can  contact us , and we can provide valuable insights we have learned while writing research over the years.

Order Original Papers & Essays

Your First Custom Paper Sample is on Us!

Timely Deliveries

No Plagiarism & AI

100% Refund

Try Our Free Paper Writing Service

Related blogs.

Connections with Writers and support

Privacy and Confidentiality Guarantee

Average Quality Score

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• 29 April 2024

How reliable is this research? Tool flags papers discussed on PubPeer

• Dalmeet Singh Chawla

You can also search for this author in PubMed   Google Scholar

RedacTek’s tool alerts users to PubPeer discussions, and indicates when a study, or the papers that it cites, has been retracted. Credit: deepblue4you/Getty

A free online tool released earlier this month alerts researchers if a paper cites studies that are mentioned on the website PubPeer , a forum scientists often use to raise integrity concerns surrounding published papers.

Studies are usually flagged on PubPeer when readers have suspicions, for example about image manipulation , plagiarism , data fabrication or artificial intelligence (AI)-generated text . PubPeer already offers its own browser plug-in that alerts users if a study that they are reading has been posted on the site. The new tool, a plug-in released on 13 April by RedacTek , based in Oakland, California, goes further — it searches through reference lists for papers that have been flagged. The software pulls information from many sources, including PubPeer’s database; data from the digital-infrastructure organization Crossref, which assigns digital object identifiers to articles; and OpenAlex , a free index of hundreds of millions of scientific documents.

It’s important to track mentions of referenced articles on PubPeer, says Jodi Schneider, an information scientist at the University of Illinois Urbana-Champaign, who has tried out the RedacTek plug-in. “Not every single reference that’s in the bibliography matters, but some of them do,” she adds. “When you see a large number of problems in somebody’s bibliography, that just calls everything into question.”

The aim of the tool is to flag potential problems with studies to researchers early on, to reduce the circulation of poor-quality science, says RedacTek founder Rick Meyler, who is based in Emeryville, California. Future versions might also use AI to automatically clarify whether the PubPeer comments on a paper are positive or negative, he adds.

Third-generation retractions

As well as flagging PubPeer discussions, the plug-in alerts users if a study, or a paper that it cites, has been retracted. There are existing tools that alert academics about retracted citations ; some can do this during the writing process, so that researchers are aware of the publication status of studies when constructing bibliographies. But with the new tool, users can opt in to receive notifications about further ‘generations’ of retractions — alerts cover not only the study that they are reading, but also the papers it cites, articles cited by those references and even papers cited by the secondary references.

The software also calculates a ‘retraction association value’ for studies, a metric that measures the extent to which the paper is associated with science that has been withdrawn from the literature. As well as informing individual researchers, the plug-in could help scholarly publishers to keep tabs on their own journals, Meyler says, because it allows users to filter by publication.

In its ‘paper scorecard’, the tool also flags any papers in the three generations of referenced studies in which more than 25% of papers in the bibliography are self-citations — references by authors to their previous works.

Future versions could highlight whether papers cited retracted studies before or after the retraction was issued, notes Meyler, or whether mentions of such studies acknowledge the retraction. That would be useful, says Schneider, who co-authored a 2020 analysis that found that as little as 4% of citations to retracted studies note that the referenced paper has been retracted 1 .

Meyler says that RedacTek is currently in talks with the scholarly-services firm Cabell’s International in Beaumont, Texas, which maintains pay-to-view lists of suspected predatory journals . These publish articles without running proper quality checks for issues such as plagiarism, but still collect authors’ fees. The plan is to use these lists to improve the tool so that it can also automatically flag any cited papers that are published in such journals.

doi: https://doi.org/10.1038/d41586-024-01247-6

Schneider, J., Ye, D., Hill, A. M. & Whitehorn, A. S. Scientometrics 125 , 2877–2913 (2020).

Article   Google Scholar  

Download references

Reprints and permissions

Related Articles

Predatory-journal papers have little scientific impact

Pioneer behind controversial PubPeer site reveals his identity

• Scientific community

Hunger on campus: why US PhD students are fighting over food

Career Feature 03 MAY 24

France’s research mega-campus faces leadership crisis

News 03 MAY 24

Mount Etna’s spectacular smoke rings and more — April’s best science images

Plagiarism in peer-review reports could be the ‘tip of the iceberg’

Nature Index 01 MAY 24

Algorithm ranks peer reviewers by reputation — but critics warn of bias

Nature Index 25 APR 24

Seeking Global Talents, the International School of Medicine, Zhejiang University

Welcome to apply for all levels of professors based at the International School of Medicine, Zhejiang University.

Yiwu, Zhejiang, China

International School of Medicine, Zhejiang University

Assistant, Associate, or Full Professor

Athens, Georgia

University of Georgia

Associate Professor - Synthetic Biology

Position Summary  We seek an Associate Professor in the department of Synthetic Biology (jcvi.org/research/synthetic-biology). We invite applicatio...

Rockville, Maryland

J. Craig Venter Institute

Associate or Senior Editor (microbial genetics, evolution, and epidemiology) Nature Communications

Job Title: Associate or Senior Editor (microbial genetics, evolution, and epidemiology), Nature Communications Locations: London, New York, Philade...

New York (US)

Springer Nature Ltd

Postdoctoral Research Fellow

Two postdoctoral positions offered at HMS to study the role hypothalamic leptin resistance in control of neuron activity

Boston, Massachusetts (US)

Boston Children’s Hospital-Ozcan Lab

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Special Features

Vendor voice.

Some scientists can't stop using AI to write research papers

If you read about 'meticulous commendable intricacy' there's a chance a boffin had help.

Linguistic and statistical analyses of scientific articles suggest that generative AI may have been used to write an increasing amount of scientific literature.

Two academic papers assert that analyzing word choice in the corpus of science publications reveals an increasing usage of AI for writing research papers. One study , published in March by Andrew Gray of University College London in the UK, suggests at least one percent – 60,000 or more – of all papers published in 2023 were written at least partially by AI.

A second paper published in April by a Stanford University team in the US claims this figure might range between 6.3 and 17.5 percent, depending on the topic.

Both papers looked for certain words that large language models (LLMs) use habitually, such as “intricate,” “pivotal,” and “meticulously." By tracking the use of those words across scientific literature, and comparing this to words that aren't particularly favored by AI, the two studies say they can detect an increasing reliance on machine learning within the scientific publishing community.

In Gray's paper, the use of control words like "red," "conclusion," and "after" changed by a few percent from 2019 to 2023. The same was true of other certain adjectives and adverbs until 2023 (termed the post-LLM year by Gray).

In that year use of the words "meticulous," "commendable," and "intricate," rose by 59, 83, and 117 percent respectively, while their prevalence in scientific literature hardly changed between 2019 and 2022. The word with the single biggest increase in prevalence post-2022 was “meticulously”, up 137 percent.

The Stanford paper found similar phenomena, demonstrating a sudden increase for the words "realm," "showcasing," "intricate," and "pivotal." The former two were used about 80 percent more often than in 2021 and 2022, while the latter two were used around 120 and almost 160 percent more frequently respectively.

• Beyond the hype, AI promises leg up for scientific research
• AI researchers have started reviewing their peers using AI assistance

Boffins deem Google DeepMind's material discoveries rather shallow

• Turns out AI chatbots are way more persuasive than humans

The researchers also considered word usage statistics in various scientific disciplines. Computer science and electrical engineering were ahead of the pack when it came to using AI-preferred language, while mathematics, physics, and papers published by the journal Nature, only saw increases of between five and 7.5 percent.

The Stanford bods also noted that authors posting more preprints, working in more crowded fields, and writing shorter papers seem to use AI more frequently. Their paper suggests that a general lack of time and a need to write as much as possible encourages the use of LLMs, which can help increase output.

Potentially the next big controversy in the scientific community

Using AI to help in the research process isn't anything new, and lots of boffins are open about utilizing AI to tweak experiments to achieve better results. However, using AI to actually write abstracts and other chunks of papers is very different, because the general expectation is that scientific articles are written by actual humans, not robots, and at least a couple of publishers consider using LLMs to write papers to be scientific misconduct.

Using AI models can be very risky as they often produce inaccurate text, the very thing scientific literature is not supposed to do. AI models can even fabricate quotations and citations, an occurrence that infamously got two New York attorneys in trouble for citing cases ChatGPT had dreamed up.

"Authors who are using LLM-generated text must be pressured to disclose this or to think twice about whether doing so is appropriate in the first place, as a matter of basic research integrity," University College London’s Gray opined.

The Stanford researchers also raised similar concerns, writing that use of generative AI in scientific literature could create "risks to the security and independence of scientific practice." ®

Narrower topics

• Large Language Model
• Machine Learning
• Neural Networks
• Tensor Processing Unit

Broader topics

• Self-driving Car

Send us news

Other stories you might like

With run:ai acquisition, nvidia aims to manage your ai kubes, intel's neuromorphic 'owl brain' swoops into sandia labs, google search results polluted by buggy ai-written code frustrate coders, java thriving after 30 years.

Forget the AI doom and hype, let's make computers useful

Us, japan announce joint ai research projects funded by nvidia, microsoft, others, don't rent out that container ship yet: cios and biz buyers view ai pcs with some caution, law prof predicts generative ai will die at the hands of watchdogs, politicians call for ban on 'killer robots' and the curbing of ai weapons, what's up with alphabet and microsoft lately profits, sales – and ai costs, jensen huang and sam altman among tech chiefs invited to federal ai safety board.

• Advertise with us

Our Websites

• The Next Platform
• Blocks and Files

Your Privacy

• Cookies Policy
• Privacy Policy
• Ts & Cs

Copyright. All rights reserved © 1998–2024

This paper is in the following e-collection/theme issue:

Published on 2.5.2024 in Vol 26 (2024)

Improving the Prognostic Evaluation Precision of Hospital Outcomes for Heart Failure Using Admission Notes and Clinical Tabular Data: Multimodal Deep Learning Model

Authors of this article:

Original Paper

• Zhenyue Gao 1 * , MS   ; 
• Xiaoli Liu 2 * , PhD   ; 
• Yu Kang 3 , MD   ; 
• Pan Hu 2 , MD   ; 
• Xiu Zhang 3 , MD   ; 
• Wei Yan 2 , MD   ; 
• Muyang Yan 2 , MD   ; 
• Pengming Yu 3 , MD   ; 
• Qing Zhang 3 , MD   ; 
• Wendong Xiao 1 , PhD   ; 
• Zhengbo Zhang 2 , PhD  

1 Beijing Engineering Research Center of Industrial Spectrum Imaging, School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing, China

2 Center for Artificial Intelligence in Medicine, The General Hospital of People's Liberation Army, Beijing, China

3 Department of Cardiology, West China Hospital, Sichuan University, Chengdu, China

*these authors contributed equally

Corresponding Author:

Zhengbo Zhang, PhD

Center for Artificial Intelligence in Medicine

The General Hospital of People's Liberation Army

28 Fuxing Road

Beijing, 100853

Phone: 86 010 68295454

Email: [email protected]

Background: Clinical notes contain contextualized information beyond structured data related to patients’ past and current health status.

Objective: This study aimed to design a multimodal deep learning approach to improve the evaluation precision of hospital outcomes for heart failure (HF) using admission clinical notes and easily collected tabular data.

Methods: Data for the development and validation of the multimodal model were retrospectively derived from 3 open-access US databases, including the Medical Information Mart for Intensive Care III v1.4 (MIMIC-III) and MIMIC-IV v1.0, collected from a teaching hospital from 2001 to 2019, and the eICU Collaborative Research Database v1.2, collected from 208 hospitals from 2014 to 2015. The study cohorts consisted of all patients with critical HF. The clinical notes, including chief complaint, history of present illness, physical examination, medical history, and admission medication, as well as clinical variables recorded in electronic health records, were analyzed. We developed a deep learning mortality prediction model for in-hospital patients, which underwent complete internal, prospective, and external evaluation. The Integrated Gradients and SHapley Additive exPlanations (SHAP) methods were used to analyze the importance of risk factors.

Results: The study included 9989 (16.4%) patients in the development set, 2497 (14.1%) patients in the internal validation set, 1896 (18.3%) in the prospective validation set, and 7432 (15%) patients in the external validation set. The area under the receiver operating characteristic curve of the models was 0.838 (95% CI 0.827-0.851), 0.849 (95% CI 0.841-0.856), and 0.767 (95% CI 0.762-0.772), for the internal, prospective, and external validation sets, respectively. The area under the receiver operating characteristic curve of the multimodal model outperformed that of the unimodal models in all test sets, and tabular data contributed to higher discrimination. The medical history and physical examination were more useful than other factors in early assessments.

Conclusions: The multimodal deep learning model for combining admission notes and clinical tabular data showed promising efficacy as a potentially novel method in evaluating the risk of mortality in patients with HF, providing more accurate and timely decision support.

Introduction

Heart failure (HF), a syndrome of impaired heart function, represents the advanced stage of various cardiac conditions [ 1 - 3 ]. With its substantial influence on both morbidity and mortality, HF poses a formidable challenge to human health and societal progress [ 4 - 6 ].

As a potentially life-threatening condition, particularly when accompanied by advanced organ dysfunction or severe complications, a considerable portion of patients with HF may require immediate access to advanced, high-technology, life-saving care, which is typically available only in intensive care units (ICUs) [ 7 ]. Studies have indicated that approximately 10% to 51% of patients with HF admitted to hospitals in the United States are subsequently admitted to ICUs [ 8 , 9 ]. It has also been found that ICU-admitted patients with HF experience significantly higher adjusted in-hospital mortality rates compared to those admitted solely to hospitals [ 10 ]. The in-hospital mortality rate for patients with HF receiving treatment in an ICU has been reported as 10.6%, in contrast to the overall in-hospital mortality rate of 4% for all patients with HF [ 11 ]. Given this substantially higher mortality rate, accurate prediction of in-hospital mortality could empower physicians to implement early interventions and tailor individualized treatments [ 12 , 13 ]. Consequently, there is an increasing need for the development of predictive models that can effectively identify individuals at a heightened risk of mortality in the ICU.

Most previous research works have applied statistical analysis or machine learning techniques using structured administrative data from electronic health records to identify significant risk predictors that trigger adverse outcomes [ 14 - 19 ]. However, HF disease often develops rapidly, and while some sensitive biomarkers, such as N-terminal pro–b-type natriuretic peptide (NT-proBNP), tend to increase in reactivity after the disease progresses, their efficiency is limited due to their high cost and inability to be measured in real time [ 20 , 21 ]. Recently, there has been a growing acknowledgment of the importance of clinical narratives in clinical decision-making [ 22 , 23 ]. The narrative notes at admission, such as chief complaint, history of present illness, physical examination, medical history, and admission medication, play a central role in health care communication. They represent a more comprehensive and personalized account of patient history and assessments [ 24 ]. Harnessing the potential of clinical narratives can largely enhance patient care and contribute to the improvement of predictive models for prognosis [ 25 , 26 ]. Exploiting the potential of clinical narratives and modeling them by multimodal deep learning (DL) approaches can enhance the precision of patient care and contribute to the improvement of predictive models for in-hospital mortality.

We aim to design a multimodal DL model and explore the infusion approaches to improve evaluation performance using tabular data and admission notes. The cross-modal model, which characterizes textual, categorical, and continuous variables separately, significantly outperforms the unimodal models on the multicenter and prospective validation sets. We believe our findings will motivate data-centric studies to more precisely characterize the illness severity of patients with HF.

Study Design

An overview of the study flow is shown in Figure 1 A. First, we acquired patients’ admission notes and tabular data, and these two single modalities were separately embedded to obtain the feature and status representation. The categorical and continuous variables in tabular data were characterized separately. Next, a feature-fusing DL network was applied to integrate the two modalities and achieved the model development. Then, a fully connected DL network was used to predict the in-hospital outcome, our primary outcome of interest. Two postexplanation approaches were adopted to increase the credibility of the model. Finally, the internal, prospective, and external validation with multiple evaluation metrics were accomplished. Our study followed the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) reporting guidelines for prognosis studies.

Data Sets and Cohorts

The cohorts for this multicenter retrospective cohort study were derived from 3 open-access clinical databases, including the Medical Information Mart for Intensive Care v1.4 (MIMIC-III; CareVue) and MIMIC-IV v1.0, collected from the Beth Israel Deaconess Medical Center in Boston from 2001 to 2008 and 2008 to 2019, respectively [ 27 , 28 ], and the eICU Collaborative Research Database v1.2 (eICU-CRD), collected from 208 hospitals in United States from 2014 to 2015 [ 29 ]. We included all first-time ICU admissions for patients with HF aged ≥16 years according to the International Classification of Diseases diagnostic codes. We excluded patients who stayed in the ICU for less than 24 hours and did not have admission notes. Patients were divided into 4 cohorts to support adequate model evaluation, including the development, internal validation, prospective validation, and external validation cohorts. These cohorts correspond to different stages of model development and evaluation. The development cohort consisted of a subset of data used to create or develop the predictive model. Internal validation, prospective validation, and external validation cohorts helped to check if the model had learned patterns that generalized well to new, unseen data. Each of these cohorts played a crucial role in different stages of model development and validation, ensuring that the predictive model was accurate, reliable, and applicable to new and diverse data sets or situations. The inclusion criteria of them are displayed in Figure 1 B.

Data Extraction

Our target is to provide early clinical decision support during ICU admissions. The 5 types of commonly recorded notes were extracted, including chief complaint, history of present illness, medical history, admission medications, and physical exam. In Figure S1 in Multimedia Appendix 1 , an example of an admission note is shown with highlights. Meanwhile, 6 types of clinical variables were collected for model development, as follows: (1) basic information of age, gender, weight, BMI, and Charlson Comorbidity Index; (2) vital signs, such as Glasgow Coma Scale, heart rate, respiratory rate, and systolic blood pressure; (3) laboratory tests, including glucose, creatinine, white blood cell, and total bilirubin; (4) urine output; (5) treatments received, including mechanical ventilation; and (6) physical frailty assessments, including activity and fall risks. Representative statistical features were calculated based on the type of variable, such as the maximum, minimum, and mean values. The median value of each feature was used to impute missing values for continuous variables except for FiO 2 ([fraction of inspired oxygen] with the imputation of 21%), with a missing ratio limitation of less than 30%. Details about all types of candidate variables are provided in Table S1 in Multimedia Appendix 1 . Their missing ratio is shown in Table S2 in Multimedia Appendix 1 .

Model Development and Output

The model was constructed based on a supervised multimodal DL framework, which mainly included feature extractors and a feature fusion module. A pretrained Bidirectional Encoder Representations from Transformers (BERT) module was used for learning the presentation of clinical notes [ 30 ]. In the preliminary experiments, we used all the text chunks (the same subset from the training set for the model) to compare the performances of different pretrained clinical BERT models. We found that clinical BERT [ 31 ] demonstrated the best comprehensive performance (Table S3 in Multimedia Appendix 1 ). In the fusion module, a gate attention mechanism [ 32 ] was introduced to aggregate the embedded features of clinical notes and tabular data using attention scores; this module finally output the predicted risk probability of in-hospital deaths through a fully connected layer (Figure S2 in Multimedia Appendix 1 ). The maximum predicted value of all text chunks for a patient was adopted as the optimal risk prediction score. Further detailed information on model building and training is present in the Multimedia Appendix 1 .

Model Explanation

In the pursuit of explicating the underlying mechanisms of the DL model and facilitating a comprehensive visualization of pivotal insights, we embarked on an intricate analysis of the pivotal terminologies instrumental in shaping predictions within the developed model. To achieve this, we used the Integrated Gradients (IG) technique [ 33 ] to enhance our comprehension of the BERT model’s inner workings and the rationale behind its predictions. This technique hinges on computing gradients with respect to input features, gauging each feature’s contribution to the model’s prediction. IG offers an intuitive understanding of model predictions by quantifying different features’ contributions, aiding clinicians and researchers in comprehending the model’s decisions [ 34 , 35 ]. At the same time, IG demonstrates stability across diverse samples and model architectures, yielding consistent explanatory outcomes, crucial in the face of clinical data diversity and complexity [ 36 ]. Consequently, it is considered a reliable analytical tool, helping to assess how each word in the input sequence influences the model's predictions for our research. Simultaneously, we harnessed the SHapley Additive exPlanations (SHAP) technique to unravel the importance of clinical variables in structured tabular data. We computed Shapley values to rank the important clinical variables. Shapley values involve a game theory–based approach to explain the prediction of DL models. They measure the contribution of a given feature value to the difference between the actual prediction and the mean prediction. It is important to note that higher SHAP values signify a heightened pertinence of specific terms in influencing the model’s predictions, whereas relatively diminished SHAP values connote a less pronounced impact. The IG technique exhibits a similar pattern.

Leveraging the IG and SHAP techniques offers valuable insights into the intricate relationship between input features and prediction outcomes, contributing to a more comprehensive understanding of the model’s decision-making process.

Model Validation

The discrimination performance of our prediction model was assessed on the internal (MIMIC, 2001-2016), prospective (MIMIC, 2017-2019) and external (eICU-CRD, 2014-2015) validation cohorts. This assessment compared the model against different single modalities covering notes, tabular data, and a combination of both. The importance of the 5 types of notes for outcome assessment was also examined separately. We trained 5 predictive models based on the tabular data and individual clinical notes. It should be mentioned that the chief complaint was absent in the external validation cohort. Three evaluation metrics were calculated along with their corresponding 95% CIs, the area under the receiver operating characteristic curve (AUROC), F 1 -score, and the area under the precision-recall curve.

Statistics Analysis

The median (IQR) values for continuous variables are presented. The t test (2-tailed) or the Wilcoxon Rank Sum Test was used when appropriate to compare survivors and nonsurvivors of HF. Categorical variables were reported by total numbers and percentages. Two-sided P values of less than .05 were considered statistically significant.

Ethical Considerations

This study was exempt from institutional review board approval due to the retrospective design and lack of direct patient intervention. All data from patients were retrospectively collected from the electronic health care records systems (in the form of third-party public databases or hospital health care systems), which originated from daily clinical work.

All data were de-identified before the analysis. Third-party public databases (MIMIC-IV, MIMIC-III, and eICU-CRD) were used in this study. The institutional review boards of the Massachusetts Institute of Technology (number 0403000206) and Beth Israel Deaconess Medical Center (number 2001-P-001699/14) approved the use of the database for research.

The requirement for individual patient consent was waived because the study did not impact clinical care, all protected health information was deidentified, and all available data in the databases were anonymous.

Patient Characteristics

A total of samples from 12,486 (14.1%) patients with HF were collected from MIMIC-III and MIMIC-IV joint data sets between 2001 and 2016; they were randomly divided into a development set and an independent internal validation set. Additionally, 1896 (18.3%) patients with HF were extracted from MIMIC-IV from 2017 to 2019 for a prospective validation set. For the external validation set, 7432 (15%) patients with HF were extracted from the eICU-CRD data set. Baseline characteristics are summarized in Table 1 . The proportion of patients with in-hospital mortality in the 4 cohorts ranges from 14% to 19%. Detailed comparisons of survivors and nonsurvivors in all study cohorts are shown in Tables S4-S7 in Multimedia Appendix 1 .

a CCI: Charlson Comorbidity Index.

b ICU: intensive care unit.

Model Performance Evaluation

We present the discrimination performance on internal, prospective, and external validation sets by receiver operating characteristic curves of the optimal models after tuning the hyperparameters ( Figure 2 ). The AUROCs of the multimodal model were significantly higher than the two unimodal models in all 3 types of validation evaluations. They were 0.838 (95% CI 0.827-0.851), 0.849 (95% CI 0.841-0.856), and 0.767 [0.762-0.772] for the internal, prospective, and external validation sets, respectively. Specifically, the design details of the unimodal models were as follows: the text-based unimodal model used clinical BERT, leveraging its capabilities in contextualizing clinical text data; on the other hand, the tabular unimodal model used a fully connected model structure, tailored to effectively process structured tabular data from the tables. More comparisons on baseline models, such as random forest and logistic regression, and all evaluation metrics for these models in the 3 validation types are presented in Table S8 in Multimedia Appendix 1 .

Contribution of Individual Part in Clinical Notes

The performance contributions of the 5 types of clinical notes (including chief complaint, history of present illness, medical history, admission medication, and physical exam) were separately evaluated by combining them with clinical variables to retrain all prediction models. Figure 3 and Table S9 ( Multimedia Appendix 1 ) display the AUROC comparisons with the full model. We found the individual contributions were much lower than the overall contribution in all validation cohorts. Specifically, medical history and physical exam contained more information that was useful in assessing the prognosis of patients with HF compared to other note types.

Clinical Notes Visualization and Interpretation

We applied the IG method to study the attribution of the prediction of a deep network to its input features, aiming to provide explanation for individual predictions. IG is computed based on the gradient of the prediction outputs with respect to the input words. Higher IG values denote the greater significance of a word to the model’s prediction, whereas lower values indicate lesser importance. We derived IG values for all tokens present in the clinical notes of each patient within the test data set, extracting those tokens with higher IG values. It is important to note that, due to BERT’s tokenization process, inputs are represented as tokens rather than individual words. For instance, the phrase “the patient has been extubated” is tokenized into “the patient has been ex ##tub ##ated” as the input sequence [ 35 ]. To enhance readability, we conducted postprocessing by excluding numbers, tokens with only 1 or 2 characters, and separators. A clinical expert assessed the clinical significance of tokens and their associated IG values in the context of mortality prediction. The sorted tokens are illustrated in Figure 4 A.

The analysis identified commonly ubiquitous clinical terms like “in,” “to,” and “with,” which were segregated due to their limited potential in distinguishing prognostic variations. Among the top 20 clinically meaningful indicators vital for mortality prediction, intriguing insights emerged upon clinical interpretation. For instance, “Failure” and “Pain,” the leading predictors, denote prevalent symptoms within ICU care and can mirror disease severity and disability. Indicators 4 and 9 align with pulmonary pathology, their elevated importance reflecting the gravity of respiratory conditions and the necessity for ICU interventions, such as mechanical ventilation. Additional indicators such as “pneumonia” and “fall” manifest acute illness, carrying prognostic weight in mortality prediction. Clinical cues, such as “status,” “reflex,” and “shock,” correspond to mental well-being, with their significance in prognosis attributed to the association of delirium with adverse outcomes.

Clinical Variables Feature Analysis

We ranked the important clinical variables using the SHAP technique. The top 20 out of 52 clinical variables ( Figure 4 ) show that for structured tabular data, the highest ranked variables also correlate with disease severity and poorer prognosis. These variables represent clinically important information, such as mental status, using the Glasgow Coma Scale, urine output, mechanical ventilation, activity, and respiratory rate measurements.

Principal Findings

This retrospective prognostic study aimed to develop, validate, and explain a multimodal DL prediction model for in-hospital outcomes in critically ill patients with HF. The model was constructed based on the admission notes and records from the first ICU admission day. Simultaneously, we compared the difference between multimodal and unimodal models and explored the individual importance of admission notes in the clinical practice of HF. We found that multimodality could further enhance the model’s ability and credibility to evaluate outcomes compared to unimodality.

Emerging clinical data sets provide an opportunity for the DL techniques to study the problem of in-hospital mortality prediction. Compared to previous related work, which mostly considers single modality or simply concatenates embeddings from different modalities, our work demonstrates a novel approach. We separately embed texts as well as categorical and continuous variables to integrate multimodal knowledge and leverage clinical notes information for better predictions. Our comprehensive experiments demonstrate that our proposed model outperforms the models using single modality (text-only AUROC: 0.701; tabular-only AUROC: 0.790) by achieving high performance (AUROC: 0.838).

The fusion method we used integrated two different modalities—unstructured clinical notes and structured clinical variables—into a universal shareable space using a transformer block. It was efficient to leverage clinical notes and integrate tabular data. Meanwhile, the novel application of an attention mechanism on clinical data enhanced the model’s ability to focus on evaluating the target in models when fusing multimodal information. Our ablation study, as shown in Figures S3-S5 ( Multimedia Appendix 1 ), on domain adaptive pretraining and task adaptive fine-tuning with multiple BERTs verified the significance of pretraining and fine-tuning, when implementing BERT models on natural language text, especially on domain-specific clinical notes.

In the multimodal model, the proportion of clinical variables, especially continuous variables, was much higher compared to other parts (Figure S6 in Multimedia Appendix 1 ). The analysis and visualization of important words in clinical notes also yielded interesting findings. The ranking of words by IG values provided face validity, indicating that some of the important words used for prediction were clinically related to diseases trajectory, such as the severity of respiratory disease or mental status. Some of the unspecified words, such as “disease,” used in diverse scenarios, were more difficult to interpret as isolated words. Lastly, some of the clinically meaningful words can change significantly with negation, such as “fall.” In the future, we will use more techniques, such as the NegEx algorithm, to consider negation of keywords to better explain the clinical words’ meanings.

There are some limitations in our study. First, our model leveraged the electronic health records data based on patients’ ICU admission and the first day of admission to predict in-hospital death risk. It did not include recorded data during the treatment, which might reduce the evaluation performance of the model. Second, we simplify the feature extraction, using the maximum, minimum, or mean statistical values to characterize all data throughout the day. Such simplification ignored the changes in time series, and it might have caused the loss of useful information. Time series data will be considered in our future study. Finally, we recommend that the model need to be calibrated using local data to avoid assessment bias.

Conclusions

In this multicenter prognostic study, we developed and validated an attention-multimodal DL model for in-hospital outcome prediction of patients with HF and explored the approaches that can improve the evaluation precision by simultaneously characterizing both admission notes and tabular data. The AUROCs of our model were significantly higher than those of unimodal models in all validation sets. The clinical variables included in the study made a particularly significant contribution to the overall results, with the data from the clinical notes exhibiting a much lower contribution. The model shows good predictive and explainable performance to potentially support the precise decision-making and disease management of critically ill patients with HF.

Acknowledgments

We gratefully acknowledge the guidance and assistance of Dr Max Shen from Beth Israel Deaconess Medical Center (BIDMC).

This work was supported in part by the National Natural Science Foundations of China (NSFC) under grants 62173032 and 62171471, the Foshan Science and Technology Innovation Special Projects (grant BK22BF005), the Regional Joint Fund of the Guangdong Basic and Applied Basic Research Fund (grant 2022A1515140109).

Qing Zhang ([email protected]), Wendong Xiao ([email protected]), and Zhengbo Zhang ([email protected]) are co-corresponding authors of this manuscript.

Authors' Contributions

ZG and XL had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. They also contributed to the conceptualization and design of the study. YK, PH, XZ, and WY contributed to the acquisition, analysis, or interpretation of data. MY and PY conducted statistical analysis. ZG and XL drafted the manuscript. PY, QZ, ZZ, and WX obtained funding for the study and supervised the study.

Conflicts of Interest

None declared.

Additional statistics.

• Katz AM, Rolett EL. Heart failure: when form fails to follow function. Eur Heart J. Feb 01, 2016;37(5):449-454. [ CrossRef ] [ Medline ]
• Ormerod JO, Ashrafian H, Frenneaux MP. Impaired energetics in heart failure - a new therapeutic target. Pharmacol Ther. Sep 2008;119(3):264-274. [ CrossRef ] [ Medline ]
• Pagliaro BR, Cannata F, Stefanini GG, Bolognese L. Myocardial ischemia and coronary disease in heart failure. Heart Fail Rev. Jan 22, 2020;25(1):53-65. [ CrossRef ] [ Medline ]
• Murphy SP, Ibrahim NE, Januzzi JL. Heart failure with reduced ejection fraction: a review. JAMA. Aug 04, 2020;324(5):488-504. [ CrossRef ] [ Medline ]
• Savarese G, Becher P, Lund L, Seferovic P, Rosano G, Coats A. Global burden of heart failure: a comprehensive and updated review of epidemiology. Cardiovasc Res. Jan 18, 2023;118(17):3272-3287. [ CrossRef ] [ Medline ]
• Ziaeian B, Fonarow GC. Epidemiology and aetiology of heart failure. Nat Rev Cardiol. Dec 2016;13(6):368-378. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Li F, Xin H, Zhang J, Fu M, Zhou J, Lian Z. Prediction model of in-hospital mortality in intensive care unit patients with heart failure: machine learning-based, retrospective analysis of the MIMIC-III database. BMJ Open. Jul 23, 2021;11(7):e044779. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Safavi KC, Dharmarajan K, Kim N, Strait KM, Li S, Chen SI, et al. Variation exists in rates of admission to intensive care units for heart failure patients across hospitals in the United States. Circulation. Feb 26, 2013;127(8):923-929. [ CrossRef ]
• van Diepen S, Bakal JA, Lin M, Kaul P, McAlister FA, Ezekowitz JA. Variation in critical care unit admission rates and outcomes for patients with acute coronary syndromes or heart failure among high‐ and low‐volume cardiac hospitals. JAHA. Mar 10, 2015;4(3):e001708. [ CrossRef ]
• Wunsch H, Angus DC, Harrison DA, Collange O, Fowler R, Hoste EAJ, et al. Variation in critical care services across North America and Western Europe*. Critical Care Medicine. 2008;36(10):2787-27e8. [ CrossRef ]
• Adams KF, Fonarow GC, Emerman CL, LeJemtel TH, Costanzo MR, Abraham WT, et al. ADHERE Scientific Advisory CommitteeInvestigators. Characteristics and outcomes of patients hospitalized for heart failure in the United States: rationale, design, and preliminary observations from the first 100,000 cases in the Acute Decompensated Heart Failure National Registry (ADHERE). Am Heart J. Feb 2005;149(2):209-216. [ CrossRef ] [ Medline ]
• Adler ED, Voors AA, Klein L, Macheret F, Braun OO, Urey MA, et al. Improving risk prediction in heart failure using machine learning. Eur J Heart Fail. Jan 2020;22(1):139-147. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Angraal S, Mortazavi BJ, Gupta A, Khera R, Ahmad T, Desai NR, et al. Machine learning prediction of mortality and hospitalization in heart failure with preserved ejection fraction. JACC Heart Fail. Jan 2020;8(1):12-21. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Soliman A, Agvall B, Etminani K, Hamed O, Lingman M. The price of explainability in machine learning models for 100-day readmission prediction in heart failure: retrospective, comparative, machine learning study. JMIR. Oct 27, 2023;25:e46934. [ FREE Full text ] [ CrossRef ] [ Medline ]
• White-Williams C, Rossi LP, Bittner VA, Driscoll A, Durant RW, Granger BB, et al. American Heart Association Council on CardiovascularStroke Nursing; Council on Clinical Cardiology;Council on EpidemiologyPrevention. Addressing social determinants of health in the care of patients with heart failure: a scientific statement from the American heart association. Circulation. Jun 02, 2020;141(22):e841-e863. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Abraham WT, Fonarow GC, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, et al. OPTIMIZE-HF InvestigatorsCoordinators. Predictors of in-hospital mortality in patients hospitalized for heart failure: insights from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF). J Am Coll Cardiol. Jul 29, 2008;52(5):347-356. [ FREE Full text ] [ CrossRef ] [ Medline ]
• McGilvray MM, Heaton J, Guo A, Masood MF, Cupps BP, Damiano M, et al. Electronic health record-based deep learning prediction of death or severe decompensation in heart failure patients. JACC Heart Fail. Sep 2022;10(9):637-647. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Li J, Liu S, Hu Y, Zhu L, Mao Y, Liu J. Predicting mortality in intensive care unit patients with heart failure using an interpretable machine learning model: retrospective cohort study. JMIR. Aug 09, 2022;24(8):e38082. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Kaya SI, Cetinkaya A, Ozcelikay G, Samanci SN, Ozkan SA. Approaches and challenges for biosensors for acute and chronic heart failure. Biosensors (Basel). Feb 16, 2023;13(2):282. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Hudson SR, Chan D, Ng LL. Change in plasma volume and prognosis in acute decompensated heart failure: an observational cohort study. J R Soc Med. Sep 08, 2016;109(9):337-346. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Goh KH, Wang L, Yeow AYK, Ding YY, Au LSY, Poh HMN, et al. Prediction of readmission in geriatric patients from clinical notes: retrospective text mining study. J Med Internet Res. Oct 19, 2021;23(10):e26486. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Clapp MA, Kim E, James KE, Perlis RH, Kaimal AJ, McCoy TH, et al. Comparison of natural language processing of clinical notes with a validated risk-stratification tool to predict severe maternal morbidity. JAMA Netw Open. Oct 03, 2022;5(10):e2234924. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Özyılmaz E, Özkan Kuşçu Ö, Karakoç E, Boz A, Orhan Tıraşçı G, Güzel R, et al. Worse pre-admission quality of life is a strong predictor of mortality in critically ill patients. Turk J Phys Med Rehabil. Mar 01, 2022;68(1):19-29. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Thapa NB, Nischay B, Sattar S, Sona T. Hospital readmission prediction using clinical admission notes. 2022. Presented at: ACSW '22: Proceedings of the 2022 Australasian Computer Science Week; February 14 - 18; Brisbane, Australia. [ CrossRef ]
• van Aken B, Papaioannou JM, Mayrdorfer M, Budde K, Gers F, Loeser A. Clinical outcome prediction from admission notes using self-supervised knowledge integration. arXiv. Preprint posted online on Feb 8, 2021. [ CrossRef ]
• Johnson AEW, Pollard TJ, Shen L, Lehman LH, Feng M, Ghassemi M, et al. MIMIC-III, a freely accessible critical care database. Sci Data. May 24, 2016;3:160035. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Johnson A, Stone D, Celi L, Pollard T. The MIMIC Code Repository: enabling reproducibility in critical care research. J Am Med Inform Assoc. Jan 01, 2018;25(1):32-39. [ CrossRef ] [ Medline ]
• Pollard TJ, Johnson AEW, Raffa JD, Celi LA, Mark RG, Badawi O. The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci Data. Sep 11, 2018;5(1):180178. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Devlin J, Chang MW, Lee K, Toutanova K. BERT: pre-training of deep bidirectional transformers for language understanding. arXiv. Preprint posted online on Oct 11, 2018. [ CrossRef ]
• Huang K, Altossar J, Ranganath R. ClinicalBERT: modeling clinical notes and predicting hospital readmission. arXiv. Preprint posted online on Apr 10, 2019. [ CrossRef ]
• Rahman W, Hasan MK, Lee S, Bagher Zadeh AA, Mao C, Morency LP, et al. Integrating multimodal information in large pretrained transformers. Proc Conf Assoc Comput Linguist Meet. Jul 2020;2020:2359-2369. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Sundararajan M, Taly A, Yan Q. Axiomatic attribution for deep networks. 2017. Presented at: Proceedings of the 34th International Conference on Machine Learning; August 6 - 11;3319-3328; Sydney, Australia. [ CrossRef ]
• Novakovsky G, Dexter N, Libbrecht MW, Wasserman WW, Mostafavi S. Obtaining genetics insights from deep learning via explainable artificial intelligence. Nat Rev Genet. Feb 03, 2023;24(2):125-137. [ CrossRef ] [ Medline ]
• Lyu W, Dong X, Wong R, Zheng S, Abell-Hart K, Wang F, et al. A multimodal transformer: fusing clinical notes with structured EHR data for interpretable in-hospital mortality prediction. AMIA Annu Symp Proc. 2022;2022:719-728. [ FREE Full text ] [ Medline ]
• Zhang Y, Tino P, Leonardis A, Tang K. A survey on neural network interpretability. IEEE Trans Emerg Top Comput Intell. Oct 2021;5(5):726-742. [ CrossRef ]

Abbreviations

Edited by G Eysenbach, T de Azevedo Cardoso; submitted 07.11.23; peer-reviewed by E Kawamoto, MO Khursheed; comments to author 05.12.23; revised version received 01.01.24; accepted 19.03.24; published 02.05.24.

©Zhenyue Gao, Xiaoli Liu, Yu Kang, Pan Hu, Xiu Zhang, Wei Yan, Muyang Yan, Pengming Yu, Qing Zhang, Wendong Xiao, Zhengbo Zhang. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 02.05.2024.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.

Princeton University

Princeton engineering, the science of static shock jolted into the 21st century.

By Scott Lyon

April 9, 2024

Static electricity has puzzled scientists for thousands of years. Above, water ions carry charge between two electrically insulating materials. The blue mesh represents the flow of charge that could be felt as a spark. Image courtesy of the researchers

Shuffling across the carpet to zap a friend may be the oldest trick in the book, but on a deep level that prank still mystifies scientists, even after thousands of years of study.

Now Princeton researchers have sparked new life into static. Using millions of hours of computational time to run detailed simulations, the researchers found a way to describe static charge atom-by-atom with the mathematics of heat and work. Their paper appeared in Nature Communications on March 23.

The study looked specifically at how charge moves between materials that do not allow the free flow of electrons, called insulating materials, such as vinyl and acrylic. The researchers said there is no established view on what mechanisms drive these jolts, despite the ubiquity of static: the crackle and pop of clothes pulled from a dryer, packing peanuts that cling to a box.

“We know it’s not electrons,” said Mike Webb , assistant professor of chemical and biological engineering , who led the study. “What is it?”

Webb first asked himself that question as a postdoctoral researcher at the University of Chicago. He puzzled over it with colleagues, baffled that such a common phenomenon could be so poorly understood. But the more they looked, the more insurmountable the questions became. “It just seemed out of reach,” he said.

It had been out of reach since Thales of Miletus first rubbed amber with fur and watched the amber (Greek: elektron ) collect feathers and dust — 26 centuries ago. Thales was one of the first people to explain nature through reason rather than supernatural forces. He played a critical role in the development of philosophy and eventually science. Despite the depth and breadth of knowledge accumulated over subsequent millennia, despite the myriad technologies born of that knowledge, science, in all that time, never cracked static. Maybe it never would.

At Princeton Webb got to talking to his colleague Sankaran Sundaresan , a leading expert in chemical reaction engineering who specializes in the flow of materials in gaseous chambers. In those environments, loaded with volatile chemicals, a stray spark could be deadly. Sundaresan had worked with static charge for decades, using reliable experimental data to predict but not fully fathom how charge moved in these systems.

“I treat that like a black box,” said Sundaresan, the Norman John Sollenberger Professor in Engineering. “We do some experiments and the experiments tell me: This is what happens. This is the charge.” He works down to the limit and carefully notes what he sees. What happens inside the black box remains a mystery.

One thing you find no matter where you look, though, according to Sundaresan, is trace amounts of water. Charged water molecules are everywhere, in nearly everything, clinging to virtually every surface on Earth. Even in extremely arid conditions, under intense heat, stray water ions pool into microscopic oases that harbor electrical charge.

Incidentally, Thales is best known not for his work on electricity but for an even grander project. He proposed that the entirety of nature was made of water, that water was the ur-substance, the essential stuff. It was the first attempt at a unified theory of everything. Aristotle wrote it all down.

Over the arc of Sundaresan’s career, he and his colleagues shrunk that black box so that the mysteries have been pushed ever deeper. But mysteries they remain.

The conversation between him and Webb led to a mutual realization: Sundaresan had decades of insight into data from reactors, and Webb could apply sophisticated atom-scale computational techniques to look at these water ions from the perspective of thermodynamics. How much energy would it take for a water ion to bolt from surface to surface? Maybe that would explain what was happening inside Sundaresan’s black box. The unresolved puzzle from Webb’s postdoc days came unlocked.

By modeling the relationship between charged water molecules and the amount of energy those molecules have available to propel them between surfaces, Webb and graduate student Hang Zhang demonstrated a very precise mathematical approximation of how electrical charge moves between two insulating materials.

In other words, they used math to simulate the movement of around 80,000 atoms. Those simulations matched real-life observations with a very high degree of precision. It turns out, in all likelihood, static shock is a function of water, and more specifically, the free energy of stray water ions. With that framework, Webb and Zhang revealed the molecular underpinnings of those familiar shocks in infinitesimal detail. They blew Sundaresan’s black box wide open. If only Thales could see.

The paper “Thermodynamic driving forces in contact electrification between polymeric materials” was published March 23 in the journal Nature Communications. Support for this work was provided by the Princeton Innovation Project X Fund and the U.S. Department of Energy. The simulations and analyses were performed using the resources of Princeton Research Computing.

Related News

Holographic displays offer a glimpse into an immersive future

Elke Weber receives BBVA Frontiers of Knowledge Award

Grad alum Avi Wigderson wins Turing Award for groundbreaking insights in computer science

Episode 4: Bethany Beardslee Winham and Chris Winham

Retro-reflectors could help future cities keep their cool

Episode 3: Haydn Seek

Michael A. Webb

Sankaran Sundaresan

Materials Science and Engineering

Related departments and centers.

Chemical and Biological Engineering

Princeton Materials Institute

An official website of the United States government

• Research @ BEA

Studies on the Value of Data

The U.S. Bureau of Economic Analysis has undertaken a series of studies that present methods for quantifying the value of simple data that can be differentiated from the complex data created by highly skilled workers that was studied in Calderón and Rassier 2022 . Preliminary studies in this series focus on tax data, individual credit data, and driving data. Additional examples include medical records, educational transcripts, business financial records, customer data, equipment maintenance histories, social media profiles, tourist maps, and many more. If new case studies under this topic are released, they will be added to the listing below.

• Capitalizing Data: Case Studies of Driving Records and Vehicle Insurance Claims | April 2024
• Private Funding of “Free” Data: A Theoretical Framework | April 2024
• Capitalizing Data: Case Studies of Tax Forms and Individual Credit Reports | June 2023

Rachel Soloveichik

JEL Code(s) E01 Published April 2024