scholarly articles higher education

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
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 13 October 2017

Rethinking higher education and its relationship with social inequalities: past knowledge, present state and future potential

• Theocharis Kromydas 1  

Palgrave Communications volume  3 , Article number:  1 ( 2017 ) Cite this article

128k Accesses

150 Citations

77 Altmetric

Metrics details

The purposes and impact of higher education on the economy and the broader society have been transformed through time in various ways. Higher education institutional and policy dynamics differ across time, but also between countries and political regimes and therefore context cannot be neglected. This article reviews the purpose of higher education and its institutional characteristics juxtaposing two, allegedly rival, conceptual frameworks; the instrumental and the intrinsic one. Various pedagogical traditions are critically reviewed and used as examples, which can potentially inform today’s policy making. Since, higher education cannot be seen as detached from all other lower levels of education appropriate conceptual links are offered throughout this article. Its significance lies on the organic synthesis of literature across social science, suggesting ways of going forward based on the traditions that already exist but seem underutilized so far because of overdependence in market-driven practices. This offers a new insight on how theories can inform policy making, through conceptual “bridging” and reconciliation. The debate on the purpose of higher education is placed under the context of the most recent developments of increasing social inequalities in the western world and its relation to the mass model of higher education and the relevant policy decisions for a continuous increase in participation. This article suggests that the current policy focus on labor market driven policies in higher education have led to an ever growing competition transforming this social institution to an ordinary market-place, where attainment and degrees are seen as a currency that can be converted to a labour market value. Education has become an instrument for economic progress moving away from its original role to provide context for human development. As a result, higher education becomes very expensive and even if policies are directed towards openness, in practice, just a few have the money to afford it. A shift toward a hybrid model, where the intrinsic purpose of higher education is equally acknowledged along with its instrumental purpose should be seen by policy makers as the way forward to create educational systems that are more inclusive and societies that are more knowledgeable and just.

Similar content being viewed by others

The impact of artificial intelligence on employment: the role of virtual agglomeration

Identity and inequality misperceptions, demographic determinants and efficacy of corrective measures

Participatory action research

Introduction.

The mainstream view in the western world, as informed by the human capital theory sees education, as an ordinary investment and the main reason why someone consumes time and money to undertake higher levels of education, is the high returns expected from the corresponding wage premium, when enters the labour market (Becker, 1964 , 1993 ). Nevertheless, things in practice are more complicated and this sequence of events is unlikely to be sustained, especially in recession periods like the one we currently live in. On the contrary, one notion of education, related somewhat to the American liberal arts tradition, is the intrinsic notion, which interprets that the purpose of education is to ‘equip people to make their own free, autonomous choices about the life they will lead’ (Bridges, 1992 : 92). There might be an economic basis underpinning this individual choice, but the intrinsic notion permits more subjective motivations, which are not necessarily affected by economic circumstances.

Robinson and Aronica ( 2009 ) argue that education, have become an impersonal linear process, a type of assembly line, similar to a factory production. They challenge this view and call for a less standardised pedagogy; more personalised to students needs as well as talents. Education is not similar to a manufacturing production-line, since students are highly concerned about the quality of education they receive as opposed to motor cars, which are indifferent to the process by which they are manufactured. Along these lines, Waters ( 2012 ), following Weber’s ( 1947 , 1968 ) rationale on the role of bureaucracy in modern societies, adds that this manufacturing process is achieved through rigid, rationalised and productively efficient but totally impersonal bureaucracy, operated in a way that sees children as raw materials for the creation of adults, which is the final product properly equipped to reproduce “itself” by being a parent to a new born “raw material” and so forth. Durkheim ( 1956 , 2006 ) sees this as a mechanism where adults exercise their influence over the younger in order to maintain the status quo they desire. However, since education entails ontological as well as epistemological implications, primary focus should be given to learning in such a way that educative and social functions could be amalgamated, rather than solely focusing on the delivery of existing knowledge per se, which becomes a reiterated process and an unchallenged absolute truth (Freire, 1970 ; Heidegger, 1988 ; Dall’ Alba and Barnacle, 2007 ).

This article focus on higher education; since it is the last stage before somebody enters the labour market and thus the instrumental view becomes more dominant over the intrinsic view, compared to the lower levels of education. Higher education, is being traditionally offered by universities. The first established university in Europe is the University of Bologna, where the term “academic freedom” was introduced as the kernel of its culture (Newman, 1996 ). Graham ( 2013 ) distinguishes between three different models of higher education. These are: the university college, the research and the technical university. He provides a historical review of the origins of these three models. The university college is the oldest one, where Christian values were the core values. Later on, when scientific knowledge questioned the universal theological truth, another type of university has been established, where research was the ultimate goal of the scholarship. This type of university has subsequently transformed by the introduction of the liberal arts tradition, flourished in the US. The research university model, originated circa 16 th century in Cambridge and established in Berlin by the introduction of the Humboldian University, shared a common aim: the pursuit of knowledge and its dissemination to the greater society. The third model of university is the technical one. It has been established in an industrial revolution context in Scotland and particularly in Glasgow in the premises of what is currently known as the University of Strathclyde. While the introduction of capitalism changed radically the structure and the format of labour relations, the technical model was based on the idea that industrial skills had to be acquired by formal education and somehow verified institutionally in order to be applied to the broader society. This is the first time where the up to then distinct fields of education and industry, started to be conceived as inextricably tight in a rather linear way.

These different models of higher education cultures and traditions still exist, but in reality, Universities worldwide follow a hybrid approach, where all traditions collaborate with each other. However, there are some universities that still carry the reputation and tradition of a specific model and to some extent this tradition differentiates them from all others. It is not the scope of this research to analyse this in detail, as the main aim is to offer an institutional and policy narrative, exploring the purpose of higher education and its relationship with social inequalities, focusing primarily on the western world.

Nowadays, in a rapidly changing word, the major debate is placed under the forms of institutional transformation of higher education. Brennan ( 2004 ), based on Trow ( 1979 , 2000 ), allocates three forms of higher education. The first one is the elite form, which main aim is to prepare and shape the mind-set of students originated from the most dominant class. The second is the mass form of higher education, which transmits the knowledge and skills acquired in higher education into the technical and economic roles students subsequently perform in the labour market. Lastly, the third is the universal form, which main purpose is to adapt students and the general population to the rapid social and technological changes.

This article reviews the contemporary trends in higher education and its widespread diffusion as interacted with the evolutions in western economies and societies, where social inequalities persist and even become wider (Dorling and Dorling, 2015 ). The narrative used in this article is more suitable to conceptualise higher education in a western world context, though we acknowledge that via globalisation, the way education and particularly higher education is delivered in the rest of the world seems to follow similar to the Western worlds paths, despite the apparent differences in culture, social and economic systems as well as writing systems. Footnote 1

An interdisciplinary and critical synthesis of the relevant literature is conducted, presenting two stances that are largely considered as rival: The instrumental one that treats higher education as an ordinary investment with particular financial yields in the labour market and the more intrinsic one which sees higher education as mainly detached from the logic of economic costs and benefits. The theoretical rivalry is apparent since in the former approach higher education is an inevitable property of labour market and thus an indispensable part of the mainstream economic neoliberal regime, whereas the latter sees no logical link between higher education and labour market purposes and therefore the content and substance of learning and knowledge acquisition in education and specifically in higher education should not be market-driven or aligned to the functions of specific economic regimes. However, this article argues that educational systems, and particularly their higher levels, are amalgamated parts of contemporary societies and therefore theories and practices need to move away from rather futile binary rationales.

The remainder of this paper explains why both the intrinsic and instrumental approaches are doomed to fail in practice when used in isolation. In a rapidly diverging and polarised world, where social inequalities rise within as well as between countries, common sense dictates social theories and practices to move towards reconciliation rather than stubborn rivalry. In that spirit, this paper argues that the intrinsic and instrumental approach are in fact complementary to each other. Such view can inform policy making towards building more inclusive educational systems; organically tight with the broader society. The narrative this article uses departs and expands on the rationale of eminent critical pedagogists such as Freire, Bronfenbrenner, Bourdieu and Kozol in order to challenge the current instrumental world-view of education, at least as this is apparent in the western world. Then the article moves into offering a reasoning for an organic synthesis of existing knowledge in order the two rival theories to be actualised in practice as a unified and reconciled pedagogical strategy. This reasoning builds on the research conducted by Durst’s ( 1999 ), Payne ( 1999 ) and Lu and Horner ( 2009 ). Durst ( 1999 ) suggests a “reflective instrumentalism”, where student’s pragmatic view that education is just a way of finding a well-paid job, operated in tandem with critical pedagogical canons, is indeed possible. Payne ( 1999 ) proposes a similar approach, where students are equipped with the necessary tools to find a job in the labour market; however educators should engage students with this knowledge in a critical way in order to be able to produce something new. Likewise Lu and Horner ( 2009 ) note that educators and students need to work together in such a way that perceptions of both are amenable to change and career choices are critically discussed in a constantly changing social context.

The purpose of higher education in western societies

Mokyr ( 2002 ) suggests that education should be integrated by both inculcation and emancipation in order to serve individual intellectual development as well as social progression. Shapiro ( 2005 ) emphasizes the need for the higher education institutions to serve a public purpose moving beyond narrow self-serving concerns, as well as to enforce social change in order to reflect the nature of a society that its members desire. More recently, in philosophical terms Barnett ( 2017 , p 10) calls for a wider conceptual landscape in higher education where “The task of an adequate philosophy of higher education…is not merely to understand the university or even to defend it but to change it”. )

The purpose of education and its meaning in the contemporary western societies has been also criticised by Bo ( 2009 ), suggesting that education has become a contradictory notion that leaves no space for emancipation since it gives no opportunity for improvisation to students. Thus, the students feel encaged within the system instead of being liberated. Bo agrees with Mokyr, who highlighted the need for recalling the basic notions of education from ancient philosophies: that education should be integrated by both inculcation and emancipation in order to serve individual intellectual development as well as social progression (Mokyr, 2002 ; Bo, 2009 ).

Not all individuals and societies agree on the purposes and roles of higher education in the modern world. However, in any case, it is a place where teaching and research can be accommodated in an organised fashion for the promotion of various types of knowledge, applied and non-applied. It is a place where money and moral values compete and collaborate simultaneously, where the development of labour market skills and competences coexist with the identification and utilisations of people’s skills and talents as well as the pursuit of employment, morality and citizenship.

The post-WWII era has been characterised by the mass model of higher education. Before this, higher education was for those belonging to higher social classes (Brennan, 2004 ). This model became the kernel of educational policies in Europe and generally, in the western world (Shapiro, 2005 ). Such policies have been boosted by the advent of Information and Communication Technologies (ICT), which enhance commercial and non-commercial bonds between countries and higher education institutions, transforming the role of higher education even further, making it rather universal (Jongbloed et al., 2008 ). Higher education’s boundaries have become vague and the predefined “social contract” between its institutions and those participated in them, is more complicated to be defined in absolute terms. Higher education institutions are now characterised by economic competition in a strict global market environment, where governments are not the key players anymore (Brennan, 2004 ).

Moreover, student demographics in higher education are constantly changing. Higher education is now an industry operating in a global market. Competition to attract talents from around the world is growing rapidly as an increasing number of countries offer additional graduate and post graduate positions to non-nationals, usually at a higher cost compared to nationals (Barber et al., 2013 ). Countries such as China or Singapore that are growing economically very rapidly are investing huge amounts of money to develop their higher education system and make it more friendly to talented people from around the world. The advent of new technologies have changed the traditional model of higher education, where physical presence is not a necessary requirement anymore (Yuan et al., 2013 ). Studying while working is much easier and therefore more mature students have now the opportunity to study towards a graduate or post-graduate degree. All these developments have increased the potential for profit; however it also requires huge amount of money to be invested in new technologies and all kinds of infrastructures and resources. The need for diversification in funding sources is simply essential and therefore all other industries become inevitably more engaged (Kaiser et al., 2014 ). On top of all these, climate change, the rise of terrorism, the prolonged economic uncertainty and the automazation of labour will likely increase cross-national and intraoccupational mobility and therefore the demand for higher education, especially in the recipient countries of the economically developed western world will inevitably rise. Summing up, higher education institutions operate under a very fluid and unpredictable environment and therefore approaches that are informed by adaptability and flexibility are absolutely crucial. The hybrid approach we propose where instrumental and intrinsic values are reconciled is along these lines.

Modern views of higher education place its function under a digital knowledge-based society, where economy dominates. Labour markets demand for skills such as technological competence and complex problem-solving by critical thinking and multitasking, which increases competition and in turn, accelerates the pace of the working day (Westerheijden et al., 2007 ). Haigh and Clifford ( 2011 ) argue that high competency, in both hard and soft skills, is not enough, as higher education needs to go deeper into changing attitudes and behaviours becoming the core of a globalised knowledge-based-economy. However, the trends of transferring knowledge and skills by universities, which “increasingly instrumentalize, professionalize, vocationalize, corporatize, and ultimately technologize education” (Thomson, 2001 : 244), have been extensively criticised in epistemological as well as in ontological terms (Bourdieu, 1998 ; Dall’ Alba and Barnacle, 2007 ). Livingstone ( 2009 ) argues that education and labour market have different philosophical departures and institutional principles to fulfill and therefore conceptualising them as concomitant economic events, with strong causal conjunctions, leads to logical fallacies. Livingstone sees the intrinsic purposes of education and contemporary labour market as rather contradictory than complimentary and any attempt to see them as the latter, leads to arbitrary and ambiguous outcomes, which in turn mislead rather than inform policy making. The current article, building on the arguments of Durst’s ( 1999 ), Payne ( 1999 ) and Lu and Horner ( 2009 ) challenges this view introducing a “bridging” rationale between the two theories, which can be also actualized in practice and inform policy making.

When education, and especially higher education, is considered as a public social right that everyone should have access to, human capital, as solely informed by the investment approach, cannot be seen as the most appropriate tool to explain the benefits an individual and society can gain from education. Citizenship can be regarded as one of these tools and perhaps concepts, such as the social and c ultural capital or habitus , which contrary to human capital acknowledge that students are not engaged with education just to succeed high returns in the labour market but apart from the economic capital, should be of equal importance when we try to offer a better explanation of the individuals’ drivers to undertake higher education. (Bourdieu, 1986 ; Coleman, 1988 ). Footnote 2 For example, Bourdieu ( 1984 ) thinks that certificates and diplomas are neither indications of academic or applied to the labour market knowledge, nor signals of competences but rather take the form of tacit criteria set by the ruling class to identify people from a particular social origin. Yet, Bourdieu does not disregard the human capital theory as invalid; however he remains very sceptical on its narrow social meaning as it becomes a property of ruling class and used as a mechanism to maintain their power and tacitly reproduce social inequalities.

Higher education attainment cannot be examined irrespectively of someone’s capabilities, as its conceptual framework presupposes a social construction of interacting and competing individuals, fulfilling a certain and, sometimes common to all, task each time. Capabilities, certainly, exist in and out of this context, as it includes both innate traits and acquired skills in a dynamic social environment. Sen ( 1993 : 30) defines capability as “a person’s ability to do valuable acts or reach valuable states of being; [it] represents the alternative combinations of things a person is able to do or be”. Moreover, Sen argues that capabilities should not be seen only as a means for succeeding a certain goal, but rather as an end itself (Sen, 1985 ; Saito, 2003 ; Walker and Unterhalter, 2007 ).

Capabilities are a prerequisite of well-being and therefore, social institutions should direct people into fulfilling this aim in order to feel satisfied with their lives. However, since satisfaction is commonly understood as a subjective concept, it cannot be implied that equal levels of life satisfaction, as these perceived by people of different demographic and socio-economic characteristics, mean social and economic equality. Usually, the sense of life satisfaction is relative to future expectations, aspirations and past empirical experiences, informed by the socio-economic circumstances people live in (Saito, 2003 ).

According to the capability approach, assessing the educational attainment of individuals or the quality of teachers and curriculum are not such useful tasks, if not complemented by the capacity of a learner to convert resources into capabilities. Sen’s ( 1985 , 1993 ) capability approach, challenges the human capital theory, which sees education as an ordinary investment undertaken by individuals. It also remains sceptical towards structuralist and post-structruralist approaches, which support the dominance of institutional settings and power over the individual acts. According to Sen ( 1985 , 1993 ), educational outcomes, as these are measured by student enrolments, their performance on tests or their expected future income, are very poor indicators for evaluating the overall purpose of education, related to human well-being. Moreover, the capability approach does not imply that education can only enhance peoples’ capabilities. It also implies that education, can be detrimental, imposing severe life-long disadvantages to individuals and societies, if delivered poorly (Unterhalter, 2003 , 2005 ).

From Sen’s writings, it is not clear whether the capability approach imply a distinction between instrumental and intrinsic values. Even if someone attempts an interpretation of the capability approach by arguing that it is only means that have an instrumental value, whereas ends only an intrinsic one, it is still unclear how can we draw a line between means and ends in a rather objective way. Escaping from this rather dualistic interpretation, a common-sense argument seems apparent: Capabilities have both intrinsic and instrumental value. Material resources can be obtained through people’s innate talents and acquired skills; however through the same resources transformed into capabilities a person who does not see this as an end but rather as a means, can also become a trusted member of the community and a good citizen, given that some kind of freedom of choice exists. Thus, resources apart from their instrumental value can also have an intrinsic one, with the caveat that the person chooses to conceive them as means towards a socially responsible end.

The American tradition in student development goes back to the liberal arts tradition, which main aim is to build a free person as an active member of a civic society. The essence of this tradition can be found in Nussbaum ( 1998 : 8)

“When we ask about the relationship of a liberal education to citizenship, we are asking a question with a long history in the Western philosophical tradition. We are drawing on Socrates’ concept of ‘the examined life,’ on Aristotle’s notions of reflective citizenship, and above all on Greek and Roman Stoic notions of an education that is ‘liberal’ in that it liberates the mind from bondage of habit and custom, producing people who can function with sensitivity and alertness as citizens of the whole world.”

Nowadays, liberal arts tradition is regarded as the delivery of interdisciplinary education across the social sciences but also beyond that, aiming to prepare students for the challenges they are facing both as professionals and as members of civic society. However, as Kozol notes in reality things are quite different (Kozol, 2005 , 2012 ). Kozol devoted much of his work examining the social context of schools in the US by focusing on the interrelationships that exist, maintained or transformed between students, teachers and parents. He points out that segregation and local disparities in the US schools are continuously increasing. The US schools and especially urban schools are seen as distinctive examples of institutions where social discrimination propagates while the US educational system currently functions as a mechanism of reproducing social inequality. Kozol is very critical on the instrumental purpose of market-driven education as this places businesses and commerce as the “key players”, since they shape the purpose, content and curriculum of education. At the same time, students, their parents as well as teachers, whose roles should have been essential, are displaced into some kind of token participants.

Hess ( 2004 ) might agree that US schools have become vehicles of increasing social inequalities but he suggest a very different to Kozol’s approach. Since schools are social institutions that operate and constantly interact with the rest of economy they have to become accountable in the way that ordinary business are, at least when it comes to basic knowledge delivery. Hess insists that all schools across the US should be able to deliver high quality basic knowledge and literacy. Such knowledge can be easily standardised and a national curriculum, equal and identical to all US school can be designed. By this, all schools are able to deliver high quality basic knowledge and all pupils, irrespective of their social background, would be able to receive it. Then, each school, teacher and pupil are held accountable for their performance and failure to meet the national standards should result in schools closed down, teachers laid off and pupils change school environment or even lose their chance to graduate. Hess distinguishes between two types of reformers; the status quo reformers who do not challenge the state control education and the common-sense reformers who are in favour of a non-bureaucratic educational system, governed by market competition, subjected to accountability measures similar to those used in the ordinary business world.

While Hess presents evidence that the problem in higher education is not underfunding but efficiency in spending, the argument he makes that schools can only reformed and flourish through the laws of market competition is not adequately backed up as there are plenty of examples in many industrial sectors, where the actual implementation of market competition instead of opening up opportunities for the more disadvantaged, has finally generated huge multinationals corporations, which operate in a rather monopolistic or at best oligopolistic environment, satisfying their own interests on the expense of the most deprived and disadvantaged members of the society. The ever growing increasing competition in the financial, pharmaceutical or IT software and hardware (Apple Microsoft, IOS and Android software etc.) sectors have not really helped the disadvantaged or the sector itself but rather created powerful “too big to fail” corporations that dominate the market if not own it.

Hess indeed believes that the US educational system apart from preparing students for the labour market has a social role to fulfil. When the purpose of higher education is solely labour market-oriented teaching and learning become inadequate to respond to the social needs of a well-functioned civic democracy, which requires active learners and critical thinkers who, apart from having a job and a profession, are able “ to frame and express their thoughts and participate in their local and national communities”(p. 4) . Creating rigorous standards for basic knowledge in all US schools is a goal that is sound and rather achievable. However, when such goals are based on a Darwinian like competition and coercion where only the fittest can survive they become rather inapplicable for satisfying the needs of human development, equity and sustainable social progress.

Bronfenbrenner’s ecological systems theory ( 1979 , 2005 , 2009 ) (subsequently named from Bronfenbrenner himself as bioecological systems theory) is also an example of schools as organic ingredients of a single concentric system that includes four sub systems; the micro, the meso, the exo and the macro as well as the chronosystem that refers to the change of the other four through time. The Micro system involves activities and roles that are experienced through interpersonal relationships such as the family, schools, religious or social institutions or any interactions with peers. The meso system includes the relationships developed between the various microsystem components, such as the relationship between school and workplace or family and schools. The exosystem comprises various interactions between systems that the person who is in the process of development does not directly participates but influence the way microsystems function and impact on the person. Some examples of exosystems are the relationships between family and peers of the developing person, family and schools, etc. The macrosystem incorporates all these things that can be considered as cultural environment and social context in which the developing person lives. Finally, the chronosystem introduces a time dimension, which encompasses all other sub-systems, subjecting them to the changes occurred through time. All these systems constantly interact, shaping a dynamic, complex but also natural ecological environment, in which a person develops its understanding of the world. In practical terms, this theory has found application in Finland, gradually transforming the Finish educational system to such a degree that is now considered the best all over the world (Määttä and Uusiautti, 2014 ; Takala et al., 2015 ). Finally, Bronfenbrenner is also an advocate that poverty and social inequalities are developed not because of differences in individual characteristics and capabilities but because of institutional constraints that are insurmountable to those from a lower socio-economic background.

Freire ( 1970 , 2009 ) criticizes the way schooling is delivered in contemporary societies. The term he uses to describe the current state of education is “banking education”, where teachers and students have very discrete roles with the former to be perceived as depositors of knowledge and the latter as depositories. This approach sees the knowledge acquired within the institutional premises of formal education as an absolute truth, where reality is perceived as something static aiming to preserve the status quo in education and in turn in society and satisfy the interests of the elite. This actual power play means that those who hold knowledge and accept its acquiring procedure as static, become the oppressors whereas those who either lack knowledge or even hold it but challenge it in order to transform it, the oppressed. From the one side the oppressors achieve to maintain their dominance over the oppressed and on the other side the oppressed accept their inferior role as an unchallenged normality where their destiny is predetermined and can never be transformed. Therefore, through this distinction of social roles, social inequalities are maintained and even intensified through time. Freire sees the “banking education” approach as a historical hubris since social reality is a process of constant transformation and hence, it is by definition dynamic and non-static. What we actually know today cannot determine our future social roles, neither can prohibit individuals from challenging and transforming it into something new (Freire, 1970 ; Giroux, 1983 ; Darder, 2003 ).

The banking education approach resembles very much the ethos of the human capital theory, where individuals utilise educational attainment as an investment instrument for succeeding higher wages in the future and also climb the levels of social hierarchy. The assumption of linearity between past individual actions and future economic and social outcomes is at the core of banking education and thus human capital theory. However, this assumption introduces a serious logical fallacy that surprisingly policy makers seem to value very little nowadays, at least in the Western societies. Freire ( 2009 ) apart from criticizing the current state of education argues that a pedagogical approach that “demythologize” and unveils reality by promoting dialogue between teachers and students create critical thinkers, who are engaged in inquiry in order to create social reality by constantly transforming it. This is the process of problem-posing education , which aligns its meaning with the intrinsic view of education that regards human development as mainly detached from the acquisition of material objects and accumulation of wealth through increased levels of educational attainment.

Originated in Germany, the term Bildung —at least as this was interpreted from 18 th century onwards, after Middle Ages era where everything was explained in the prism of a strict and theocratic society- shaped the philosophy by which the German educational system has been functioning even until nowadays (Waters, 2016 ). Bildung aims to provide the individual education with the appropriate context, through which can reach high levels of professional development as well as citizenship. It is a term strongly associated with the liberation of mind from superstition and social stereotypes. Education is assumed to have philosophical underpinnings but it needs, as philosophy itself as a whole does too, to be of some practical use and therefore some context needs to be provided Footnote 3 (Herder, 2002 ).

For Goethe ( 2006 ) Bildung , is a self-realisation process that the individual undertakes under a specific context, which aims to inculcate altruism where individual actions are consider benevolent only if they are able to serve the general society. Although Bildung tradition, from the one hand, assumes that educational process should be contextualised, it approach context as something fluid that is constantly changing. Therefore, it sees education as an interactive and dynamic process, where roles are predetermined; however at the same time they are also amenable to constant transformation (Hegel, 1977 ). Consequently, this means that Bildung tradition is more closely to what Freire calls problem-posing education and therefore to the intrinsic notion of education. Weber ( 1968 ), looked on the Bildung tradition as a means to educate scientists to be involved in policy making and overcome the problems of ineffective bureaucracy. Waters ( 2016 ) based on his experiences with teaching in German higher education argue that the Bildung tradition is still apparent today in the educational system in Germany.

However, higher education, as an institution, involves students, teachers, administrators, policy makers, workers, businessmen, marketers and generally, individuals with various social roles, different demographic characteristics and even different socio-economic backgrounds. It comes natural that their interests can be conflicting and thus, they perceive the purpose of higher education differently.

Higher education expansion and social inequalities: contemporary trends

Higher education enrolment rates have been continuously rising for the last 30 years. In Europe, and especially in the Anglo-Saxon world, policies are directed towards widening the access to higher education to a broader population (Bowl, 2012 ). However, it is very difficult for policy-makers to design a framework towards openness in higher education, mainly due to the heterogeneity of the population the policies are targeted upon. Such population includes individuals from various socio-economic, demographic, ethnic, innate ability, talent orientation or disability groups, as well as people with very different social commitments and therefore the vested interests of each group contradict each other, rendering policy-making an extremely complicated task (CFE and Edge Hill University, 2013 ).

A collection of essays, edited by Giroux and Myrsiades ( 2001 ), provided valuable insights to the humanities and social sciences literature regarding the notion of corporate university and its implications to society’s structure. As Williams ( 2001 : 18) notes in one of this essays:

“Universities are now being conscripted directly as training grounds for the corporate workforce…university work has been more directly construed to serve not only corporate-profit agendas via its grant-supplicant status, but universities have become franchises in their own right, reconfigured to corporate management, labor, and consumer models and delivering a name-brand product”.

Chang et al. ( 2013 ) argues that institutional purposes do not always coincide with the expectations students have from their studies. In most cases, students hold a more pragmatic and instrumental understanding towards the purpose of higher education, primarily aiming for a better-paid and high quality jobs.

Arum and Roksa ( 2011 ) claim that students during their studies in higher education make no real progress in critical thinking and complex problem-solving. Nonetheless, it is notable that those who state that they seek some “deeper meaning” in higher education, looking at a broader picture of things, tend to perform better than those who see university through instrumental lenses (Entwistle and Peterson, 2004 ). These findings question the validity of the instrumental view in higher education as it seems that those that are intrinsically motivated to attend higher education, end up performing much better in higher education and also later on in the labour market. Therefore, in practice, the theoretical rivalry between the intrinsic and instrumental approach operate in a rather dialectic manner, where interactions between social actors move towards a convergence, despite the focus given by policy makers on the instrumental view.

Bourdieu ( 1984 , 1986 , 1998 , 2000 ) based on his radical democratic politics, argued that education inequalities are just a transformation of social inequalities and a way of reproduction of social status quo. Aronowitz ( 2004 ) acknowledged that the main function of public education in the US is to prepare students to meet the changes, occurred in contemporary workplaces. Even if this instrumental model involves the broad expansion of educational attainment, it also fails to alleviate class-based inequalities. He is in line with Bourdieu’s argument that social class relations are reproduced through schooling, as schools reinforce, rather than reduce, class-based inequalities. More recently, similar findings from various countries are very common in the literature (Chapman et al., 2011 ; Stephens et al., 2015 )

Apple ( 2001 ) argues that despite neoliberalism’s claims that privatisation, marketization, harmonisation and generally the globalisation of educational systems increase the quality of education, there are considerable findings in numerous studies that show that the expansion of higher education happens in tandem with the increase of income inequality and the aggravation of racial, gender and class differences. Gouthro ( 2002 ) argues that there has been a misrepresentation of the basic notions that characterise the purpose of education, such as critical thinking, justice and equity. Ganding and Apple ( 2002 ) went one step further by suggesting an alternative solution, which lies on the decentralisation of educational systems, using the “Citizen School” as an example of an educational institution, which prioritises quality in education and its provision to impoverished people. Finally, they call for a radical structural reform on educational systems worldwide, where the relationship between various social communities and the state is based on social justice and not on power.

Brown and Lauder ( 2006 ) investigated the impact of the fundamental changes on education, as related to the influence that various socio-economic and cultural factors have on policy making. Remaining sceptical against the empirical validity of human capital theory, they conclude that it cannot be guaranteed that graduates will secure employment and higher wages. Contrary to Card and Lemieux’s ( 2001 ) findings, the authors argue that when the wage-premium is not measured by averages, but is split in deciles within graduates, it is only the high-earning graduates that have experienced an increasing wage-gap during this period. Increasing incidences of over-education, due to an ever-increasing supply of graduates compared to the relatively modest growth rates of high-skilled jobs, have also been observed. Any differences in pay, between graduates and non-graduates, can be ascribed more to the stagnation of non-graduates' pay, rather than to graduates’ additional pay, because of their higher educational attainment. More recently, Mettler ( 2014 ) argues that the focus on corporate interests in policy making in the US has transformed higher education into a caste system that reproduces and also intensifies social inequalities.

There are evidence, which illustrate that families play a distinctive role in encouraging children’s abilities and traits through a warm and friendly family environment. As higher education requires a significant amount of money to be invested, families with high-income have more chances and means to promote their children’s abilities and traits as well as their career prospects, when compared with the low-income ones. Certainly, there are other factors, which can affect children’s prospects, but the advantage in favour of high-income families is relatively apparent in the empirical literature (Solon, 1999 ).

Livingstone and Stowe ( 2007 ), based on the General Social Survey (GSS), conducted an empirical study on the school completion rates partitioning individuals into family and class origin, residential area as well as race and gender. They focused on the relatively low completion rates of low-class individuals, from the inner city and rural areas of the US. Their findings reveal that working-class children are being discriminated on their school completion rates, compared with the mid- and high-class children. Race and gender discrimination has been detected in rural areas but not in inner cities and suburb areas, where the completion rates are more balanced.

Stone ( 2013 ), finally sees things from a very different perspective, where inequalities exist mainly because of simply bad luck. He argues in favour of lots, when a university has to decide whether to accept an applicant or not. Even if, an argument like this seems highly controversial, it consists of something that has been implemented in many countries, several times in the past (Hyland, 2011 ). The argument that an individual deserves a place in university just because he/she scored higher marks in a standardised sorting examination test does not prove that he/she will perform better in his/her subsequent academic tasks. Likewise, if an individual, who failed to secure a place in university due to low marks, was given a chance to enter university through a different procedure, he/she might have performed exceptionally well. Yet, human society cannot solely depend on lotteries and computer random algorithms, but sometimes, up to a certain point and in the name of fairness and transparency, there is a strong case for also looking on the merits for using one (Stone, 2013 ).

Furthermore, Lowe ( 2000 ) argued that the widening of higher education participation can create a hyper-inflation of credentials, causing their serious devaluation in the labour market. This relates to the concept of diploma disease, where labour markets create a false impression that a higher degree is a prerequisite for a job and therefore, induce individuals to undertake them only for the sake of getting a job (Dore, 1976 ; Collins, 1979 ). This situation can create a highly competitive credential market, and even if there are indications of higher education expansion, individuals from lower social class do not have equal opportunities to get a degree, which can lead them to a more prestigious occupational category. This is, in turn, very similar to the Weberian theory of educational credentialism, where credentials determine social stratum (Brown, 2003 ; Karabel, 2006 ; Douthat, 2005 ; Waters, 2012 ).

The concept of credential inflation has been extensively debated from many scholars, who question the role of formal education and the usefulness of the acquisition of skills within universities (Dore 1997 ; Collins, 1979 ; Walters, 2004 ; Hayes and Wynard, 2006 ). Evans et al. ( 2004 ) focuses on the tacit skills, which cannot be acquired by formal learning, mainly obtained by work and life experience as well as informal learning. These skills are competences related to the way a complex situation could be best approached or resemble to personal traits, which can be used for handling unforeseen situations.

Policy implications

Higher educational attainment that leads to a specific academic degree is a dynamic procedure, but with a pre-defined end. This renders the knowledge acquired there, as obsolete. Policies, such as Bologna Declaration supports an agenda, where graduates should be further encouraged to engage with on-the-job training and life-long education programmes (Coffield, 1999 ). Other scholars argue that institutions should have a broader role, acknowledging the benefits that higher educational attainment bring to societies as a whole by the simultaneous promotion of productivity, innovation and democratisation as well as the mitigation of social inequalities (Harvey, 2000 ; Hayward and James, 2004 ). Boosting employability for graduates is crucial and many international organisations are working towards the establishment of a framework, which can ensure that higher education satisfies this aim (Diamond et al., 2011 ). Yet, this can have negative side-effects making the employability gap between high- and low-skilled even wider, since there is no any policy framework specifically designed for low-skilled non-graduates on a similar to Bologna Declaration, supranational context. Heinze and Knill ( 2008 ) argue that convergence in higher education policy-making, as a result of the Bologna Process, depends on a combination of cultural, institutional and socio-economic national characteristics. Even if, it can be assumed that more equal countries, in terms of these characteristics, can converge much easier, it is still questionable if and how much national policy developments have been affected by the Bologna Declaration.

However, the political narrative of equal opportunities in terms of higher education participation rates does not seem very convincing (Brown and Hesketh, 2004 ; The Milburn Commission, 2009 ). It appears that a consensus has been reached in the relevant literature that there is a bias towards graduates from the higher social classes, but it has been gradually decreasing since 1960 (Bekhradnia, 2003 ; Tight, 2012 ). Nonetheless, despite the fact that, during the last few decades, there has been an improvement in the participation rates for the most vulnerable groups, such as women and ethnic minorities, the inequality is still obvious in some occasions (Greenbank and Hepworth, 2008 ). Machin and Van Reenen ( 1998 ) trace the causes of the under-participation in an intergenerational context, arguing that the positive relationship between parental income and participation rates is apparent even from the secondary school. Likewise, Gorard ( 2008 ) identifies underrepresentation on the previous poor school performance, which leads to early drop-outs in the secondary education, or into poor grades, which do not allow for a place in higher education. Other researchers argue that paradoxically, educational inequality persists even nowadays, albeit the policy orientation worldwide towards the widening of higher education participation across all social classes (Burke, 2012 ; Bathmaker et al., 2013 ).

There are different aspects on the purpose of higher education, which particularly, under the context of the ongoing economic uncertainty, gain some recognition and greater respect from academics and policy-makers. Lorenz ( 2006 ) notes that the employability agenda, which is constantly promoted within higher education institutions lately, cannot stand as a sustainable rationale in a diverse global environment. This harmonisation and standardisation of higher education creates permanent winners and losers, centralising all the gains, monetary and non-monetary, towards the most dominant countries, particularly towards Anglo-phone countries and specific industries and therefore social inequalities increase between as well as within countries. Some scholars call this phenomenon as Englishization (Coleman, 2006 ; Phillipson, 2009 ).

Tomusk ( 2002 , 2004 ) positioned education within the general framework of the recent institutional changes and the rapid rise of the short-term profits of the financial global capital. Specifically, the author sees World Bank as a transnational organisation. Given this, any loan agreement planned from the World Bank regarding higher education reforms in developing countries, has the same ultimate, but tacit, goal, which is the continuous rise of the national debt and in turn, the vitiation of national fiscal and monetary policies, in order the human resources of the so called “recipient countries”, to be redistributed in favour of a transnational dominant class.

Hunter ( 2013 ) places the debate under a broader political framework, juxtaposing neo-liberalism with the trends formulated by the OECD. She concludes that OECD is a very complex and multi-vocal organisation and when it comes to higher education policy suggestions, there is not any clear trend, especially towards neo-liberalism. This does not mean that economic thinking is not dominant within the OECD. This is, in fact, OECD’s main concern and it is clear to all. Hunter ( 2013 : 15–16) accordingly states that:

“Some may feel offended by the vocational and economic foci in OECD discourse. Many would like to see HE held up for “higher” ideals. However, it is fair for OECD to be concerned with economics. They do not deny that they are primarily an organization concerned with economics. It is up to us, the readers, politicians, scholars, voters, teachers, administrators, and policy makers, to be aware that this is an economic organization and be careful of from whom we get our assumptions”.

Hyslop-Margison ( 2000 ) investigated how the market economy affects higher education in Canada, when international organisations and Canadian business interfere in higher education policy making, under the support of government agencies. He argues that such economy-oriented policies deteriorate curriculum theory and development.

Letizia ( 2013 ) criticises market-oriented reforms, enacted by The Virginia Higher Education Opportunity Act of 2011, placing them within the context of market-driven policies informed by neoliberalism, where social institutions, such as higher education, should be governed by the law of free market. According to Letizia, this will have very negative implications to the humanistic character of education, affecting people’s intellectual and critical thinking, while perpetuating social inequalities.

The term Mcdonaldisation has been also used recently to capture functional similarities and trends in common, between higher education and ordinary commercial businesses. Thus, efficiency, calculability, predictability and maximisation are high priorities in the American and British educational systems and because of their global influence, these characteristics are being expanding worldwide (Hayes and Wynard, 2006 ; Garland, 2008 ; Ritzer, 2010 ).

The notion of Mcdonaldisation is very well explained by Garland ( 2008 , no pagination):

“Mcdonaldisation can be seen as the tendency toward hyper-rationalisation of these same processes, in which each and every task is broken down into its most finite part, and over which the individual performing it has little or no control becoming all by interchangeable. It may be argued that the labour processes involved in advanced technological capitalism increasingly depend on either the handling and processing of information, or provision of services requiring instrumentalised forms of communication and interaction, just as the same “professional” roles frequently consist of largely mechanized, functional tasks requiring a minimum of individual input or initiative, let alone creative or critical thought, a process illustrated in blackly comic by the 1999 film Office Space”.

Realistically, higher education cannot be solely conceptualised by the human capital approach and similar quantitative interpretations, as it has cultural, psychological, idiosyncratic and social implications. Additionally, Hoxby ( 1996 ) argued that policy environment and systems of governance in higher education play a significant role to an individuals’ decision-making process to obtain further education and unfortunately, policy makers regard this aspect as static that can never be transformed.

Lepori and Bonaccorsi ( 2013 ), following Latour and Woolgar’s ( 1979 ) rationale of the high importance of vested interest in scientific endeavours, argue that higher education trends are too complex to be reduced and captured adequately, by the use of economic indicators as related to the labour market. However, the market and money value of higher education should not be neglected, especially in developing countries, as there is evidence that it can help people escape the vicious cycle of poverty and therefore it has a practical and more pragmatic purpose to fulfil (Psacharopoulos and Patrinos, 2004 ). According to World Bank ( 2013 ), education can contribute to a significant decrease of the number of poor people globally and increase social mobility when it manages to provides greater opportunities for children coming from poor families. There are also other studies that do not only focus to strict economic factors, but also to the contribution of educational attainment to fertility and mortality rates as well as to the level of health and the creation of more responsible and participative citizens, bolstering democracy and social justice (Council of Europe, 2004 ; Osler and Starkey, 2006 ; Cogan and Derricott, 2014 ).

Mountford-Zimdars and Sabbagh ( 2013 ), analysing the British Social Attitudes (BSA) survey, offer a plausible explanation on why the widening of participation in higher education is not that easy to be implemented politically, in the contemporary western democracies. The majority of the people, who have benefited from higher educational attainment in monetary and non-monetary terms, are reluctant to support the openness of higher education to a broader population. On the contrary, those that did not succeed or never tried to secure a place in a higher education institute, are very supportive of this idea. This clash of interests creates a political perplexity, making the process of policy-making rather dubious. Therefore, the apparent paradox of the increase in higher educational attainment, along with a stable rate in educational inequalities, does not seem that strange when vested interests of certain groups are taken into account.

Moreover, the decision for someone to undertake higher education is not solely influenced by its added value in the labour market. Since an individual is exposed to different experiences and influences, strategic decisions can easily change, especially when these are taken from adolescents or individuals in their early stages of their adulthood. Given this, perceptions and preferences do change with ageing and this is why there are some individuals who drop out from university, others who choose radical shifts in their career or others who return to education after having worked in the labour market for many years and in different types of jobs.

Higher education has expanded rapidly after WWII. The advent of new technologies dictates the enhancement of people’s talents and skills and the creation of a knowledge-based-economy, which in turn, demands for even more high-skilled workers. Policy aims for higher education in the western world is undoubtedly focusing on its diffusion to a broader population. This expansion is seen as a policy instrument to alleviate social and income inequalities. However, the implementation of such policies has been proved extremely difficult in practise, mainly because of existent conflicted interests between groups of people, but also because of its institutional incapacity to target the most vulnerable. Nonetheless, it has been observed a constant marketization process in higher education, making it less accessible to people from poor economic background. Concerns on the persistence of policy-makers to focus primarily on the economic values of higher education have been increasingly expressed, as strict economic reasoning in higher education contradicts with political claims for its continuing expansion.

On the other hand, there are studies arguing that the instrumental model can make the transition of graduates into the labour market smoother. Such studies are placed under the mainstream economics framework and are also informed by policy decisions implemented by the Bologna Process, where competitiveness, harmonisation and employability are the main policy axes. The Bologna Process and various other institutions (e.g., the EU, World Bank, OECD) have provided a framework under which higher education can be seen as inextricably linked with labour market dynamics; however, the intrinsic notion of higher education is treated more as a nuisance and less as a vital component on this framework. Nevertheless, this makes the job competition between graduates much more intense and also creates very negative implications for those that remain with low qualifications as they effectively become socially and economically marginalised.

The purpose of higher education and its role in modern societies remains a heated philosophical debate, with strong practical and policy implications. This article sheds more light to this debate by presenting a synthetic narrative of the relevant literature, which can be used as a basis for future theoretical and empirical research in understanding contemporary trends in higher education as interwoven with the evolutions in the broader socio-economic sphere. Specifically, two conflicting theoretical stances have been discussed. The mainstream view primarily aims to assist individuals to increase their income and their relative position in the labour market. On the other hand, the intrinsic notion focus on understanding its purpose under ontological and epistemological considerations. Under this conceptual framework, the enhancement of individual creativity and emancipation are in conflict with the contemporary institutional settings related to power, dominance and economic reasoning. This conflict can influence people’s perceptions on the purpose of higher education, which can in turn perpetuate or otherwise revolutionise social relations and roles.

However, even if the two theoretical stances presented are regarded as contradictory, this article argues that, in practical terms, they can be better seen as complementing each other. From one hand, using an instrumental perspective, an increase in higher education participation, focusing particularly on the most vulnerable and deprived members of society, can alleviate problems of income and social inequalities. The instrumental view of education has a very important role to play if focused on lower-income social classes, as it can become the mechanism towards the alleviation of income inequalities. On the other hand, apart from the pecuniary, there are also other non-pecuniary benefits associated with this, such as the improvement in the fertility and mortality and general health level rates or the boost of active democracy and citizenship even within workplaces and therefore a shift of higher education towards its intrinsic purposes is also needed. (Bowles and Gintis, 2002 ; Council of Europe, 2004 ; Brennan, 2004 ; Brown and Lauder, 2006 ; Wolff and Barsamian, 2012 ).

Summing up, education is not a simply just another market process. It is not just an institution that supply graduates as products that have some predetermined value in the labour market. Consequently, acquired knowledge in education verified by college degrees is neither a necessary nor a sufficient condition for the labour market to create appropriate jobs, where graduates utilise and expand this knowledge. In fact, the increasing costs of higher education, mostly due to its internationalisation, and the rising levels of job mismatch create a rather gloomy picture of the current economic environment, which seems to preserve the well-paid jobs mostly to those from a certain socio-economic class background. At the same time, poor students are vastly disadvantaged to more wealthy ones, considering the huge differences in terms of higher as well as their past education, their parent’s education and also certain elitist traditions that work towards perpetuating power relations in favour of the dominant class.

As Castoriadis ( 1997 ) notes, it is impossible to separate education from its social context. We, as human beings, acquire knowledge, in the sense of what Castoriadis calls paideia , from the day we born until the day we die. We are being constantly developed and transformed along with the social transformations that happen around us. The transformation on the individual is in constant interaction with social transformations, where no cause and effect exists. Formal schooling has become nowadays an apathetic task where no real engagement with learning happens, while its major components such as educators, families and students are largely disconnected with each other. Educators, cynically execute the teaching task that a curriculum dictates each time, families’ main concern is to attach a market value to their children educational attainment, “labelling” them with a credential that the labour market allegedly desires, while students pay attention to anything else apart from the knowledge they get per se and therefore they care too little for its quality and also its practical use.

To tackle the ever-growing social inequalities due to the narrow economic policy making in education, we need a radical shift towards policies that are informed from Freire’s problem-posing education and Sen’s capabilities approach, get insights in terms of structure from Bronfenbrenner’s bioecological systems theory, while giving context according to the Bildung tradition also acknowledging that education, apart from instrument, is a vehicle towards liberation, cultural realisation as well as social transformation. In practical terms, real-world examples from Finland or Germany can be used, which policy makers from around the world should start paying more attention to, moving away from narrow and sterile instrumentalism that has spectacularly failed to tackle social inequalities.

In the context of a modern world where monetary costs and benefits are the basis of policy arguments, a massification and broader diffusion of higher education to a much broader population implies marketisation and commercialisation of its purpose and in turn its inclusion on an economy-oriented model where knowledge, skills, curriculum and academic credentials inevitably presuppose a money-value and have a financial purpose to fulfil. The policy trends towards an economy-based-knowledge, through a strict instrumental reasoning, rather than the alleged knowledge-based-economy seems to persist and prevail, albeit its poor performance on alleviating income and social inequalities. Yet, in a global context of a prolonged economic stagnation and a continuous deterioration of society’s democratic reflexes, a shift towards a model, where knowledge is not subdued to economic reasoning, can inform a new societal paradigm of a genuine knowledge-based-economy, where economy would become a means rather than an ultimate goal for human development and social progress.

Data availability

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

For example, Confucian tradition is very rich, when it comes to education and human development. It is indeed very interesting to see how the basic principles of Confucian education, such as humanism, harmony and hierarchy, has been transformed through time and especially after the change in China’s economic model by Den Xiaoping’s reforms towards a more open economic system and along this a more business-oriented and globalised educational system. Perhaps the Chinese tradition in education, which mainly regards education as a route to social status and material success based on merit and constant examination can explain why the human capital theory is more applicable. On the other hand, additional notions in the Confucian tradition that education should be open to all, irrespective of the social class each person belongs to (apart perhaps from women and servants that were rather considered as human beings with limited social rights), its focus on ethics and its purpose to prepare efficient and loyal practitioners for the government introduces an apparent paradox with human capital theory but not necessarily with the instrumental view of education. This contradiction deserves to be appropriately and thoroughly examined in a separate analysis before it is contrasted to the Western tradition. For this reason the current research focuses only on the Western world leaving the comparison analysis with educational traditions found around the world, among them the Confucian tradition, as a task that will be conducted in the near future.

The use of capital in Bourdieu is criticised by a stream of social science scholars as rather promiscuous and unfortunate (Goldthorpe, 2007 ). They argue that a paradox here is apparent as in English linguistic etymological terms, the word capital implies, if not presupposes market activity. The same time Bourdieu criticises Becker’s human capital tradition as solely market-driven and a tacit way where the ruling class maintain their power through universities and other institutions. Waters ( 2012 ) argue that the use of the term “capital” in both Becker’s and Bourdieu’s writings is unfortunate, while both use the term to mean different things. Bourdieu’s understanding on the nature of “habitus” is a much more applicable term to explain the social role of education systems. Habitus is not capital, even if there is constant interaction between the two. Becker on the other hand, seem to neglect social and cultural capital as well as Bourdieu’s notion of habitus, which in turn is about the reproduction of society and power relations by universities and other institutions.

Some might have valid ontological objections on this, in terms of the purpose of philosophy as a whole; however the concept of Bildung has given education a role within society that moves away from individualism and the constant pursuit of material objects as ultimate means of well-being.

Apple M (2001) Comparing neo-liberal projects and inequality in education. Comp Educ 37(4):409–423

Article   Google Scholar  

Aronowitz S (2004) Against schooling: Education and social class. Soc Text 22(2):13–35

Arum R, Roksa J (2011) Academically adrift: Limited learning on college campuses. University of Chicago Press, Chicago, IL

Google Scholar  

Barber M, Donnelly K, Rizvi S, Summers L (2013) An avalanche is coming: Higher education and the revolution ahead. Institute for Public Policy Research, London, p 78

Barnett R (2017) Constructing the university: Towards a social philosophy of higher education. Educ Philos Theory 49(1):78–88

Bathmaker AM, Ingram N, Waller R (2013) Higher education, social class and the mobilisation of capitals: Recognising and playing the game. Br J Sociol Educ 34(5-6):723–743

Becker GS (1964) Human capital: A theoretical and empirical analysis with special reference to education. National Bureau of Economic Research, New York

Becker GS (1993) Human capital: A theoretical and empirical analysis with special reference to education, 3rd edn.. National Bureau of Economic Research, Chicago

Book   Google Scholar  

Bekhradnia B (2003) Widening participation and fair access: An overview of the evidence. Higher Education Policy Institute, London

Bo W (2009) Education as both Inculcation and emancipation [Online] Institute of Social Economy and Culture 1-6, Peking University Education Review. Available at http://www2.hawaii.edu/~pesaconf/zpdfs/30bo.pdf

Bourdieu P (1984) Distinction: A social critique of the judgement of taste. Cambridge, MA: Harvard university press.

Bourdieu P (1986) The forms of capital. In: Richardson J (ed) Handbook of Theory and Research for the Sociology of Education. Greenwood, Westport, CT, p 241–258

Bourdieu P (1998) Practical reason: On the theory of action. Stanford University Press, Stanford, CA.

Bourdieu P (2000) Participant objectivation: Breaching the boundary between anthropology and sociology- How? Lecture on the occasion of the presentation of the Huxley Memorial Medal for 2000. Royal Anthropological Institute, London, 6 Dec

Bowl M (2012) The contribution of further education and sixth-form colleges to widening participation [Online] York: Higher Education Academy. Available at: http://www.heacademy.ac.uk/resources/detail/WP_syntheses/Bowl

Bowles S, Gintis H (2002) Schooling in capitalist America revisited. Sociol Educ 75(1):1–18

Brennan J (ed) (2004) The social role of the contemporary university: contradictions, boundaries and change. In: TenYears on: changing education in a changing world. Center for Higher Education Research and Information, Open University Press, Maidenhead, UK, p 23–54

Bridges D (1992) Enterprise and liberal education. J Philos Educ 26(1):91–98

Bronfenbrenner U (1979) The ecology of human development: experiments by nature and design. Am Psychol 32:513–531

Bronfenbrenner, U (2005). Making human beings human: bioecological perspectives on human development. Sage.

Bronfenbrenner U (2009) The ecology of human development . Harvard university press.

Brown P (2003) The opportunity trap: Education and employment in a global economy. Eur Educ Res J 2(1):141–179

Brown P, Hesketh A (2004) The mismanagement of talent: Employability and jobs in the knowledge economy. Oxford University Press, Oxford

Brown P, Lauder H (2006) “Globalisation, knowledge and the myth of the magnet economy”. Globalisation, Societies and Education 4(1):25–57

Burke PJ (2012) The right to higher education: Beyond widening participation. Routledge, Oxon

Card D, Lemieux T (2001) Can falling supply explain the rising return to college for younger men? A cohort-based analysis”. Q J Econ 116(2):705–746

Article   MATH   Google Scholar  

Castoriadis C (1997) The Castoriadis reader. Blackwell, Oxford

CFE and Edge Hill University (2013) The uses and impact of HEFCE funding for widening participation. HEFCE, Bristol, http://www.hefce.ac.uk/media/hefce/content/pubs/indirreports/2013/usesandimpactofwpfunding/The%20uses%20and%20impact%20of%20HEFCE%20funding%20for%20widening%20participation.pdf

Chang SS, Stuckler D, Yip P, Gunnell D (2013) Impact of 2008 global economic crisis on suicide: Time trend study in 54 countries. Br Med J 347:1–15

Chapman C, Laird J, Ifill N, KewalRamani A (2011) Trends in High School Dropout and Completion Rates in the United States: 1972-2009 . Compendium Report. NCES 2012-006. National Center for Education Statistics

Coffield F (1999) Breaking the consensus: Lifelong learning and social control. Br Educ Res J 25(4):479–499

Cogan J, Derricott R (eds) (2014) Citizenship for the 21st century: An international perspective on education. Routledge, London

Coleman JA (2006) English-medium teaching in European higher education. Lang Teach 39(1):1–14

Coleman JS (1988) Social capital in the creation of human capital. Am J Sociol 94(Supplement):S95–S120

Collins R (1979) Credential society: Historical sociology of education and stratification. Academic Press, New York

Council of Europe (Ed.) (2004) All European study on education for democratic citizenship policies. Council of Europe Publishing, Strasbourg, http://www.coe.int/t/dg4/education/edc/Source/Resources/Pack/AllEuropeanStudyEDCPolicies_En.pdf

Dall’ Alba G, Barnacle R (2007) An ontological turn for higher education. Stud High Educ 32(6):679–691

Darder A (2003) The critical pedagogy reader. Psychology Press, New York

Diamond A, Walkley L, Scott-Davies S (2011) Global graduates into global leaders. CIHE, London

Dore R (1976) The diploma disease: Education, qualification and development. Allen & Unwin, London

Dore R (1997) Reflections on the diploma disease twenty years later. Assessment in Education 4(1):189–205

Dorling D, Dorling D (2015) Injustice: Why social inequality still persists. Policy Press, Bristol

Douthat RG (ed) (2005) Privilege: Harvard and the education of the ruling class. Hyperion, New York

Durkheim E (1956) Education and Sociology. Free Press, Glencoe, IL

Durkheim E (2006) Education: Its nature and its role. In: Lauder, et al. (Eds) Education, globalisation & social change. Taylor and Francis Ltd, San Francisco, p 76–88

Durst RK (1999) Collision Course: Conflict, Negotiation, and Learning in College Composition. National Council of Teachers of English, Urbana, IL

Entwistle NJ, Peterson ER (2004) Conceptions of learning and knowledge in higher education: Relationships with study behaviour and influences of learning environments. Int J Educ Res 41(6):407–428

Evans K, Hodkinson P, Unwin L (eds) (2004) Working to learn: Transforming learning in the workplace. Routledge, London

Freire P (1970) Pedagogy of the Oppressed. Herder& Herder, New York

Freire P (2009) Pedagogy of the Oppressed 30th Anniversary Eds. Continuum, New York

Ganding LA, Apple M (2002) Can education challenge neoliberalism? The citizen school and the struggle for democracy in Porto Alegre, Brazil. Social Justice 29(4):26–40

Garland C (2008) The McDonaldization of higher education: Notes on the UK experience. Fast Capitalism 4(1). Available at: http://www.uta.edu/huma/agger/fastcapitalism/4_1/garland.html

Giroux H, Myrsiades K (eds) (2001) Beyond the corporate university: Culture and pedagogy in the new millennium. Rowan and Littlefield Publishers, Lanham

Giroux HA (1983) Theory and resistance in education: A pedagogy for the opposition. Bergin & Garvey, South Hadley, MA

Goethe J (2006) The Sorrows of Young Werther. Penguin, UK, (Vol. 10)

Goldthorpe JH (2007) “Cultural Capital”: some critical observation. In: Scherer S, Pollak R, Otte G, Gangl M (eds) From Origin to Destination. Campus, Frankfurt am Main, p 78–101

Gorard S (2008) Who is missing from higher education? Camb J Educ 38(3):421–437

Gouthro PA (2002) Education for sale: At what cost? Lifelong learning and the marketplace. Int J Lifelong Educ 21(4):334–346

Graham G (2013) The university: A critical comparison of three ideal types. In: Sugden R, Valania M, Wilson JR (eds) Leadership and cooperation in academia reflecting on the roles and responsibilities of university faculty and management. Edward Elgar Publishing Limited, Cheltenham, UK, p 1–16

Greenbank P, Hepworth S (2008) Working class students, the career decision-making process: a qualitatitive study. Higher Education Careers Service Unit, Manchester

Haigh M, Clifford VA (2011) Integral vision: A multi-perspective approach to the recognition of graduate attributes. High Educ Res Dev 30(5):573–584

Harvey L (2000) New realities: The relationship between higher education and employment. Tert Educ Manag 6(1):3–17

Article   MathSciNet   Google Scholar  

Hayes D, Wynard R (eds) (2006) The McDonaldization of higher education, 2nd edn. Bergin and Harvey, Westport, CT

Hayward G, James S (eds) (2004) Balancing the skills equation: Key issues for policy and practice. Policy Press, Bristol

Hegel GWF (1977) The Phenomenology of Spirit. Translated by A. V. Miller. Oxford: Oxford University Press

Heinze T, Knill C (2008) Analysing the differential impact of the Bologna Process: Theoretical considerations on national conditions for international policy convergence. High Educ 56(4):493–510

Heidegger M (1998) Plato’s doctrine of truth. In: McNeill W (ed) Pathmarks. Cambridge University Press, Cambridge, p 155–182

Chapter   Google Scholar  

Herder JG (2002) Herder: Philosophical Writings. Translated and Edited by M. Foster. Cambridge University Press, Cambridge

Hess FM (2004) Common sense school reform. Palgrave MacMillan, New York

Hoxby CM (1996) How teachers’ unions affect education production. Q J Econ 111(3):671–718

Hunter CP (2013) Shifting themes in OECD country reviews of higher education. High Educ 66(6):707–723

Hyland A ( 2011) Entry to higher education in Ireland in the 21st century. National Council for Curriculum and Assessment/Higher Education Authority (NCCA/HEA) Seminar, Discussion Paper, Cork, Ireland. 21 September 2011. Available at: http://hea.ie/assets/uploads/2017/04/Aine-Hyland_Entry-to-Higher-Education-in-Ireland-in-21st-Century-2011.pdf

Hyslop-Margison EJ (2000) The market economy discourse on education: Interpretation, impact, and resistance. Alta J Educ Res 46(3):203–213

Jongbloed B, Enders J, Salerno C (2008) Higher education and its communities: interconnections, interdependencies and a research agenda. Higher Education 56(3):303–324

Kaiser F, Maassen P, Meek L, van Vught F, de Weert E, Goedegebuure L (eds) (2014) Higher education policy: An international comparative perspective . Pergamont Press, Oxford.

Karabel J (2006) The chosen: The hidden history of admission and exclusion at Harvard, Yale, and Princeton. Houghton Mifflin Harcourt, Boston, MA

Kozol J (2005) The shame of the nation: The restoration of apartheid schooling in America. Crown Publishing, New York

Kozol J (2012) Savage inequalities: Children in America’s schools. Broadway Books, New York

Latour B, Woolgar S (1979) Laboratory life: The social construction of scientific facts. Sage, Beverly Hills

Lepori B, Bonaccorsi A (2013) The socio-political construction of a European Census of higher education institutions: Design, methodological and comparability Issues. Minerva 51(3):271–293

Letizia A (2013) Dialectical constellations of progress: New visions of public higher education for the twenty-first century. J Crit Thought Prax 2(1):55–82

Livingstone (2009) Education & jobs: Exploring the gaps. University of Toronto Press, Toronto

Livingstone DW, Stowe S (2007) Class, race, space, and unequal educational outcomes in the United States: Beyond dichotomies. In: Kincheloe J, Steinberg S (eds.) Cutting class: Socioeconomic status and education. Rowman and Littlefield Publishers, Plymouth, p 97–119

Lorenz C (2006) Will the universities survive the European Integration? Higher education policies in the EU and in the Netherlands before and after the Bologna Declaration. Sociologia Internationalis 44(1):1–28

Lowe J (2000) International examinations: The new credentialism and reproduction of advantage in a globalising world. Assessment in Education 7(3):363–377

Lu MZ, Horner B (2009) Composing in a global-local context: Careers, mobility, skills. Coll Eng 72(2):113–133

Määttä K, Uusiautti S (eds) (2014) Early child care and education in Finland . Routledge, Abingdon

Machin S, Van Reenen J (1998) Technology and changes in skill structure: evidence from seven OECD countries. Q J Econ 113(4):1215–1244

Mettler S (2014) Degrees of inequality: How the politics of higher education sabotaged the American dream. Basic Books, New York

Milburn AC (2009) Unleashing aspiration: the final report of the panel on fair access to the professions. Cabinet office, London

Mokyr J (2002) The gifts of Athena: Historical origins of the knowledge economy. Princeton University Press, Princeton, NJ

Mountford-Zimdars A, Sabbagh D (2013) Fair access to higher education: A comparative perspective. Comp Educ Rev 57(3):359–368

Newman JH (1996) The idea of a university. University Press, Yale

Nussbaum MC (1998) Cultivating humanity. Harvard University Press, Cambridge, MA

Osler A, Starkey H (2006) Education for democratic citizenship: A review of research, policy and practice 1995–2005. Research Papers in Education 21(4):433–466

Payne D (1999) Composition, interdisciplinary collaboration, and the postindustrial concern. JAC 19(4):607–632

Phillipson R (2009) Linguistic imperialism continued. Routledge, London and New York

Psacharopoulos G, Patrinos HA (2004) Human capital and rates of return. In: Johnes G, Johnes J (eds) International handbook on the economics of education. Elgar, Cheltenham, p 1–57

Ritzer G (2010) McDonaldization: The Reader, 3rd edn. Pine Forge Press, Thousand Oaks

Robinson K, Aronica L (2009) The element: How finding your passion changes everything. Viking, New Work

Saito M (2003) Amartya Sen’s Capability approach to education: A critical exploration. J Philos Educ 37(1):17–33

Sen A (1985) Commodities and capabilities. Oxford University Press, Amsterdam, North Holland, reprinted by Delhi

Sen A (1993) Capability and well-being. In: Nussbaum M, Sen A (eds) The Quality of Life. Clarendon Press, Oxford, p 30–53

Shapiro HT (2005) A larger sense of purpose: Higher education and society. Princeton University Press, Princeton

Solon G (1999) Intergenerational mobility in the labor market. Handbook of labor economics 3(Part A):1761–1800

Stephens NM, Brannon TN, Markus HR, Nelson JE (2015) Feeling at home in college: Fortifying school‐relevant selves to reduce social class disparities in higher education. Soc Issue Policy Rev 9(1):1–24

Stone P (2013) Access to higher education by the luck of the draw. Comp Educ Rev 57(3):577–599

Takala M, Wickman K, Uusitalo-Malmivaara L, Lundström A (2015) Becoming a special educator–Finnish and Swedish students’ views on their future professions. Educ Inq 6(1):25–51

Tight M (2012) Researching higher education. Open University Press, Maidenhead, UK

Thomson I (2001) Heidegger on ontological education, or: how we become what we are. Inquiry 44(3):243–268

Tomusk V (2002) The rise of the transnational capitalist class and World Bank “aid” for higher education. International Studies in Sociology of Education 12(3):335–352

Tomusk V (2004) Three Bolognas and a pizza pie: notes on institutionalization of the European higher education system. Int Stud Sociol Educ 14(1):75–96

Trow M (1979) Elite and mass higher education: American models and European realities. National Board of Universities, Stockholm

Trow M (2000) From mass higher education to universal access: The American advantage. Minerva 37(4):303–328

Unterhalter E (2003) The Capabilities approach and gendered education: An examination of south African complexities. Theory Res Educ 1(1):7–22

Unterhalter E (2005) Global inequality, capabilities, social justice: The millennium development goal for gender equality in education. Int J Educ Dev 25(2):111–122

Walker M, Unterhalter E (eds) (2007) Amartya Sen’s capability approach and social justice in education. Palgrave Macmillan, New York

Walters D (2004) A comparison of the labour market outcomes of postsecondary graduates of various levels and fields over a four-cohort period. Can J Sociol 29(1):1–27

MathSciNet   Google Scholar  

Waters T (2012) Schooling, childhood, and bureaucracy: Bureaucratizing the child. Palgrave Macmillan, New York

Waters T (2016) “Teaching like you do in America” –personal reflections from teaching across borders in Tanzania and Germany. Palgrave Communications. 1 (2015) (on-line) Available at: https://doi.org/10.1057/palcomms.2015.26

Weber M (1947) The Theory of Social and Economic Organization. Free Press, Glencoe, III

Weber M (1968) Economy and Society (2 volumes). University of California, Berkeley, CA

Westerheijden DF, Stensaker B, Rosa MJ (eds) (2007) Quality assurance in higher education: Trends in regulation, translation and transformation. Springer, Dordrecht, NL

Williams JJ (2001) Franchising the University. In: I Giroux H, Myrsiades K (eds) Beyond the corporate university: Culture and pedagogy in the new millennium. Rowan and Littlefield, Lanham, p 15–28

Wolff R, Barsamian D (2012) Occupy the economy: Challenging capitalism. City Lights Books, San Francisco

World Bank (2013) The Human Capital report. World Bank, Washington DC

Yuan L, Powell S, CETIS J (2013) MOOCs and open education: Implications for higher education. JISC CETIS (White Paper)

Download references

Author information

Authors and affiliations.

University of Glasgow, Glasgow, UK

Theocharis Kromydas

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Theocharis Kromydas .

Additional information

Competing interests: The author declares that there are no competing financial interests.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

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

Reprints and permissions

About this article

Cite this article.

Kromydas, T. Rethinking higher education and its relationship with social inequalities: past knowledge, present state and future potential. Palgrave Commun 3 , 1 (2017). https://doi.org/10.1057/s41599-017-0001-8

Download citation

Received : 14 February 2017

Accepted : 30 August 2017

Published : 13 October 2017

DOI : https://doi.org/10.1057/s41599-017-0001-8

Share this article

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

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

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Hyper-ambition and the replication crisis: why measures to promote research integrity can falter.

• Yasemin J. Erden

Journal of Academic Ethics (2024)

Differentiated socio-ecological system approach for vulnerability and adaptation assessment in the Central Himalaya

• Praveen Kumar
• Christine Fürst
• P. K. Joshi

Mitigation and Adaptation Strategies for Global Change (2024)

Democracy and Lifelong Learning in Africa

• Chimere O. Iheonu
• Simplice A. Asongu

Journal of the Knowledge Economy (2024)

Research- vs. government-driven physical activity policy monitoring: a systematic review across different levels of government

• Sven Messing
• Antonina Tcymbal
• Peter Gelius

Health Research Policy and Systems (2023)

The neglect of researchers during the first COVID-19 pandemic induced national lockdown in India: inside the lives of JNU’s research scholars

• Yangchen Roy
• Somashree Das

Higher Education (2023)

Quick links

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

The Review of Higher Education

Penny A. Pasque, The Ohio State University; Thomas F. Nelson Laird, Indiana University, Bloomington

Journal Details

The Review of Higher Education  is interested in empirical research studies, empirically-based historical and theoretical articles, and scholarly reviews and essays that move the study of colleges and universities forward. The most central aspect of  RHE  is the saliency of the subject matter to other scholars in the field as well as its usefulness to academic leaders and public policymakers. Manuscripts submitted for  RHE  need to extend the literature in the field of higher education and may connect across fields and disciplines when relevant. Selection of articles for publication is based solely on the merits of the manuscripts with regards to conceptual or theoretical frameworks, methodological accurateness and suitability, and/or the clarity of ideas and gathered facts presented. Additionally, our publications center around issues within US Higher Education and any manuscript that we send for review must have clear implications for US Higher Education. 

Guidelines for Contributors

Manuscripts should be typed, serif or san serif text as recommended by APA 7th edition (e.g., 11-point Calibri, 11-point Arial, and 10-point Lucida Sans Unicode, 12-point Times New Roman, 11-point Georgia, 10-point Computer Modern) double-spaced throughout, including block quotes and references. Each page should be numbered on the top right side of the page consecutively and include a running head. Please supply the title of your submission, an abstract of 100 or fewer words, and keywords as the first page of your manuscript submission (this page does not count towards your page limit). The names, institutional affiliations, addresses, phone numbers, email addresses and a short biography of authors should appear on a separate cover page to aid proper masking during the review process. Initial and revised submissions should not run more than 32 pages (excluding abstract, keywords, and references; including tables, figures and appendices). Authors should follow instructions in the 7th edition Publication Manual of the American Psychological Association; any manuscripts not following all APA guidelines will not be reviewed. Please do not change fonts, spacing, or margins or use style formatting features at any point in the manuscript except for tables. All tables should be submitted in a mutable format (i.e. not a fixed image). Please upload your manuscript as a word document. All supporting materials (i.e., tables, figures, appendices) should be editable in the manuscript or a separate word document (i.e., do not embedded tables or figures). For a fixed image, please upload a separate high-resolution JPEG.

Authors should use their best judgment when masking citations. Masking some or all citations that include an author’s name can help prevent reviewers from knowing the identities of the authors. However, in certain circumstances masking citations is unnecessary or could itself reveal the identities of manuscript authors. Because authors are in the best position to know when masking citations will be effective, the editorial team will generally defer to them for these decisions.

Manuscripts are to be submitted in Word online at  mc.manuscriptcentral.com/rhe . (If you have not previously registered on this website, click on the “Register here” link to create a new account.) Once you log on, click on the “Author Center” link and then follow the printed instructions to submit your manuscript.

The term “conflict of interest” means any financial or other interest which conflicts with the work of the individual because it (1) could significantly impair the individual’s objectivity or (2) could create an unfair advantage for any person or organization. We recommend all authors review and adhere to the ASHE Conflict of Interest Policy before submitting any and all work. Please refer to the policy at  ashe.ws/ashe_coi

Please note that  The Review of Higher Education  does not require potential contributors to pay an article submission fee in order to be considered for publication.  Any other website that purports to be affiliated with the Journal and that requires you to pay an article submission fee is fraudulent. Do not provide payment information. Instead, we ask that you contact the  RHE  editorial office at  [email protected]  or William Breichner the Journals Publisher at the Johns Hopkins University Press  [email protected] .

Author Checklist for New Submissions

Page Limit.  Manuscripts should not go over 32 pages (excluding abstract, keywords, and references; including tables, figures and appendices.)

Masked Review.  All author information (i.e., name, affiliation, email, phone number, address) should appear on a separate cover page of the manuscript. The manuscript should have no indication of authorship. Any indication of authorship will result in your manuscript being unsubmitted.

Formatting.  Manuscripts should be typed, serif or san serif text as recommended by APA 7th edition (e.g., 11-point Calibri, 11-point Arial, and 10-point Lucida Sans Unicode, 12-point Times New Roman, 11-point Georgia, 10-point Computer Modern), double-spaced throughout, including block quotes and references, and each page should be numbered on the top right side of the page consecutively. Authors should follow instructions in the 7th edition Publication Manual of the American Psychological Association; this includes running heads, heading levels, spacing, margins, etc.. Any manuscripts not following APA 7th edition will be unsubmitted. [Please note, the  RHE  editorial team recommends 12-pt Times New Roman font to ensure proper format conversion within the ScholarOne system.]

Abstract.  All manuscripts must include an abstract of 100 words or fewer, and keywords as the first page of your manuscript submission (this page does not count towards your page limit).

Author Note.  An Author’s note may include Land Acknowledgments, Disclosure Statement (i.e., funding sources), or other acknowledgments. This should appear on your title page (not in the masked manuscript).  

Tables.  All tables should be editable. Tables may be uploaded in the manuscript itself or in a separate word document. All tables must be interpretable by readers without the reference to the manuscript. Do not duplicate information from the manuscript into tables. Tables must present additional information from what has already been stated in the manuscript.

Figures.  Figures should be editable in the manuscript or a separate word document (i.e., no embedded tables). For fixed images, please upload high-resolution JPEGs separately.

References.  The reference page should follow 7th edition APA guidelines and be double spaced throughout (reference pages do not count toward your page limit). 

Appendices.  Appendices should generally run no more than 3 manuscript pages. 

Additional Checklist for Revised Submissions

Revised manuscripts should follow the checklist above, with the following additional notes: 

Page Limit.  Revised manuscripts should stay within the page limit for new submissions (32 pages). However, we do realize that this is not always possible, and we may allow for a couple of extra pages for your revisions. Extensions to your page length will be subject to editor approval upon resubmission, but may not exceed 35 pages (excluding abstract, keywords, and references).

• Author Response to Reviewer Comments.  At the beginning of your revised manuscript file, please include a separate masked statement that indicates fully [1] all changes that have been made in response to the reviewer and editor suggestions and the pages on which those changes may be found in the revised manuscript and [2] those reviewer and editor suggestions that are not addressed in the revised manuscript and a rationale for why you think such revisions are not necessary. This can be in the form of a table or text paragraphs and must appear at the front of your revised manuscript document. Your response to reviewer and editor comments will not count toward your manuscript page limit. Please note that, because you will be adding your response to the reviewer and editor feedback to the beginning of your submission, this may change the page numbers of your document unless you change the pagination and start your manuscript itself on page 1. The choice is yours but either way, please ensure that you reference the appropriate page numbers within your manuscript in these responses. Additionally, when you submit your revised manuscript, there will be a submission box labeled “Author Response to Decision Letter”. You are not required to duplicate information already provided in the manuscript, but instead may use this to send a note to the reviewer team (e.g., an anonymous cover letter or note of appreciation for feedback). Please maintain anonymity throughout the review process by NOT including your name or by masking any potentially identifying information when providing your response to the reviewer's feedback (both in documents and the ScholarOne system).

Editorial Correspondence

Please address all correspondence about submitting articles (no subscriptions, please) to one or both of the following editors:

Dr. Penny A. Pasque, PhD Editor, Review of Higher Education 341 C Ramseyer Hall 29 W. Woodruff Avenue The Ohio State University Columbus, OH 43210 email:  [email protected]

Dr. Thomas F. Nelson Laird, PhD Editor, Review of Higher Education 201 North Rose Avenue Indiana University School of Education Bloomington, IN 47405-100 email:  [email protected]

Submission Policy

RHE publishes original works that are not available elsewhere. We ask that all manuscripts submitted to our journal for review are not published, in press or submitted to other journals while under our review. Additionally, reprints and translations of previously published articles will not be accepted.

Type of Preliminary Review

RHE utilizes a collaborative review process that requires several members of the editorial team to ensure that submitted manuscripts are suitable before being sent out for masked peer-review. Members of this team include a Editor, Associate Editor and Managing Editors. Managing Editors complete an initial review of manuscripts to ensure authors meet RHE ’s Author Guidelines and work with submitting authors to address preliminary issues and concerns (i.e., APA formatting). Editors and Associate Editors work together to decide whether it should be sent out for review and select appropriate reviewers for the manuscript.

Type of Review

When a manuscript is determined as suitable for review by the collaborative decision of the editorial team, Editors and/or Associate Editors will assign reviewers. Both the authors’ and reviewers’ are masked throughout the review and decision process.

Criteria for Review

Criteria for review include, but are not limited to, the significance of the topic to higher education, completeness of the literature review, appropriateness of the research methods or historical analysis, and the quality of the discussion concerning the implications of the findings for theory, research, and practice. In addition, we look for the congruence of thought and approach throughout the manuscript components.

Type of Revisions Process

Some authors will receive a “Major Revision” or “Minor Revision” decision. Authors who receive such decisions are encouraged to carefully attend to reviewer’s comments and recommendations and resubmit their revised manuscripts for another round of reviews. When submitting their revised manuscripts, authors are asked to include a response letter and indicate how they have responded to reviewer comments and recommendations. In some instances, authors may be asked to revise and resubmit a manuscript more than once.

Review Process Once Revised

Revised manuscripts are sent to the reviewers who originally made comments and recommendations regarding the manuscript, whenever possible. We rely on our editorial board and ad-hoc reviewers who volunteer their time and we give those reviewers a month to provide thorough feedback. Please see attached pdf for a visual representation of the RHE workflow .

Timetable (approx.)

• Managing Editor Technical Checks – 1-3 days
• Editor reviews and assigns manuscript to Associate Editors – 3-5 days
• Associate Editor reviews and invites reviewers – 3-5 days
• Reviewer comments due – 30 days provided for reviews
• Associate Editor makes a recommendation –  5-7 days
• Editor makes decision – 5-7 days
• If R&R, authors revise and resubmit manuscript – 90 days provided for revisions
• Repeat process above until manuscript is accepted or rejected -

Type of review for book reviews

Book reviews are the responsibility of the associate editor of book reviews. Decisions about acceptance of a book review are made by that associate editor.

The Hopkins Press Journals Ethics and Malpractice Statement can be found at the ethics-and-malpractice  page.

The Review of Higher Education expects all authors to review and adhere to ASHE’s Conflict of Interest Policy before submitting any and all work. The term “conflict of interest” means any financial or other interest which conflicts with the work of the individual because it (1) could significantly impair the individual’s objectivity or (2) could create an unfair advantage for any person or organization. Please refer to the policy at ashe.ws/ashe_coi .

Guidelines for Book Reviews

RHE publishes book reviews of original research, summaries of research, or scholarly thinking in book form. We do not publish reviews of books or media that would be described as expert opinion or advice for practitioners.

The journal publishes reviews of current books, meaning books published no more than 12 months prior to submission to the associate editor in charge of book reviews.

If you want to know whether the RHE would consider a book review before writing it, you may email the associate editor responsible for book reviews with the citation for the book.

Reviewers should have scholarly expertise in the higher education research area they are reviewing.

Graduate students are welcome to co-author book reviews, but with faculty or seasoned research professionals as first authors.

Please email the review to the associate editor in charge of book reviews (Timothy Reese Cain, [email protected] ), who will work through necessary revisions with you if your submission is accepted for publishing.

In general, follow the APA Publication Manual, 7th edition.

Provide a brief but clear description and summary of the contents so that the reader has a good idea of the scope and organization of the book. This is especially important when reviewing anthologies that include multiple sections with multiple authors.

Provide an evaluation of the book, both positive and negative points. What has been done well? Not so well? For example the following are some questions that you can address (not exclusively), as appropriate:

What are the important contributions that this book makes?

What contributions could have been made, but were not made?

What arguments or claims were problematic, weak, etc.?

How is the book related to, how does it supplement, or how does it complicate current work on the topic?

To which audience(s) will this book be most helpful?

How well has the author achieved their stated goals?

Use quotations efficiently to provide a flavor of the writing style and/or statements that are particularly helpful in illustrating the author(s) points. 

If you cite any other published work, please provide a complete reference.

Please include a brief biographical statement immediately after your name, usually title and institution. Follow the same format for co authored reviews. The first author is the contact author.

Please follow this example for the headnote of the book(s) you are reviewing: Stefan M. Bradley. Upending the Ivory Tower: Civil Rights, Black Power, and the Ivy League. New York: New York University Press, 2018. 465 pp. $35. ISBN 97814798739999.

Our preferred length is 2,000–2,500 words in order for authors to provide a complete, analytical, review. Reviews of shorter books may not need to be of that length.

The term “conflict of interest” means any financial or other interest which conflicts with the work of the individual because it (1) could significantly impair the individual’s objectivity or (2) could create an unfair advantage for any person or organization. We recommend all book reviewers read and adhere to the ASHE Conflict of Interest Policy before submitting any and all work. Please refer to the policy at ashe.ws/ashe_coi

NOTE: If the Editor has sent a book to an author for review, but the author is unable to complete the review within a reasonable timeframe, we would appreciate the return of the book as soon as possible; thanks for your understanding.

Please send book review copies to the contact above. Review copies received by the Johns Hopkins University Press office will be discarded.

Penny A. Pasque,         The Ohio State University

Thomas F. Nelson Laird,         Indiana University-Bloomington

Associate Editors

Angela Boatman,         Boston College

Timothy Reese Cain (including Book Reviews),         University of Georgia

Milagros Castillo-Montoya,         University of Connecticut

Tania D. Mitchell,         University of Minnesota

Chrystal A. George Mwangi       George Mason University

Federick Ngo,        University of Nevada, Las Vegas

Managing Editors

Stephanie Nguyen,         Indiana University Bloomington

Monica Quezada Barrera,         The Ohio State University

Editorial Board

Sonja Ardoin,         Clemson University

Peter Riley Bahr,        University of Michigan

Vicki Baker,      Albion College

Allison BrckaLorenz,        Indiana University Bloomington

Nolan L. Cabrera,        The University of Arizona

Brendan Cantwell,        Michigan State University

Rozana Carducci,        Elon University

Deborah Faye Carter,         Claremont Graduate University

Ashley Clayton,         Louisiana State University

Regina Deil-Amen,         The University of Arizona 

Jennifer A. Delaney,     University of Illinois Urbana Champaign

Erin E. Doran,    Iowa State University

Antonio Duran,   Arizona State University 

Michelle M. Espino,        University of Maryland 

Claudia García-Louis,        University of Texas, San Antonio

Deryl Hatch-Tocaimaza,        University of Nebraska-Lincoln

Nicholas Hillman,        University of Wisconsin-Madison

Cindy Ann Kilgo,        Indiana University-Bloomington

Judy Marquez Kiyama,  University of Arizona

Román Liera,        Montclair State University

Angela Locks,        California State University, Long Beach

Demetri L. Morgan,  Loyola University Chicago

Rebecca Natow,         Hofstra University 

Z Nicolazzo,        The University of Arizona

Elizabeth Niehaus,        University of Nebraska-Lincoln

Robert T. Palmer,        Howard University

Rosemary Perez,        University of Michigan

OiYan Poon,         Spencer Foundation 

Kelly Rosinger,        The Pennsylvania State University

Vanessa Sansone,         The University of Texas at San Antonio

Tricia Seifert,        Montana State University

Barrett Taylor,         University of North Texas 

Annemarie Vaccaro,  University of Rhode Island

Xueli Wang,        University of Wisconsin-Madison

Stephanie Waterman,         University of Toronto 

Rachelle Winkle-Wagner,         University of Wisconsin-Madison

Association for the Study of Higher Education Board of Directors

The Review of Higher Education is the journal of Association for the Study Higher Education (ASHE) and follows the ASHE Bylaws and Statement on Diversity. 

ASHE Board of Directors

Abstracting & Indexing Databases

• Current Contents
• Web of Science
• Dietrich's Index Philosophicus
• IBZ - Internationale Bibliographie der Geistes- und Sozialwissenschaftlichen Zeitschriftenliteratur
• Internationale Bibliographie der Rezensionen Geistes- und Sozialwissenschaftlicher Literatur
• Academic Search Alumni Edition, 9/1/2003-
• Academic Search Complete, 9/1/2003-
• Academic Search Elite, 9/1/2003-
• Academic Search Premier, 9/1/2003-
• Current Abstracts, 9/1/2003-
• Education Research Complete, 3/1/1997-
• Education Research Index, Sep.2003-
• Education Source, 3/1/1997-
• Educational Administration Abstracts, 3/1/1991-
• ERIC (Education Resources Information Center), 1977-
• MLA International Bibliography (Modern Language Association)
• Poetry & Short Story Reference Center, 3/1/1997-
• PsycINFO, 2001-, dropped
• Russian Academy of Sciences Bibliographies
• TOC Premier (Table of Contents), 9/1/2003-
• Scopus, 1996-
• Gale Academic OneFile
• Gale OneFile: Educator's Reference Complete, 12/2001-
• Higher Education Abstracts (Online)
• ArticleFirst, vol.15, no.3, 1992-vol.35, no.2, 2011
• Electronic Collections Online, vol.20, no.1, 1996-vol.35, no.2, 2011
• Periodical Abstracts, v.26, n.4, 2003-v.33, n.3, 2010
• PsycFIRST, vol.24, no.3, 2001-vol.33, no.1, 2009
• Personal Alert (E-mail)
• Education Collection, 7/1/2003-
• Education Database, 7/1/2003-
• Health Research Premium Collection, 7/1/2003-
• Hospital Premium Collection, 7/1/2003-
• Periodicals Index Online, 1/1/1981-7/1/2000
• Professional ProQuest Central, 07/01/2003-
• ProQuest 5000, 07/01/2003-
• ProQuest 5000 International, 07/01/2003-
• ProQuest Central, 07/01/2003-
• Psychology Database, 7/1/2003-
• Research Library, 07/01/2003-
• Social Science Premium Collection, 07/01/2003-
• Educational Research Abstracts Online
• Research into Higher Education Abstracts (Online)
• Studies on Women and Gender Abstracts (Online)

Abstracting & Indexing Sources

• Contents Pages in Education   (Ceased)  (Print)
• Family Index   (Ceased)  (Print)
• Psychological Abstracts   (Ceased)  (Print)

Source: Ulrichsweb Global Serials Directory.

1.8 (2022) 3.2 (Five-Year Impact Factor) 0.00195 (Eigenfactor™ Score) Rank in Category (by Journal Impact Factor): 185 of 269 journals, in “Education & Educational Research”

© Clarivate Analytics 2023

Published quarterly

Readers include: Scholars, academic leaders, administrators, public policy makers involved in higher education, and all members of the Association for the Study of Higher Education (ASHE)

Print circulation: 761

Print Advertising Rates

Full Page: (4.75 x 7.5") - $450.00

Half Page: (4.75 x 3.5") - $338.00

2 Page Spread - $675.00

Print Advertising Deadlines

September Issue – July 15

December Issue – October 15

March Issue – January 15

June Issue – April 15

Online Advertising Rates (per month)

Promotion (400x200 pixels) – $338.00

Online Advertising Deadline

Online advertising reservations are placed on a month-to-month basis.

All online ads are due on the 20th of the month prior to the reservation.

General Advertising Info

For more information on advertising or to place an ad, please visit the Advertising page. 

eTOC (Electronic Table of Contents) alerts can be delivered to your inbox when this or any Hopkins Press journal is published via your ProjectMUSE MyMUSE account. Visit the eTOC instructions page for detailed instructions on setting up your MyMUSE account and alerts.  

Also of Interest

Vasti Torres, Indiana University

Patti Duncan, Oregon State University

Georgia L. Irby, College of William & Mary

Benjamin Fraser, The University of Arizona

Jeraldine Kraver, University of Northern Colorado; Peter Kratzke, University of Colorado, Boulder

Chrysogonus Siddha Malilang, Malmö University, Sweden

Carine Bourget, University of Arizona

Megan Corbin, West Chester University

Joseph Michael Sommers, Central Michigan University

Hopkins Press Journals

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Student engagement and wellbeing over time at a higher education institution

Roles Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft

* E-mail: [email protected]

Affiliation Computer Science, University of Exeter, Exeter, United Kingdom

Roles Data curation, Methodology, Software

Affiliation School of Psychology, University of Exeter, Exeter, United Kingdom

Roles Conceptualization, Data curation, Investigation, Methodology, Writing – review & editing

Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Writing – review & editing

Roles Conceptualization, Funding acquisition, Methodology, Supervision, Writing – original draft, Writing – review & editing

• Chris A. Boulton, 
• Emily Hughes, 
• Carmel Kent, 
• Joanne R. Smith, 
• Hywel T. P. Williams

• Published: November 27, 2019
• https://doi.org/10.1371/journal.pone.0225770
• Reader Comments

Student engagement is an important factor for learning outcomes in higher education. Engagement with learning at campus-based higher education institutions is difficult to quantify due to the variety of forms that engagement might take (e.g. lecture attendance, self-study, usage of online/digital systems). Meanwhile, there are increasing concerns about student wellbeing within higher education, but the relationship between engagement and wellbeing is not well understood. Here we analyse results from a longitudinal survey of undergraduate students at a campus-based university in the UK, aiming to understand how engagement and wellbeing vary dynamically during an academic term. The survey included multiple dimensions of student engagement and wellbeing, with a deliberate focus on self-report measures to capture students’ subjective experience. The results show a wide range of engagement with different systems and study activities, giving a broad view of student learning behaviour over time. Engagement and wellbeing vary during the term, with clear behavioural changes caused by assessments. Results indicate a positive interaction between engagement and happiness, with an unexpected negative relationship between engagement and academic outcomes. This study provides important insights into subjective aspects of the student experience and provides a contrast to the increasing focus on analysing educational processes using digital records.

Citation: Boulton CA, Hughes E, Kent C, Smith JR, Williams HTP (2019) Student engagement and wellbeing over time at a higher education institution. PLoS ONE 14(11): e0225770. https://doi.org/10.1371/journal.pone.0225770

Editor: Marina Della Giusta, University of Reading, UNITED KINGDOM

Received: April 23, 2019; Accepted: November 12, 2019; Published: November 27, 2019

Copyright: © 2019 Boulton et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Survey response data can be found at 10.5281/zenodo.3480070.

Funding: This research was supported by the Effective Learning Analytics project at the University of Exeter. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Engagement with learning is believed to be an important factor in student success in higher education. Engagement has been defined in different ways in the literature [ 1 ], but is considered here to refer to the active commitment and purposeful effort expended by students towards all aspects of their learning, including both formal and informal activities [ 2 ]. Student engagement has been shown to be related to success in both online learning [ 3 – 5 ] and more traditional campus-based higher education settings [ 6 – 8 ]. However, engagement can be difficult to measure. In most studies of online-only education (e.g. [ 9 – 13 ]), student engagement is measured from the interactions a student has within a virtual learning environment (VLE). This may be a reasonable approach for digital-only contexts where a large proportion of learning activities occur through this channel. In contrast, in a traditional, face-to-face learning, university environment, VLE usage only captures one dimension of student learning activity and full engagement with learning is much harder to measure. The numerous and varied interactions students have with their learning programmes, including lectures, seminars, peer group discussions and ad hoc interactions with teaching staff, as well as other aspects of campus life such as participation in sports and student societies, are harder to record, requiring innovative methods for their capture [ 14 , 15 ].

Exploration of the relationship between student engagement and success raises the important question of how “success” is defined. Most obviously, success relates to academic performance, such as final grades (e.g. [ 6 – 8 , 16 ]), but success is also often discussed in terms of retention and completion of a course of learning (e.g. [ 7 , 10 , 13 , 17 – 19 ]). It is important to consider that students may have different motivations for attending university, including, for example, social or sporting aims alongside conventional academic goals. Thus, in seeking to link engagement to success, there is value in adopting a more holistic view of student motivations and appropriate measures of outcomes. Furthermore, it is important to note that engagement and success, however measured, are dynamic and should be expected to vary within and between individuals over the duration of academic study.

There is increasing interest in learning analytics [ 20 – 25 ], which may use either static attributes of students (e.g. demographics, socioeconomic indicators, previous attainment) or dynamic attributes based on digital traces of learning behaviour to understand many aspects of the student experience, including student engagement. Traditionally, such studies have primarily made use of “found” data from institutional databases and “by-product” data from digital learning platforms. This kind of data, which is not collected for the purpose of pedagogical research, has limitations. The records that are collected institutionally tend to relate to either the administration of higher education (e.g., demographic data, recruitment/retention statistics) or to the core components of academic performance (e.g., grades, progression, completion). Data collected as the by-product of student learning activities on digital platforms such as VLEs (e.g. [ 8 – 10 ] only offers a partial view of a complex whole. For example, previous work that examined the relationship between academic performance and engagement at a traditional University found that VLE usage alone is a relatively poor predictor of academic performance in this context [ 8 ], while another study showed that VLE usage was a useful predictor of outcomes for online learning but not significant for face-to-face learning [ 9 ].

Dispositional learning analytics (see [ 26 ]), on the other hand, seeks to combine digital trace data (e.g., those generated by engagement in online learning activities) with learner data (e.g., dispositions, attitudes, and values assessed via self-report surveys). By doing so, recent research has found that learning dispositions (e.g., motivation, emotion, self-regulation) strongly and dynamically influence engagement and academic performance over time (e.g., [ 27 – 29 ]). In addition, this research suggests that the predictive value added by consideration of learner data might be time-dependent: learner data seems to play a critical role up until the point that feedback from assessment or online activities becomes available. This raises the possibility that whether incorporating learner dispositions into learning analytics models is useful depends on learning context (i.e., online only versus campus-based institutions).

Another limitation of learning analytics based solely on digital traces, is that these sources often cannot capture subjective aspects of student life, such as wellbeing and satisfaction, which are rarely routinely measured. Relationships between student engagement and wellbeing, or between wellbeing and success, have consequently been less well studied for higher education than that between engagement and success (but see [ 30 , 31 ]). One project that has moved beyond by-product data and used deliberate collection of digital records to measure student behaviour and wellbeing is the StudentLife study at Dartmouth College in the USA [ 14 ]. This project supplied mobile phones to student participants in a term-long study that attempted to capture a multi-dimensional and longitudinal view of student behaviour. Findings used aspects of student life that had previously been inaccessible to researchers, including social interactions and physical activity patterns, to predict academic performance [ 16 ] and also to diagnose wellbeing issues [ 14 , 32 ]. While the StudentLife study showed that deliberate data collection using digital methods can access important aspects of the subjective student experience, it does not address the difficulty of doing so using the kinds of by-product digital records and institutional data that are routinely collected and used as input into learning analytics.

The importance of student wellbeing for academic outcomes, and the relationships between wellbeing and engagement, remain open research questions for higher education. Wellbeing is a loosely defined concept that may include a number of different dimensions, including satisfaction, positive affect (e.g. enjoyment, gratitude, contentment) and negative affect (e.g. anger, sadness, worry) [ 33 , 34 ]. Many studies have explored the relationship between wellbeing and academic performance, commonly finding a positive association, e.g. in US college undergraduates [ 35 , 36 ] and among high school students [ 37 ]. The relationship between engagement and wellbeing is less well studied in higher education, but a positive association has been found in other working environments [ 34 ]. A recent government report on student mental health and wellbeing in UK universities found increasing incidence of mental illness, mental distress and low wellbeing [ 38 ]. The same study found that these negative wellbeing factors had a substantial harmful impact on student performance and course completion; by extension, students with positive wellbeing are likely to perform better and complete their studies. Another study by the UK Higher Education Academy focused on methods for promoting wellbeing in higher education, as well as identifying several pedagogical benefits [ 39 ].

Here we report on a longitudinal survey of student learning behaviours at a traditional campus-based university in the United Kingdom. Our survey was designed to capture multiple dimensions of student engagement and wellbeing over time, deliberately using self-report to look beyond digital traces and institutional records. An initial questionnaire included questions to characterise individual students on different dimensions including learning style and motivations for study. Subsequent waves captured student learning behaviours and engagement with a wide variety of learning systems (both offline and online) and activities, as well as their subjective feelings of satisfaction and wellbeing. The survey ran in 10 waves spanning a teaching semester, vacation and exam period, allowing observation of changes over time.

This study aims to complement the growing body of work that uses digital trace data to measure engagement, with a more subjective offline approach that captures a fuller representation of the student experience. Our research goals are to understand how engagement and wellbeing vary over time, as well as to determine a multidimensional view of student learning behaviours and patterns. Addressing these questions will make an important contribution to the academic study of student engagement and will help to identify other learning dispositions (e.g., engagement) that might be of value to combine with digital trace data in learning analytic models. Findings may also offer instrumental benefit by helping to guide institutional decision-making around interventions and student support.

The cohort for the survey consisted of 1st year and 2nd year undergraduate students at a research-intensive campus-based university in the United Kingdom. Students were invited to participate via emails containing a link to survey registration. In addition, recruitment booths were set up at the university’s main campus and researchers approached students to invite them to participate. Students were incentivised by entry into a prize draw to win gift vouchers for a well-known online retailer, with 10 prizes available in each wave. There were 10 waves in all. To incentivise continued participation, there was an additional final prize draw with larger prizes available to students who had completed 80% of surveys. Every participant explicitly gave their consent to their data being analysed for research purposes.

The survey ran from February to June 2017. Of the 10 waves, Waves 1–7 were released weekly during the Spring term, followed by a break for the Easter vacation period. Waves 8–10 were released fortnightly during the Summer term, which at this institution was mostly taken up with revision and examinations. Responses were received asynchronously, so although the survey was released in waves, we analyse the data over a continuous time interval spanning 19 weeks.

Our longitudinal survey consisted of a series of questions that students completed in every wave. To measure engagement with learning, we asked respondents to report their participation in each of 17 different learning activities (see Table 1 ), measured as the number of days in the past 7 days they had performed that activity. These activities were selected to represent the variety of online and offline activities, as well as social and academic activities, available to students at the university. To give context, we also asked respondents to report whether they had an assessment due in the past 7 days.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0225770.t001

Effort over the preceding week was assessed with two items assessed on a 5-point Likert scale (specifically, “How engaged were you with your studies?”; “How much effort did you put into your studies?”, 1 = not at all, 5 = very much). The mean response from each student was used to form a reliable scale (Pearson’s r = 0.78, p < .001). Well-being over the last week was assessed with four items that asked about happiness in general (e.g., “How happy did you feel about your life in general?”) and in relation to their programme of study (e.g., “How well do you feel you are doing in your course?”, 1 = not at all, 5 = very much). Responses were averaged to form a reliable scale (Cronbach’s α = 0.69).

In addition to the longitudinal survey questions, we also asked further questions in Wave 1 to determine their self-reported learning engagement style and motivation for attending university.

Engagement with learning was assessed with 10 items adapted from the Student Engagement in Schools Questionnaire (SESQ; [ 40 ]). Participants indicated the extent of their agreement with the statements on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Principal components analysis with varimax rotation extracted two factors, accounting for 53% of the variance. The first factor was characterised by the items assessing cognitive engagement (e.g., “When I study, I try to understand the material better by relating it to things I already know”), and items were averaged to form a cognitive engagement scale (α = 0.73). The second factor was characterised by the items assessing behavioural engagement (e.g., “In my modules, I work as hard as I can”), and items were averaged to form a behavioural engagement scale (α = 0.75).

Participants indicated their agreement with six different reasons for attending university (1 = not at all, 5 = very much). Principal components analysis with varimax rotation extracted two factors, accounting for 57% of the variance. The first factor was characterised by the items assessing social motivations (e.g., “To socialise with friends”), and items were averaged to form a social motivations scale (α = 0.62). The second factor was characterised by the items assessing academic motivations (e.g., “To get good grades”), and items were averaged to form an academic motivations scale (α = 0.48). The original survey is shown in Supplementary Information ( S1 File ).

The survey and following analysis were undertaken in accordance with the guidelines of the British Psychological Society. All participants provided informed consent prior to participation and were free to withdraw at any time without penalty. The survey and analysis received ethical approval from the University of Exeter Psychology Ethics Committee prior to commencement of data collection.

Our analysis is based on both static and dynamic variables from the survey responses for each student. Static variables include the motivation and engagement style measurements that were calculated from Wave 1. An additional static variable was also used to measure student academic performance across the term in which the survey was conducted, using grade data from the university database; for this metric, a student grade variable was calculated as their credit-weighted average grade from all the modules they took during the term in which the survey was conducted. Dynamic variables include the engagement and wellbeing measurements recorded in every wave. To allow comparison between static variables and dynamic variables, we take the mean value for the dynamic variable (e.g., the mean number of days per week that a student participated in a learning activity, or their mean effort scale score). Correlations between variables are measured using the Pearson correlation coefficient and measure correlations between both the static and dynamic variables. In both cases, all data is used in the correlation measurement, such that there is one record per student who answered in Wave 1, and all the responses are used to calculate the correlation between the dynamic variables.

Dynamic variables were used to analyse trends in behaviour over time, such as trends in engagement and wellbeing. To allow analysis of trends across the whole cohort, we created time series for engagement and wellbeing variables using a moving average across all responses with a 7-day window size. To ensure robustness, we made sure there were at least 10 responses in each window for which a mean was calculated. Since counts were lower during vacation and examination periods, we restricted our trend analysis to term-time only. Trends in these time series were calculated using the Kendall rank correlation coefficient, which counts the proportion of concordant pairs (both x i >x j and y i >y j or x i <x j and y i <y j ). Using time as one of the variables, this gives a measure of tendency in the range [– 1 , 1 ], with a score of -1 if the time series is always decreasing, a score of +1 if the time series is always increasing, and a score of 0 if there is no overall trend.

Our analysis involved looking for differences in behaviour between sub-populations within our respondent cohort (e.g. splitting the cohort into those who did or did not have an assessment due each week). We present differences in the mean values between the two distributions and then use a Mann-Whitney U-test to determine if the distributions are significantly different. We use these non-parametric tests since the distributions of values are typically non-normal and vary in shape between different variables. We also have a small sample size once the distributions have been split. However, we still present the difference in mean values, rather than the difference in median values, since the discrete nature of our data (e.g., integer values in range 0–7, which for some variables have an inter-quartile range of 0 to 1) means that medians are sometimes too coarse-grained to show differences even where the distributions are significantly different.

Survey response

Overall, we had responses from 175 unique students, 174 of which answered the Wave 1 survey including questions to determine engagement style and motivations. We had 1050 responses overall, giving an average of exactly 6 responses per student.

Fig 1 shows the number of responses received over time during the 19-week period that the survey was active. There is an expected decline in the number of responses over time as participants lose interest or for other reasons drop out of the cohort. Despite this, we still have a reasonably steady and high response rate during the Spring term (left of the grey shaded area). There is a significant drop off in survey participation during the Easter break (grey shaded area), before the response rate recovers during the Summer term, although not to the levels seen previously (right of great shaded area). The Summer term in our survey is dominated by revision and exams, which suggests we might see different student behaviour.

Grey shaded region refers to the Easter break between semesters. Spring Term is to the left of the grey region, Summer Term to the right. Vertical dotted lines indicate the weeks in which a survey email was sent and a responder lottery was held to incentivise participation. Note that students could answer a survey wave in the following week, hence a lower amount of first-week responses is observed when compared to the 174 students that answered the first wave of the survey.

https://doi.org/10.1371/journal.pone.0225770.g001

Table 2 shows some demographics of our survey respondents (n = 175), compared to the entire student population (n = 15646). We find that our survey respondents are slightly biased towards being female and in their first year of study. The students who took the survey also have slightly higher marks than the student population. The number of students in the Life and Environmental Sciences college is greater than expected, with less representation of students from the Social Sciences and International Studies college and the Medical School. The low numbers from the Medical School reflect the fact that this School is based on a different campus to where physical recruitment of participants occurred.

https://doi.org/10.1371/journal.pone.0225770.t002

Respondent characteristics

The Wave 1 survey included one-time questions intended to allow construction of engagement style and motivation scores for each individual student (see Methods ). The distributions of these scores are shown in Fig 2 . Due to the nature of these measurements, and the fact that they are only measured once, they make up part of our ‘static’ data and can be thought of as measuring a student’s underlying dispositions. They suggest that generally students reported slightly higher levels of behavioural engagement than cognitive engagement, although there was a bigger spread in behavioural engagement scores. Most of the students who responded to our survey reported higher academic motivation than social motivation for attending university.

Students were asked a one-time set of questions to determine their engagement type and motivations (see Methods ) and as such this is a static measurement. Dotted lines show the minimum and maximum scores, solid lines show the interquartile range, and points show the medians.

https://doi.org/10.1371/journal.pone.0225770.g002

Relationships among student characteristics, average engagement and performance

Fig 3 shows the distributions of values from the longitudinal survey questions used to measure dynamic variables related to engagement with different learning activities and levels of student wellbeing. The plots show all responses from all students aggregated together, with the various learning activities ordered according to their mean usage level. The distributions suggest that activities that are most directly associated with learning (e.g. using the VLE, using the info app, using the Internet for learning, attending a teaching session) are used much more frequently than those that are not (e.g. using sports facilities, talking to a year representative, using SU facilities). This is consistent with the finding above that most students in the sample had stronger academic than social motivations for attending university. Distributions of scores on the “effort” and “happy” scales derived from the wellbeing questions asked each week (see Methods ) show that both metrics have a broad absolute range but a relatively narrow interquartile range. These metrics cannot be usefully compared.

The underlying survey questions were asked in all waves and as such these are dynamic variables. Plot shows minimum and maximum scores (dotted lines), the interquartile range (solid lines) and median values (points). For this analysis all student responses were pooled.

https://doi.org/10.1371/journal.pone.0225770.g003

Next, we related the various static variables to each other and to the mean values for the various dynamic variables for each student in our cohort. Table 3 shows (Spearman’s) correlations between static variables across the cohort for: engagement style, motivation, grades, wellbeing, and engagement levels. Statistical significance is indicated in Table 3 ; henceforth we only discuss correlations with statistical significance at level p <0.05, unless stated explicitly. For the dynamic variables, we use the mean reported level across all responses for each student. Grades are analysed using the average credit-weighted module grade from the term in which the survey was carried out (see Methods ).

https://doi.org/10.1371/journal.pone.0225770.t003

We find relatively strong positive correlation (ρ = 0.36) between levels of the two engagement styles (behavioural and cognitive). Behavioural engagement is correlated positively with academic motivation for attending university (ρ = 0.15) but correlated negatively with social motivation (ρ = -0.22). Behavioural engagement is very strongly positively correlated with effort (ρ = 0.55) and positively correlated with grades (ρ = 0.24). Cognitive engagement, on the other hand, is not correlated with grades (ρ = 0.02) but is positively correlated with happiness (ρ = 0.30). Cognitive engagement is also often positively correlated with participation in the various learning activities, with several positive correlations: seeing a lecturer (ρ = 0.32); going to the library (ρ = 0.28); using social media for learning (ρ = 0.18); and using the Internet for learning (ρ = 0.24). Cognitive engagement is negatively correlated with viewing lecture recordings (ρ = -0.16). Interestingly, behavioural engagement was typically uncorrelated with participation in learning activities except negatively with attending scheduled teaching sessions (ρ = -0.16); and viewing lecture recordings (ρ = -0.17).

The two types of motivation (academic and social) are not significantly correlated with each other (ρ = 0.14), but social motivation is correlated negatively with grades (ρ = -0.25). Academic motivation is significantly correlated with wellbeing scales for both effort (ρ = 0.28) and happiness (ρ = 0.29), whereas social motivation is not. Regarding participation in learning activities, the pattern of correlations makes intuitive sense. Academic motivation is weakly positively correlated with two academic activities: info app usage (ρ = 0.22); and VLE usage (ρ = 0.23). Social motivation is positively correlated with one core academic activity, attending a teaching session (ρ = 0.26), but is also positively correlated with several activities that are less directly academic and have a social aspect: working with friends (ρ = 0.19), using sports facilities (ρ = 0.46), using retail facilities (ρ = 0.23), using catering facilities (ρ = 0.23), using social media for learning (ρ = 0.21), and attending clubs or societies (ρ = 0.36).

It is interesting to note that the only significant correlations between student academic performance (measured by average grades) and levels of participation in learning activities are negative. Perhaps less surprising are negative correlations between grades and participation in “social” activities: using retail facilities (ρ = -0.22); and using catering facilities (ρ = -0.32). It is hard to explain the negative correlations between grades and attending a teaching session (ρ = -0.17). We return to this topic in the Discussion.

The wellbeing scales (effort and happiness) are positively correlated with each other (ρ = 0.30): students who put in more effort report greater happiness. Effort is positively correlated with several non-compulsory learning activities: using the VLE (ρ = 0.27); going to the library (ρ = 0.31); using career services (ρ = 0.30); using social media for learning (ρ = 0.36); and using the Internet for learning (ρ = 0.50). Effort is also positively correlated with using retail facilities (ρ = 0.27), perhaps suggesting more time spent on campus. Happiness is uncorrelated with core learning activities but is positively correlated with more social activities: using SU facilities (ρ = 0.28); and going to clubs or societies (ρ = 0.36).

Table 3 shows many positive correlations between levels of participation in various learning activities. Without listing all the pairwise relationships here, we find that 50% of activity pairs are significantly positively correlated, with no activity pairs negatively correlated. This suggests that students who engage more with learning do so in a holistic manner, with raised participation across a variety of learning activities.

Temporal trends and correlations

Next, we consider trends or changes in behaviour during the Spring term ( Fig 4 ), looking first at time series of reported participation levels for each learning activity (see Methods ). Since we use a moving average to give robust values, and since survey response rate falls outside term time, we restrict our analysis to the period within the Spring term (Waves 1–7, prior to the grey shaded area in Fig 1 ). We use a moving average equal to one week (7 days) and when doing this, the lowest number of responses in any window is 17 (on the last day of term), suggesting the plotted values are reliable. Apart from the final two days of term, all the windows have 38 or more responses within them. Trends are calculated using Kendall’s tau correlation coefficient (see Methods ). For ease of viewing, we have split the learning activities into ‘Online’ learning activities ( Fig 4a ), ‘Offline’ learning activities ( Fig 4b ) and ‘Other’ activities ( Fig 4c ). We also plot time series for wellbeing variables ( Fig 4d ).

Time series are calculated as a moving average using data from all students. Trends and significance are calculated using Kendall’s tau correlation coefficient.

https://doi.org/10.1371/journal.pone.0225770.g004

There is a general downward trend in participation with learning activities over the Spring term. Of the ‘Online’ systems ( Fig 4a ), all of them have a significantly downward trend as the term goes on: using the VLE (τ = -0.72); using the info app (τ = -0.65); using the Internet for learning (τ = -0.85); using social media for learning (τ = -0.67); and accessing lecture recordings (τ = -0.47). Three of the ‘Offline’ systems also decrease over the term ( Fig 4b ): attending teaching sessions (τ = -0.91); accessing the library (τ = -0.20); viewing past exams (τ = -0.56). Since teaching activities are scheduled with a roughly uniform density throughout the term, the downward trend in engagement with learning activities is notable. A similar trend is seen for many of the ‘Other’ activities ( Fig 4c ): going to clubs or societies (τ = -0.70); using the sports facilities (τ = -0.32); using retail facilities (τ = -0.83); using catering facilities (τ = -0.63); talking to a year rep (τ = -0.49); using SU facilities (τ = -0.68). There are no learning activities that show an increase in participation over the term.

Looking at trends in the wellbeing variables over the term, we see that effort increases slightly but not significantly (τ = 0.10). However, happiness increases significantly (τ = 0.36), suggesting that students report greater happiness as the term progresses. We cannot say whether this increase in self-reported happiness is related to the concurrent decrease in engagement, though it is tempting to speculate.

Table 4 shows correlations between the dynamic variables measuring participation in learning activities and wellbeing. This analysis shows whether there are temporal associations between levels of participation in different activities (e.g., if a student does more of one activity, does this correspond to more engagement in other activities). The striking observation in this analysis is that nearly all pairwise relationships between dynamic variables show significant positive correlations, with a small number of exceptions. This indicates a pattern whereby student learning activity varies holistically; students may be more or less active, but when they are active, they are active across a wide range of activities and behaviours. Again, the two wellbeing scales are correlated with each other (ρ = 0.37). Overall, 83% of the pairwise relationships between learning activities show a positive correlation over time (compared to 50% for the averaged data shown in Table 3 ). We find two significant negative correlations: between viewing past exam papers and visiting a lecturer (ρ = -0.08) and attending a teaching session (ρ = -0.13). This is most likely because Table 4 uses time-resolved information and is affected by the switch between attendance at scheduled teaching sessions during the Spring term and using past exams to revise for upcoming exams during the Summer term.

https://doi.org/10.1371/journal.pone.0225770.t004

Impact of assessments on engagement and wellbeing

To determine the impact of assessments (e.g., coursework, class tests, final exams, etc.) on student engagement and wellbeing, we split our dataset into “assessment week” responses (those responses where the student answered that there was an assessment due in the 7-day reporting period) and “non-assessment week” responses (where no assessments were due). Note that “assessment weeks” are temporally heterogeneous and specific to the individual; that is, the assessment/non-assessment weeks are not temporally correlated across the cohort. This rules out effects from globally correlated hidden variables such as, for example, campus wide events, external media stories, etc. For each set of responses, we create distributions for each dynamic variable and then measure the differences between these distributions using the difference in means and Mann-Whitney U-tests (see Methods ). Results are shown in Fig 5 . The bars in Fig 5 plot the difference in mean values for each distribution, with positive differences referring to increased participation in assessment weeks. Bar colours indicate whether the difference between the distributions is statistically significant according to the Mann-Whitney U-test.

Bars show the difference in mean values for reported score distributions for (upper panel) participation in each learning activity measured in days, or (lower panel) levels of effort and happiness on scale 1–5. Positive values indicate an increase in assessment weeks. Bar colours indicate statistical significance for the difference between distributions calculated from a Mann-Whitney U-test (blue—significant positive difference, red—significant negative difference, white—not significant).

https://doi.org/10.1371/journal.pone.0225770.g005

Fig 5 (upper panel) shows the mean difference for assessment weeks and non-assessment weeks in the reported number of days of participation in each learning activity. We find increased participation in all learning activities during assessment weeks, except using career services, which had significantly less usage when an assessment was due. Of the activities with increased participation, 9 of the 15 increases were significant. Interestingly, increased participation in assessment weeks extends across a mix of activity types; for example, there is greater attendance at clubs and societies when assessments are due. Overall, the analysis suggests there is higher engagement with most learning activities when assessments are due.

We also look for differences in the wellbeing variables of effort and happiness between assessment weeks and non-assessment weeks ( Fig 5 , lower panel). We find that there is a significant increase in the effort levels students report when an assessment is due. There is also, perhaps surprisingly, a slight increase in happiness, although this is not significant.

Relationships between behaviour and wellbeing

To explore the relationship between engagement with learning activities and reported wellbeing, we again split our dataset, this time into sets of responses where the student reported high/low levels of effort and high/low levels of happiness for that week. Since both variables are measured on an integer scale from 1 (low) to 5 (high), we use a threshold of 3 to split the cohort in each case, creating datasets for those who responded below 3 and those who reported 3 or above. This gives comparator sets for students who report “high effort” or “low effort” and students who report “happy” or “not happy”. Results are shown in Fig 6 .

Bars show the difference in mean scores (in days) from the distributions of participation levels for different learning activities. Positive values indicate higher participation by the (left) high effort and right (happy) students. Bar colour indicates significant differences between the distributions according to a Mann-Whitney U-test (blue—significant positive difference, red—significant negative difference, white—not significant).

https://doi.org/10.1371/journal.pone.0225770.g006

As expected, we find that 16 of the 17 learning activities show higher mean participation levels by high effort students, and for 10 of these the difference between the distributions is significant ( Fig 6 , left panel). Happy students have higher mean participation levels in all activities that students who are not happy ( Fig 6 , right panel). However, these differences are generally smaller than those for high vs low effort groups. When comparing the left and right panels in Fig 6 , there is a significant increase in going to the Sports Park and using catering facilities for happier students, whereas rates of viewing past exams are only significantly increased for high effort students.

In planning this research, we expected to find different patterns of engagement among students, such as individuals showing more engagement with certain systems and less with others. This might be driven by students’ personal preferences (e.g., [ 27 , 28 ]) or by the teaching activities prescribed and/or preferred by different disciplines and programmes (see e.g. [ 8 , 41 ]). Instead we find that students who are engaged with learning tend to be engaged with all learning activities and systems; engagement appears to be a holistic phenomenon (Tables 3 and 4 ). The only exception to this pattern is a negative correlation between attending scheduled teaching sessions and viewing past exam papers. This might be explained by the separation (for most students) of learning and revision, with exam papers used for revision after scheduled teaching has finished. The strong correlation between all forms of engagement with learning has possible instrumental value for the design of systems to monitor student engagement, since it suggests that engagement could be effectively tracked using only a subset of engagement metrics as indicators. Monitoring of engagement might be used to identify anomalies or changes in behaviour of individuals, for example, to assist tutors in providing support and pastoral care. Indeed, the predictive analytics project at Nottingham Trent University (NTU Student Dashboard), which calculates engagement scores based on five online resources (VLE access, library usage, attendance, assignment submissions, and card swipes), has identified a positive relationship between student engagement and both progression and attainment. Moreover, this information, when communicated to students and staff, has been used to provide more targeted support to students from pastoral tutors (see [ 42 ]).

A feature of our survey design is the ability to measure variables at a campus-based university that would otherwise be difficult to access. Of the 17 learning activities recorded by our survey, only four could be tracked digitally with current methods (VLE, info app, past exam views and recorded lecture viewing), with the rest not routinely measured. Furthermore, this study provides temporally resolved data on student wellbeing, giving the opportunity to explore relationships between engagement and wellbeing.

Engagement and wellbeing are shown in this study to be positively related. Looking longitudinally across the survey ( Table 4 ), we find 13 forms of engagement were positively (and significantly) correlated with at least one of the wellbeing variables, either effort or happiness. Reasonably, one could suggest a possible feedback loop where increasing engagement increases academic performance, which in turn increases wellbeing (happiness and grades are correlated; Table 4 ), which then increases engagement. Alternatively, students with greater background levels of wellbeing may be more likely to engage with learning (see also [ 30 , 31 ]). This study cannot separate these potential mechanisms, since it only shows correlation and cannot assign causality.

The responses to our survey show a broad sample of student engagement at the university where the study was based. The survey was widely advertised and contains responses from students across all disciplines. However, in common with most survey studies, it relies on voluntary participation and we had no control over who would participate (see also [ 43 ]). This may introduce bias into our results. For example, we find that the students who responded scored much higher on academic motivation than on social motivation ( Fig 2 ), but this may be an artefact of self-selection bias in the sample of survey respondents, such that academically motivated students who are engaged with learning were more likely to participate (see also [ 43 , 44 ]). Indeed, analysis of the demographic data of respondents suggests that certain disciplines were over-sampled. This might limit the generalizability of our findings to the whole cohort, given that there are likely to be disciplinary differences in the extent to which students are expected to engage with various learning systems (see [ 8 ]). Furthermore, since this study was based at a single university in the UK, it may not represent students at other universities in the UK or worldwide. We encourage other researchers to repeat our study at other institutions in order to consolidate our findings. We make our survey design available in the Supplementary Information ( S1 File ) to facilitate this.

Another caveat to our results is that differences between student workloads associated with different learning activities are not considered. In previous work, we have shown that the amount of observed VLE usage differs between different disciplines [ 8 ], explained by the differing requirements of different disciplines, programmes and modules. For example, a humanities student is likely to have a balance of learning activities that differs from an engineering student, with resulting variation in the time they spend on the VLE. In addition, the number of scheduled lectures and other contact hours will differ between disciplines, with students taking STEM subjects typically having more contact hours than those taking arts or humanities subjects which require more self-study. It is possible that these differences might affect some of our findings. For example, the correlation between attending scheduled teaching sessions and student happiness might be influenced by the fraction of sessions attended, rather than the absolute number; a student who attends 100% of 4 scheduled sessions might be happier than a student who attends 50% of 8 scheduled sessions, even though the number of attended sessions remains the same. This kind of difference might mask or confound some relationships, so it is possible that a study sample stratified on discipline or programme would give a more nuanced picture of the relationships between engagement and wellbeing. With a larger sample size, we would have been able to create disciplinary subsets of students to explore this aspect, but our sample size did not permit this here.

One interesting dimension of student engagement that we are yet to explore within our survey is how well students predict their own usage of various learning systems; that is, do they accurately report their usage of digital tools? Results given here are based on student self-report rather than documented usage of different systems. In general, students might mis-report their behaviour either by mistake or deliberately, for whatever reason. If self-reported data in the current survey are inaccurate, it might raise the interesting question of whether some students systematically under- or over-report their levels of engagement with learning, and whether students who misreport perform better or worse academically (see [ 45 , 46 ]). We will return to this question in future work. If self-report and documented data (where available) do not agree, it raises the question of which sources show a more accurate picture of student behaviour and which are more important in relation to student wellbeing.

We can only speculate why there is an observed decrease in engagement during the academic term. It could be because students like to get ahead at the start of term and work harder or engage more to do this. The larger drop off in engagement at the end of term may be due to students having assessments that are not due until after the break and therefore not needing to work as much as they do during the middle of term. The rise in reported effort during the term (although not statistically significant) is interesting in relation to the decrease in reported engagement. The observed increase in happiness towards the end of term seems to be robust but is hard to explain; we speculate that perhaps students become happier as they start to receive assessment outcomes, or maybe they are simply looking forward to the end of term. This may be at odds with the correlations between engagement and wellbeing discussed previously. However, we believe that the correlations are picking out individual student behaviours, whereas these trends reflect the whole population.

Our research identified strong differences in behaviour between students who have an assessment due and those who do not. This gives us confidence that our survey can identify meaningful results, despite the limited sample size. We also find strong differences in behaviour between those students who feel engaged as well as happy. Finding that students who are happy are engaging more is an important result for our understanding of student wellbeing. Coupled with mechanisms to routinely measure engagement, it could assist tutors to identify students who are suffering with poor wellbeing and might benefit from intervention or greater support.

Supporting information

S1 file. questions used in survey completed by participants..

The original survey was completed using survey software Qualtrics.

https://doi.org/10.1371/journal.pone.0225770.s001

Acknowledgments

This project aims to make effect use of data to help students reach their full academic potential while studying at the University of Exeter.

• View Article
• Google Scholar
• 12. Na KS, Tasir Z, editors. Identifying at-risk students in online learning by analysing learning behaviour: A systematic review. 2017 IEEE Conference on Big Data and Analytics (ICBDA); 2017 16–17 Nov. 2017.
• 14. Wang R, Chen F, Chen Z, Li T, Harari G, Tignor S, et al. StudentLife: assessing mental health, academic performance and behavioral trends of college students using smartphones. Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing; Seattle, Washington. 2632054: ACM; 2014. p. 3–14.
• 15. Kent C, Boulton CA, Williams HTP. Towards Measurement of the Relationship between Student Engagement and Learning Outcomes at a Bricks-and-Mortar University. Sixth Multimodal Learning Analytics (MMLA) Workshop and the Second Cross-LAK Workshop co-located with 7th International Learning Analytics and Knowledge Conference (LAK 2017); Vancouver, Canada2017.
• 16. Wang R, Harari G, Hao P, Zhou X, Campbell AT. SmartGPA: how smartphones can assess and predict academic performance of college students. Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing; Osaka, Japan. 2804251: ACM; 2015. p. 295–306.
• 19. Dekker GW, Pechenizkiy M, Vleeshouwers JM. Predicting Students Drop Out: A Case Study. 2nd Educational Data Mining; Cordoba, Spain: ERIC; 2009.
• 23. Sclater N, Peasgood A, Mullan J. Learning Analytics in Higher Education: A review of UK and international practice. Joint Information of Systems Committee (JISC). CC by 4.0 Licence: UK; 2016.
• 24. Sclater N, Mullan J. Learning analytics and student success: Assessing the evidence. Joint Information of Systems Committee (JISC). CC by 4.0 License: UK; 2016.
• 26. Shum SB, Crick RD, editors. Learning dispositions and transferable competencies: pedagogy, modelling and learning analytics. Proceedings of the 2nd international conference on learning analytics and knowledge; 2012: ACM.
• PubMed/NCBI
• 38. Thorley C. Not By Degrees: Not by degrees: Improving student mental health in the UK’s universities. IPPR; 2017.
• 40. Lam SF, Jimerson SR. Exploring student engagement in schools internationally: Consultation paper. Chicago, IL: International School Psychologist Association; 2008.
• 42. Lawther S, Foster E, Mutton J, Kerrigan M. Can the Use of Learning Analytics Encourage Positive Student Behaviours? In: Janes G, Nutt D, Taylor P, editors. Student Behaviour and Positive Learning Cultures: SEDA; 2016. p. 13–21.

Collaborative Learning in Higher Education: Evoking Positive Interdependence

Karin scager.

† Department of Social Sciences, Utrecht University, 3508 TC Utrecht, The Netherlands

Johannes Boonstra

‡ Department of Biology, Utrecht University, 3584 CH Utrecht, The Netherlands

Ton Peeters

Jonne vulperhorst, fred wiegant.

This study focuses on factors increasing the effectiveness of collaborative learning. Results show that challenging, open, and complex group tasks that required the students to create something new and original evoked effective collaboration.

Collaborative learning is a widely used instructional method, but the learning potential of this instructional method is often underused in practice. Therefore, the importance of various factors underlying effective collaborative learning should be determined. In the current study, five different life sciences undergraduate courses with successful collaborative-learning results were selected. This study focuses on factors that increased the effectiveness of collaboration in these courses, according to the students. Nine focus group interviews were conducted and analyzed. Results show that factors evoking effective collaboration were student autonomy and self-regulatory behavior, combined with a challenging, open, and complex group task that required the students to create something new and original. The design factors of these courses fostered a sense of responsibility and of shared ownership of both the collaborative process and the end product of the group assignment. In addition, students reported the absence of any free riders in these group assignments. Interestingly, it was observed that students seemed to value their sense of achievement, their learning processes, and the products they were working on more than their grades. It is concluded that collaborative learning in higher education should be designed using challenging and relevant tasks that build shared ownership with students.

INTRODUCTION

Students may learn a lot from working in groups, but the learning potential of collaboration is underused in practice ( Johnson et al ., 2007 ), particularly in science education ( Nokes-Malach and Richey, 2015 ). Collaborative, cooperative, and team-based learning are usually considered to represent the same concept, although they are sometimes defined differently ( Kirschner, 2001 ); we consider these concepts comparable and use the term “collaboration” throughout the paper. In collaborative learning, students participate in small-group activities in which they share their knowledge and expertise. In these student-driven activities, the teacher usually acts as a facilitator ( Kirschner, 2001 ).

Several decades of empirical research have demonstrated the positive relationship between collaborative learning and student achievement, effort, persistence, and motivation (for reviews, see Slavin, 1990 ; Webb and Palinscar, 1996 ; Barron, 2000 ; Johnson et al ., 2007 ). Collaborative learning potentially promotes deep learning, in which students engage in high-quality social interaction, such as discussing contradictory information ( Visschers-Pleijers et al ., 2006 ). In science education, a deep-learning approach is crucial for understanding concepts and complex processes ( Van Boxtel, 2000 ). Understanding of these concepts involves a process of conceptual change, a process particularly activated in collaborative learning, whereby students interact by explaining to and questioning one another critically ( Van Boxtel et al ., 2000 ; Linton et al ., 2014 ). In previous papers, we have explored and emphasized the relevance of collaborative learning in undergraduate biology courses ( Wiegant et al ., 2012 , 2014 ). By comparing university student achievement in a biology course in individual and group settings, Linton et al . (2014) found that students in group settings achieved significantly better with respect to conceptual understanding in comparison with students in courses with an individual setting. Besides these cognitive benefits, collaborative learning provides social skills needed for future professional work in the field of science.

Just forming groups, however, does not automatically result in better learning and motivation ( Salomon and Globerson, 1989 ; Gillies, 2004 ; Khosa and Volet, 2013 ). In their study of university students’ preferences for collaborative learning, Raidal and Volet (2009) found an overwhelming preference for individual forms of learning. Students are hesitant about group work because of the occurrence of “free riders,” logistical issues, or interpersonal conflicts ( Livingstone and Lynch, 2000 ; Aggarwal and O’Brien, 2008 ; Pauli et al ., 2008 ; Shimazou and Aldrich, 2010 ; Hall and Buzwell, 2012 ). As a result, students might opt for a strategic approach by dividing the work and merely using a stapler to “integrate” their work into a group paper. Johnson and Johnson (1999) refer to groups showing this kind of superficial behavior as “pseudo learning groups.” In turn, the resulting lack of synthesis can be disappointing for teachers. Dividing work also implies that students lose the potential learning effect of collaborating, since the extent to which students benefit from working with other students depends on the quality of their interactions ( Van Boxtel et al ., 2000 ; King, 2002 ; Palinscar and Herrenkohl, 2002 ; Volet et al ., 2009 ). Insight into factors that facilitate collaborative learning is critical for understanding how collaboration can be used effectively in higher education. Therefore, in the present study, we explore factors that optimize the quality of collaboration, using examples of effective group work in five different life sciences courses.

POTENTIAL FACTORS ENHANCING THE EFFECTIVENESS OF COLLABORATIVE LEARNING

Social interaction is crucial for effective collaboration ( Volet et al ., 2009 ). Learning outcomes of collaborative-learning groups have been found to depend on the quality of student discussions, including argumentation ( Teasley, 1995 ; Chinn et al ., 2000 ), explaining ideas to one another ( Veenman et al ., 2005 ), and incorporating and building on one another’s ideas ( Barron, 2003 ). These interactions with peers are assumed to promote students’ cognitive restructuring ( Webb, 2009 ). Explaining things to one another and discussing subject matter may lead to deeper understanding, to recognition of misconceptions, and to the strengthening of connections between new information and previously learned information ( Wittrock, 1990 ). The question of how to organize collaboration in a way that promotes these kinds of interactions is paramount.

Decades of research on group work have resulted in the identification of various factors that potentially enhance the effectiveness of collaboration. These factors can be differentiated as primary factors (design characteristics) and secondary or mediating factors (group-process characteristics). Regarding primary factors, groups need to be small (three to five students) to obtain meaningful interaction ( Lou et al ., 2001 ; Johnson et al ., 2007 ). With respect to group composition, mixed-ability groups have been found to increase performance for students of lower ability, but this composition does not necessarily benefit high-ability students ( Webb et al ., 2002 ). Equal participation, however, has been shown to be more important for students’ achievement than group composition, because students are more likely to use one another’s knowledge and skills fully when all students participate to the same extent ( Woolley et al ., 2015 ). Heterogeneity, with respect to diversity of perspectives and styles, has been found to increase learning, particularly in groups working on tasks that require creativity ( Kozhevnikov et al ., 2014 ). The nature of the task has been shown to be an important factor as well. Open and ill-structured tasks promote higher-level interaction and improve reasoning and applicative and evaluative thinking to a greater extent than closed tasks ( Gillies, 2014 ). In addition, complex tasks provoke deeper-level interactions than simple tasks ( Hertz-Lazarowitz, 1989 ).

Concerning secondary or intermediate factors affecting group work, positive interdependence theory is one of the best-founded theories explaining the quality of interaction in collaborative learning ( Slavin, 1990 ; Johnson and Johnson, 1999 , 2009 ; Gully et al ., 2002 ). According to this theory, collaboration is enhanced when positive interdependence exists among group members. This is achieved when students perceive the contribution of each individual to be essential for the group to succeed in completing the assigned activity ( Johnson and Johnson, 2009 ). Positive interdependence results in both individual accountability and promotive interaction. Individual accountability is defined as having feelings of responsibility for completing one’s own work and for facilitating the work of other group members. A sense of mutual accountability is necessary to avoid free riding ( Johnson and Johnson, 2009 ), which occurs when one or more group members are perceived by other members as failing to contribute their fair share to the group effort ( Aggarwal and O’Brien, 2008 ). Promotive interaction has been described as students encouraging and facilitating one another’s efforts to accomplish group goals, both with respect to group dynamics and the subject matter ( Johnson and Johnson, 2009 ).

Methods of inducing positive interdependence interaction are either reward or task based ( Johnson et al ., 2007 ). Reward-based interdependence structures the reward in such a way that students’ individual grades depend on the achievement of the whole team. According to Slavin (1991 , 1995 ), collaborative learning is rarely successful without group rewards. In higher education, however, findings on the effects of reward-based interdependence are inconclusive. The main concern is that rewards stimulate extrinsic motivation and may be detrimental to intrinsic motivation ( Parkinson and St. George, 2003 ). Intrinsically motivated students put effort into a task because they are interested in the task itself, while extrinsically motivated students are interested in the reward or grade ( Deci and Ryan, 2000 ). Strong incentives, such as grades, could steer student motivation toward the reward and subsequently reduce the task to being a means to an end. Serrano and Pons (2007) , however, found that using rewards (individual grades) created high positive interdependence in group work at a university level. They concluded that the reward structure did direct students’ motivation toward final grades, while the task still aroused the interest of the students. In contrast, Sears and Pai (2012) found that rewards were not crucial factors affecting group behavior. Their study showed that groups continued to work even after the reward was removed, whereas the efforts of students working individually decreased after the reward was removed.

In structured task-based interdependence, students are forced to exchange information; this can be achieved by assigning group members different roles, resources, or tasks (the “jigsaw” method) or by “scripting” the process, which involves giving students a set of instructions on how they should interact and collaborate ( Kagan, 1994 ; Dillenbourg, 2002 ). The effects of task structuring on collaborative learning are, however, not clear ( Fink, 2004 ; Hänze and Berger, 2007 ; Serrano and Pons, 2007 ). Hänze and Berger (2007) observed no differences in achievement between students who worked in jigsaw-structured groups and students who worked individually. In contrast, the observations of Brewer and Klein (2006) indicated that students in groups with given roles plus rewards interacted significantly more frequently than students in groups with given rewards only or in groups without structured interdependence factors. (Over)structuring interaction processes, on the other hand, could threaten intrinsic motivation and disturb natural interaction processes ( Dillenbourg, 2002 ). Although it is widely accepted that positive interdependence has been shown to be crucial in evoking social interaction, in practice, university students often tend to merely go through the motions and choose the solution requiring the least effort, which explains why positive interdependence often does not emerge ( Salomon and Globerson, 1989 ). Additional methods are necessary to encourage quality interactions that enhance learning. Moreover, the mixed results of university education studies concerning structuring interdependence—using either rewards or task structuring—do not solve the challenge of how to create interdependence without disturbing the intrinsic motivation of students. Forcing students to interact could endanger student autonomy and motivation, while merely putting students together has been shown to be ineffective.

THE CURRENT STUDY

Despite the considerable amount of research on collaborative learning, less is known about how to structure university-level group work in order to capitalize on the benefits of collaborative learning. The studies discussed earlier focused on primary and secondary education and are not fully applicable to higher education, because students in undergraduate classes may have different schedules and often have not met before. Moreover, group work of university students is mostly organized outside class hours in the absence of teachers. Furthermore, literature in this area may be limited in applicability, as many studies of factors affecting collaboration have used (quasi)experimental designs, in which outcomes of two or three designs were compared ( Johnson and Johnson, 2009 ). A restriction of this method is that only the hypothesized independent variables are studied, while other important factors contributing to effectiveness might be overlooked. In our study, we approached the theme retrospectively, investigating the learning of student groups known to have collaborated and achieved highly, according to their teachers. Rather than focusing on learning outcomes, we explored how group work in these courses was structured. Understanding the factors that facilitate students’ collaboration is critical to understanding how this approach to learning can be used more effectively in higher education. We explicitly focused on positive examples of effective collaborative learning, as best practices should be communicated to others ( Dewey, 1929 , p.11).

In the current study, we selected five different life sciences undergraduate courses that comprised successful group-work assignments. The specific question this study aimed to address was, according to the students, what factors increased collaboration in these courses? By uncovering the factors that make collaborative learning fruitful, we aim to provide useful guidelines for instructors implementing collaborative learning.

Participants

The present study involved focus group interviews with nine groups of second- and third-year students of five different undergraduate life sciences courses. We depended heavily on these focus group interviews to develop our understandings. They allowed us to gain insight into students’ perspectives, which is important because, to a large degree, students’ perspectives of instruction affect what they do and learn ( Shuell, 1996 ). Furthermore, the group exchanges of experiences and perspectives promoted breadth, as well as depth, in our understandings of the cognitive, behavioral, and situational factors contributing to the effectiveness of the collaboration. The particular courses were selected because they all implemented group work that, according to teacher assessments and student evaluations, was very effective. We approached the instructors of these courses with the request to ask their students to volunteer in focus group discussions. Students were willing to participate in these focus group discussions, although not all students were able to meet at the scheduled times. No specific reward was promised for participating in focus group discussions.

Between two and 10 students participated in each of the nine focus group interviews (see Table 1 ).

Course, number of focus group interviews, and students per interview

Course Descriptions

We focused on five courses that were all small-enrollment, upper-division courses in which 15–35 students participated per course. In all courses, collaborative activities occurred during class hours but also outside of class. In some courses, the out-of-class cooperative activities even exceeded the in-class activities.

Course A: The first course was part of a biology honors program. In this part of the program, groups of second-year bachelor’s students (12–19 students) were assigned the group task of writing a popular science book about a biology topic of their choice. Students had to perform all the activities necessary to produce the book. The project was strongly student-led, and students assigned themselves tasks necessary for finishing the project. The assignment comprised an entire academic year, starting in September and finishing in May/June as an extracurricular activity. More details of this course are described elsewhere ( Wiegant et al ., 2012 ).

Course B: Students in the immunology course, mostly third-year students, were assigned the task of writing, in groups of four, a short research project on an immunological topic. The assignment was structured in three parts: in part 1, groups designed a draft of their proposal; in part 2, the groups peer reviewed the draft of another group; and in part 3, the groups received the draft and comments of yet another group, which they had to finish and present. The assignment comprised approximately half of the course.

Course C: In the advanced cell biology course, three small teams of four or five students collaborated intensively during a semester of 15 weeks to formulate three PhD proposals within an overarching theme. Because the course was student-led, the teachers refrained from guiding the students in their decisions, instead taking a facilitating role by asking critical questions and providing feedback. As a result of the project, the teams presented and defended their research program and the three research proposals before a jury of experts. More details of this course are given elsewhere ( Wiegant et al ., 2011 , 2014 ; Scager et al ., 2014 ).

Course D: The objective of the molecular cell biology course was to learn to design a research project in groups of four. In this course, students were required to complete multiple assignments, such as reviewing a paper, developing a research proposal, designing experiments, and writing and defending their proposals. Groups met with their supervisor once a week and were supposed to keep the course coordinator informed on their progress. Final grades were based on individual (40%) and group (60%) components.

Course E: As a part of the pharmacy course, third-year students, in groups of four to six participants, were required to analyze the quality of a specific pharmacotherapy. The assignments were authentic and were provided by external commissioning companies. The group assignment counted for 70% of the final grade (50% group report and presentation; 20% individual reflection).

The interviews were semistructured and included two basic questions: 1) “What factors made group work effective in this course (as opposed to other experiences you have had)?” and 2) “What was the added value in this course of working in a group (as opposed to working individually)?” The addition of “as opposed to …” was aimed to encourage students’ thinking process; we did not ask students to elaborate on these opposing experiences. Interviewers stimulated and moderated discussions, ensuring depth as well as diversity. To focus and structure the interviews and to stimulate the sharing of discussion outcomes, we listed the answers to the two questions on a flip chart.

First, the intentions of the interview were clarified, followed by an explanation of the confidential nature of the interview. All students agreed and gave permission for the interviews to be audiotaped. All of the authors conducted one or more interviews, with the first author (K.S.) moderating them. The focus group interviews were held in or near the classroom associated with each of the specific courses. The interviews were ∼60 minutes each and were transcribed verbatim.

Detecting Factors That Facilitated Group Work.

Data were analyzed by the first and fourth authors (K.S. and J.V.) in three partially overlapping stages. Stage 1 comprised reading and rereading the transcripts to identify text units relevant to the subject of challenge. Given the aim of the focus group interviews, this meant ignoring small talk and sorting discussion units related to the two interview questions into focal issues. Stage 2 comprised identifying and coding themes related to the two main interview questions regarding 1) factors and 2) added value, using NVivo version 10 (a qualitative data-analysis computer software package). First, open coding was applied. The answers to both questions, however, evoked answers that pointed to intermediary variables affecting the outcomes of collaboration. For example, the question regarding factors brought forward the importance of the assignment being complex enough to make students feel mutually interdependent, while for the question regarding added value, students referred back to how the complexity of the assignment stimulated them to discuss, build on, and learn from one another’s ideas. The interactions provoked by the complexity of the task seemed to connect complexity with learning outcomes. Therefore, when axial coding was applied, we decided to develop three clusters of codes focused on the factors of effective collaboration, the mediating variables, and the added value of collaboration. Subsequently, selective coding was applied, wherein codes were clustered into larger sets informed by theory ( Braun and Clarke, 2006 ). Only factors that were mentioned in more than half of the focus groups were kept. This resulted in two sets of factors. The first set of factors related to the design of the group assignment (autonomy, group size, task design, and teacher expectations). The second set consisted of mediating variables related to the working processes of the groups (team and task regulation, promotive interaction, interdependence, responsibility, and mutual support and motivation).

Reliability and Validity.

Reliability is considered in terms of equivalence and internal consistency ( Sim and Wright, 2000 ). Reliability was ensured by intercoder consistency ( Burla et al ., 2008 ). Given the complexity and inhomogeneity of group discourse, agreement testing was constrained to core concepts or themes of substantive importance ( Kidd and Parshall, 2000 ). The equivalence of coding was addressed by selecting 20% of the data and comparing the coding of two secondary raters (10% each) for consistency, which yielded a kappa coefficient of 0.85. This strength of agreement is considered to be “nearly perfect” ( Everitt, 1996 ). Internal consistency was acquired by having one team member moderating all (but one) of the interviews ( Kidd and Parshall, 2000 ). The emergence of substantively similar viewpoints of the focus groups on the core issues across the five different courses supported content validity ( Kidd and Parshall, 2000 ). Furthermore, we assessed content validity by independent coding and by comparing this with theory in extant literature ( Morgan and Spanish, 1985 ; Torn and McNichol, 1998 ).

Factors That Contributed to the Effectiveness of the Collaboration

Eight factors were found to have a positive effect on the effectiveness of the collaboration. These factors are presented in Table 2 : 1) design factors: the design of the course and/or the assignment (the autonomy of the students, task characteristics, teacher expectations, and group size); and 2) process factors: the way students interacted and organized their work (team and task regulation, interdependence, promotive interaction, and mutual support and motivation).

Factors that contributed to the effectiveness of the collaboration

a “Source” refers to how many of the nine interviews the topic was discussed in; “reference” refers to the total number of times the topic was discussed.

Table 2 shows that autonomy and the density and complexity of the task were the factors most frequently mentioned by the students as contributing to the effectiveness of the collaboration. Team and task regulation, positive interdependence, and promotive interaction were perceived by students as the most important factors with respect to the way they processed the assignments. In the next section, we describe the results more elaborately, starting with the design features of these courses that are considered to enhance collaboration processes.

Design Factors

The autonomy the groups experienced was mentioned in all focus groups, indicating the importance of this factor to the effectiveness of collaboration. Autonomy was manifested in allowing student groups to choose their own topics (e.g., for their research plans) and giving them independence in organizing their processes. Statements such as “It was our own thing” occurred frequently in all nine focus group discussions. The references to “our thing” indicate that the students made choices as a group, which could have restricted individual feelings of autonomy. The students, however, did not seem to have experienced clear boundaries between individual and group autonomy. Even though their personal ideas may have been overruled by the team, they still felt autonomous, because they made decisions democratically. As one of the students said, “When you participate in the decision process it is easier to accept than when the decision is made by the teacher.”

Two features of the task were perceived as important contributors to the effectiveness of the group work. First, the density and complexity of the task was crucial. The group task needed to be extensive enough for the group members to really need one another’s contributions to finish in time and complex enough to require them to discuss their work and provide one another with feedback. Second, students perceived the relevance of the task at hand to be an important feature. The task relevance was found in different aspects, depending on the assignment. For the biology honors groups, for example, the process of writing a popular science book and getting it published increased their feelings of doing something significant. The cell biology and immunology groups emphasized the relevance of doing research, in terms of formulating a relevant proposal in the same way as it is done “in the real world.”

In terms of rewards , students emphasized that the inherent value of the end product, such as an article, a research proposal, or a book, stimulated them to achieve, which relates back to the perceived task relevance. As a student of the biology honors course said, “We have also had other group projects …, but that was taken less seriously, because you, well it was nice, but well, the result wouldn’t reach beyond the classroom, while in this project it will.” There were no grades involved in this particular course, which students appreciated, because they believed the end product to be more important than a grade. Also, in other groups, discussions about assessment were learning and/or reward oriented rather than grade oriented; for example, in one of the pharmacy groups it was said: “You are in a learning process, and I think sometimes that it is a shame that it should end in a grade—that creates a tension. And if things go wrong, that could be very beneficial for your learning, but it can also happen that you do not receive a high grade for it.”

In all of the interviews, students mentioned that it was crucial that the task was the core project in the course at that time, as students of the immunology course stated: “I think also because this is not something you do on the side, but this is the only thing we do at the moment, it is the main activity.” The fact that students’ final grades depended primarily on the group assignment was mentioned in some groups. Students emphasized that in previous experiences with group assignments they had not collaborated as intensively because their final grade did not depend largely on the team assignment.

Finally, group size was considered a factor stimulating collaboration in seven of the groups, specifically related to the level of responsibility students felt. Groups of three or four were believed to be optimal: “Otherwise, you get a sort of diffuse responsibility …, and with four you are clearly responsible for an important part of the process.”

Process Factors

The need for team and task regulation was mentioned most frequently in the focus group discussions as an important factor increasing the effectiveness of collaboration. Students divided tasks, appointed team leaders, and set their own deadlines. Organizing frequent face-to-face meetings was very helpful, according to students: “That we met each other physically, instead of doing everything by mail or chat, like in other projects. This works much better, if you can look each other in the eyes it is way faster and more efficient to manage and decide things …. It also increases the pressure, everybody prepares for a meeting.” The quote in Table 2 indicates the direct relation between the autonomy of the groups and their dedication to following their self-made group regulations.

As shown in Table 2 , students in all nine focus groups experienced a sense of positive interdependence in terms of needing one another in order to succeed and achieve their goal. The feeling of responsibility was discussed in six groups. The related issue of “uneven contribution” was discussed in all nine of the focus groups: students did experience differences in power and effort between team members. Interestingly, students did not perceive this as free riding. According to the students, some degree of uneven contribution is only natural; the students all did their best, but as the students said, “There weren’t students who contributed less; there were only students who contributed more.” According to the students, this uneven contribution was due to power differences, not to disinterest or laziness. Students showed empathy for their peers who contributed less: “The strong people might go too hard for the other people to be able to catch up.” This may have caused frustration in students who felt they were lagging behind, as one of them revealed: “You have that responsibility that drives you and then you feel the need to do more, but perhaps that is beyond your capabilities at that point.” Some of the groups discussed the issue of uneven contribution while working on their projects, but always, they stated, in an “understanding and respectful way.” Furthermore, students in all nine interviews mentioned the fact that the variety among students was useful and enhanced the discussions: “working in a group consisting of clones of yourself” would not be as interesting, one of the pharmacy groups stated.

All nine groups mentioned the need for promotive interaction several times, drawing attention to the need to discuss content to accomplish team goals. They mentioned several indicators of promotive interaction: discussions, exchange of information, and arguments, building on one another’s ideas, explaining to one another, providing and processing peer feedback, and asking one another critical questions. According to the students, these discussions enhanced their understanding, and they also learned how to discuss, voice their opinion, explain, listen to others, accept feedback, and reflect on their own work.

Last, but not least, students talked enthusiastically about the way they supported and motivated one another. There was explicit help and pep talks, and, perhaps even more importantly, implicit mutual inspiration effected by them perceiving the motivation of their peers.

Finally, we found one contextual factor (not included in Table 2 ) contributing to collaboration: the shared motivation of students to get the best out of the group assignment. Students mostly linked their having similar motivations to the fact that they were in their second or third year (four of the five courses were third-year courses). First, the students already knew one another: “When you are in your first year, you do not know each other, and some people are a bit insecure, so to say. But now we know each other, so we may scold each other all we can.” Furthermore, students suggested being equally motivated, because the unmotivated students had already left in previous years.

CONCLUSIONS AND DISCUSSION

The purpose of the current study was to find factors that enhance student collaboration. The collaboration processes (task and team regulation, mutual support and motivation, positive interaction) used by these students were distinctly effective. During these processes, positive interdependence was clearly present, supporting the notion that positive interdependence is a crucial factor affecting the effectiveness of collaboration ( Johnson and Johnson, 2009 ). Although the interview data do not allow causal relations between design factors and collaboration processes to be inferred, it seems reasonable to assume that positive interdependence was evoked by a combination of the nature of the task (autonomous, relevant, dense and complex, group rewards), the prominent placement of the group assignment within the course, and the group size.

The results indicate that positive interdependence was an important factor contributing to the effectiveness of collaboration. The positive effect of interdependence on student achievement has already been well documented (for reviews, see Slavin, 1990 ; Webb and Palinscar, 1996 ; Johnson et al ., 2007 ). Although we disassembled the factors contributing to collaboration in the analysis , we assume interdependence does not consist of a single factor but rather is constructed through the interaction between motivated students and design factors (the nature of the task and student autonomy). Furthermore, the fact that the final grade depended primarily on the group assignment can be expected to have contributed to students’ interdependence, which would concur with the findings of Slavin (1991) . Interestingly, however, these students seemed to value the learning process and the products they were working on more than their grades. Our finding, that a sense of achievement rather than a grade was of greater importance in motivating interdependence, contradicts findings of Slavin (1991) and Tsay and Brady (2010) . Tsay and Brady (2010) found that the degree of active participation of university students in collaborative groups was affected by the importance they attached to grades: students who perceived grades as highly important were more active collaborators.

The enthusiasm of the students when speaking of the way they supported and motivated one another and regulated the team and task processes indeed indicates the occurrence of strong self-regulatory processes. Although some structure was provided beforehand in all five courses (e.g., final deadlines), students were perceived to be autonomous in the planning and regulation of their work, which they said added to their motivation to follow their own rules and planning. This direct relationship between perceived autonomy and self-regulatory behavior is aligned with self-determination theory ( Deci and Ryan, 2000 ). According to Deci and Ryan (2000) , when teachers are supportive of student autonomy, students are motivated to internalize the regulation of their learning activities, whereas when teachers are controlling, self-regulated motivation is undermined. The self-regulatory social processes of these students, encouraged by the autonomy they were provided, were the most important factors increasing the effectiveness of their collaboration in these five cases.

Individual accountability is an important aspect within the theory of positive interdependence. Interestingly, instead of accountability, students used the word “responsibility.” The difference between responsibility and accountability is meaningful, because accountability is focused on the end result, or being answerable for your actions to relevant others, while responsibility is related to the task. Responsibility is viewed as having a higher level of autonomy and involves the ability to self-regulate actions free of external motivational pressure. In contrast, the accountable actor is subject to external oversight, regulation, and mechanisms of punishment ( Bivins, 2006 ). The term “responsibility” more appropriately fits the collaboration in these cases, as one of our participants illustrates: “You feel the responsibility to other people in your group, because as soon as soon as you drop the ball, the rest have to work harder.” This student does not refer to consequences externally imposed on him, but he feels responsibility toward others. The effect this has may be the same as when students are forced to be accountable because of reward- or task-based structures, as suggested by Johnson and Johnson (2009) ; however, the nature of the motivation is more intrinsically than extrinsically induced.

Related to the issue of accountability or responsibility is the problem of free riding, which is one of the main problems of group work in higher education ( Livingstone and Lynch, 2000 ; Aggarwal and O’Brien, 2008 ; Pauli et al ., 2008 ; Shimazou and Aldrich, 2010 ). In the interviews in which the issue of free riding came up, however, groups did not seem to have experienced the phenomenon. A putative explanation for the lack of free-riding behavior is the incidence of accountability ( Slavin, 1991 ; Johnson and Johnson, 2009 ; Onwegbuezie et al ., 2009 ), as students definitely felt responsible for the end result. The way students spoke about their group members, however, was in terms of mutual trust rather than accountability. Students recognized differences in contribution but did not perceive this as problematic. They were empathic toward differences between students. If there were negative feelings at all, the low contributors were more apt to feel frustrated, indicating that the differences in contribution were, as Hall and Buzwell (2012) have suggested, involuntary and due to inadequacy rather than apathy or laziness.

In the five courses of this study, the combination of design factors seems to have prevented free riding. Although the causal nature of the relationship between design features of the group work and effective group processing cannot be claimed in the current study, the results indicate that, in particular, perceived autonomy and the challenging nature of the task evoked students’ motivation to make an effort. The relevance of the tasks, which required students to produce something new (to them) and something original and tangible, motivated students. The tasks were also open and complex, which are features that have already been found to promote deeper-level interactions than simple tasks ( Hertz-Lazarowitz, 1989 ; Cohen, 1994 ). Autonomy was a factor frequently mentioned as contributing to the effectiveness of the group work. Contradictory to Johnson and Johnson’s (2009) recommendation for teachers to structure processes, students of these courses designated the autonomy they had in choosing their topic and in organizing the process, as one of the factors increasing their motivation. Results from organizational research show that autonomy can, in fact, increase teamwork achievement, but only when positive interdependence is high ( Langfred, 2000 ). Autonomy combined with low interdependence decreases achievement, indicating that autonomy should be combined with challenging tasks. Although autonomy and level of challenge in a group assignment appears to be vital, instructors in different settings may need to use greater scaffolding.

Future Research and Concluding Remarks

It is important to keep in mind the small sample and restricted context when interpreting these findings. Although the results have been obtained in small-enrollment, upper-division courses, we think that our findings might also be transferable to large-enrollment courses, provided students will be working in self-directed small groups on substantial and relevant projects. As generalizability requires data on large populations, the findings of our five cases within a restricted context are not necessarily representative of the larger population. We believe, however, that there are strong reasons for our findings to be deemed “transferable” ( Lincoln and Guba, 1985 ) to comparable situations. While generalization is applied by researchers, transferability is a process performed by the readers of research ( Metcalfe, 2005 ). Unlike generalizability, transferability does not involve broad claims but invites readers of research to make connections between elements of a study and their own experiences ( Barnes et al ., 2012 ). According to Berliner (2002 , p. 19), implementing scientific findings is always difficult in education, “because humans in schools are embedded in complex and changing networks of social interaction.” Therefore, we do not claim to have produced broadly generalizable findings but instead invite the reader to identify how the findings can be transferred to his or her situation. Similar studies with data from other university contexts, such as other countries or other class settings, would help in understanding how the conditions that facilitate collaborative learning relate to different settings.

We assume, however, that the concept of evoking, rather than enforcing, positive interdependence by increasing autonomy and the challenge level of the task provides relevant insights for discourse on effective design of group work within life sciences education. Students in life sciences education, in general, are quite experienced in working in groups and in regulating their own work. Autonomy, combined with a challenging task, evoked interdependence and generated interaction as well as student motivation in these five cases. Structuring the process, for example by scripting, seems unnecessary for promoting student interaction. It was, in Dillenbourg’s (2002) words, not necessary to “didactisise” collaborative interactions or to disturb the autonomy and natural interactions of students. Moreover, structuring the process could have impeded the feeling of autonomy, which is crucial for student motivation (Deci and Ryan, 2000). Brewer and Klein (2006) came to a similar conclusion in their investigation of the influence of types of interdependence (roles, rewards, roles plus rewards, no structure) on student interaction. The groups with no structured interdependence had significantly more cognitive interactions involving content discussion than the other groups, indicating that structuring interdependence is not always necessary with university students. We suggest that collaborative learning with university students should be designed using challenging and relevant tasks that build shared ownership with students.

Acknowledgments

Drs. Kristin Denzer, Mario Stassen, and Fons Cremers are gratefully acknowledged for encouraging their students to participate in the interviews.

• Aggarwal P, O’Brien C. Social loafing on group projects: structural antecedents and effect on student satisfaction. J Market Educ. 2008; 30 :255–264. [ Google Scholar ]
• Barnes J, Conrad K, Demont-Heinrich C, Graziano M, Kowalski D, Neufeld J, Zamora J, Palmquist M. Generalizability and transferability. 2012. Writing@CSU, Colorado State University http://writing.colostate.edu/guides/guide.cfm?guideid=65 (accessed 14 November 2015).
• Barron B. Achieving coordination in collaborative problem-solving groups. J Learn Sci. 2000; 9 :403–436. [ Google Scholar ]
• Barron B. When smart groups fail. J Learn Sci. 2003; 12 :307–359. [ Google Scholar ]
• Berliner DC. Educational research: the hardest science of all. Educ Res. 2002; 31 :18–20. [ Google Scholar ]
• Bivins TH. Responsibility and accountability. In: Fitzpatrick K, Bronstein C, editors. Ethics in Public Relations: Responsible Advocacy. Thousand Oaks, CA: Sage; 2006. pp. 19–38. [ Google Scholar ]
• Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006; 3 :77–101. [ Google Scholar ]
• Brewer S, Klein JD. Type of positive interdependence and affiliation motive in an asynchronous collaborative learning environment. Educ Technol Res Dev. 2006; 54 :331–354. [ Google Scholar ]
• Burla L, Knierim B, Barth KL, Duetz M, Abel T. From text to codings: intercoder reliability assessment in qualitative content analysis. Nurs Res. 2008; 57 :111–117. [ PubMed ] [ Google Scholar ]
• Chinn CA, O’Donnell AM, Jinks TS. The structure of discourse in collaborative learning. J Exp Educ. 2000; 69 :77–97. [ Google Scholar ]
• Cohen EG. Restructuring the classroom: conditions for productive small groups. Rev Educ Res. 1994; 64 :1–35. [ Google Scholar ]
• Deci EL, Ryan RM. The “what” and “why” of goal pursuits: human needs and the self-determination of behavior. Psychol Inq. 2000; 11 :227–268. [ Google Scholar ]
• Dewey J. The Sources of a Science of Education. New York: Liveright; 1929. [ Google Scholar ]
• Dillenbourg P. Over-scripting CSCL: the risks of blending collaborative learning with instructional design. In: Kirschner PA, editor. Three Worlds of CSCL. Can We Support CSCL. Heerlen: Open University Nederland; 2002. pp. 61–91. [ Google Scholar ]
• Everitt B. Making Sense of Statistics in Psychology: A Second-Level Course. Oxford, UK: Oxford University Press; 1996. [ Google Scholar ]
• Fink DL. Beyond small groups: harnessing the extraordinary power of learning teams. In: Michaelsen LK, Bauman Knight A, Fink LD, editors. Team-Based Learning: A Transformative Use of Small Groups in Teaching. Sterling, VA: Stylus; 2004. pp. 3–26. [ Google Scholar ]
• Gillies R. Cooperative learning: developments in research. Int J Educ Psychol. 2014; 3 :125–140. [ Google Scholar ]
• Gillies RM. The effects of cooperative learning on junior high school students during small group learning. Learn Instr. 2004; 14 :197–213. [ Google Scholar ]
• Gully S, Incalcaterra K, Joshi A, Beaubien J. A meta-analysis of team-efficacy, potency, and performance: interdependence and level of analysis as moderators of observed relationships. J Appl Psychol. 2002; 87 :819–832. [ PubMed ] [ Google Scholar ]
• Hall D, Buzwell S. The problem of free-riding in group projects: looking beyond social loafing as reason for non-contribution. Active Learn High Educ. 2012; 14 :37–49. [ Google Scholar ]
• Hänze M, Berger R. Cooperative learning, motivational effects, and student characteristics: an experimental study comparing cooperative learning and direct instruction in 12th grade physics classes. Learn Instr. 2007; 17 :29–41. [ Google Scholar ]
• Hertz-Lazarowitz R. Cooperation and helping in the classroom: a contextual approach. Int J Res Educ. 1989; 13 :113–119. [ Google Scholar ]
• Johnson DW, Johnson RT. Making cooperative learning work. Theory Pract. 1999; 38 :67–73. [ Google Scholar ]
• Johnson DW, Johnson RT. An educational psychology success story: social interdependence theory and cooperative learning. Educ Res. 2009; 38 :365–379. [ Google Scholar ]
• Johnson DW, Johnson RT, Smith K. The state of cooperative learning in postsecondary and professional settings. Educ Psychol Rev. 2007; 19 :15–29. [ Google Scholar ]
• Kagan S. Cooperative Learning, 2nd ed. San Clemente, CA: Kagan; 1994. [ Google Scholar ]
• Khosa DK, Volet SE. Promoting effective collaborative case-based learning at university: a metacognitive intervention. Stud High Educ. 2013; 38 :870–889. [ Google Scholar ]
• Kidd PS, Parshall MB. Getting the focus and the group: enhancing analytical rigor in focus group research. Qual Health Res. 2000; 10 :293–308. [ PubMed ] [ Google Scholar ]
• King A. Structuring peer interaction to promote high-level cognitive processing. Theory Pract. 2002; 41 :33–39. [ Google Scholar ]
• Kirschner PA. Using integrated electronic environments for collaborative teaching/learning. Res Dialogue Learn Instruction. 2001; 2 :1–9. [ Google Scholar ]
• Kozhevnikov M, Evans C, Kosslyn SM. Cognitive style as environmentally sensitive individual differences in cognition: a modern synthesis and applications in education, business and management. Psychol Sci. 2014; 15 :3–33. [ PubMed ] [ Google Scholar ]
• Langfred CW. The paradox of self-management: individual and group autonomy in work groups. J Organ Behav. 2000; 21 :563–585. [ Google Scholar ]
• Lincoln YS, Guba EG. Naturalistic Inquiry. London: Sage; 1985. [ Google Scholar ]
• Linton DL, Farmer JK, Peterson E. Is peer interaction necessary for optimal active learning. CBE Life Sci Educ. 2014; 13 :243–252. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Livingstone D, Lynch K. Group project work and student-centred active learning: two different experiences. Stud High Educ. 2000; 25 :325–345. [ Google Scholar ]
• Lou Y, Abrami P, d’Apollonia S. Small group and individual learning with technology: a meta-analysis. Rev Educ Res. 2001; 71 :449–521. [ Google Scholar ]
• Metcalfe M. Generalisation: learning across epistemologies. Forum: Qualitative Social Research. 2005; 6 (1) www.qualitative-research.net/fqs-texte/1-05/05-1-17-e.htm (accessed 5 January 2016) [ Google Scholar ]
• Morgan DL, Spanish MT. Social interaction and the cognitive organisation of health-relevant behavior. Sociol Health Illness. 1985; 7 :401–422. [ Google Scholar ]
• Nokes-Malach TJ, Richey JE. When is it better to learn together? Insights from research on collaborative learning. Educ Psychol Rev. 2015; 27 :645–656. [ Google Scholar ]
• Onwegbuezie AJ, Collins KMT, Jiao QG. Performance of cooperative learning groups in a postgraduate education methodology course. Active Learn High Educ. 2009; 10 :265–277. [ Google Scholar ]
• Palincsar AS, Herrenkohl LR. Promoting thinking through peer learning. Theory Pract. 2002; 41 :26–32. [ Google Scholar ]
• Parkinson TJ, St. George AM. Are the concepts of andragogy and pedagogy relevant to veterinary undergraduate teaching. J Vet Med Educ. 2003; 30 :247–253. [ PubMed ] [ Google Scholar ]
• Pauli R, Mohiyeddini C, Bray D, Michie F, Street B. Individual differences in negative group work experiences in collaborative student learning. Educ Psychol. 2008; 28 :47–58. [ Google Scholar ]
• Raidal SL, Volet SE. Preclinical students’ predispositions towards social forms of instruction and self-directed learning: a challenge for the development of autonomous and collaborative learners. High Educ. 2009; 57 :577–596. [ Google Scholar ]
• Salomon G, Globerson T. When teams do not function the way they ought to. Int J Educ Res. 1989; 13 :89–99. [ Google Scholar ]
• Scager K, Akkerman SF, Pilot A, Wubbels T. Challenging high ability students. Stud High Educ. 2014; 9 :659–680. [ Google Scholar ]
• Sears DA, Pai H-H. Effects of cooperative versus individual study on learning and motivation after reward-removal. J Exp Educ. 2012; 80 :246–262. [ Google Scholar ]
• Serrano JM, Pons RM. Cooperative learning: we can also do it without task structure. Intercultural Educ. 2007; 18 :215–230. [ Google Scholar ]
• Shimazou J, Aldrich H. Group work can be gratifying: understanding and overcoming resistance to cooperative learning. Coll Teach. 2010; 58 :52–57. [ Google Scholar ]
• Shuell TJ. Teaching and learning in a classroom context. In: Berliner DC, Calfee RC, editors. Handbook of Educational Psychology. New York: MacMillan; 1996. pp. 726–763. [ Google Scholar ]
• Sim C, Wright J. Research in Health Care: Concepts and Methods. Cheltenham, UK: Nelson Thomes; 2000. [ Google Scholar ]
• Slavin RE. Cooperative learning. Rev Educ Res. 1990; 50 :315–342. [ Google Scholar ]
• Slavin RE. Group rewards make groupwork work. Educ Leadership. 1991; 48 (5):89–91. [ Google Scholar ]
• Slavin RE. Cooperative Learning: Theory, Research, and Practice. Boston: Allyn Bacon; 1995. [ Google Scholar ]
• Teasley SD. The role of talk in children’s peer collaboration. Dev Psychol. 1995; 31 :207–220. [ Google Scholar ]
• Torn A, McNichol E. A qualitative study utilizing a focus group to explore the role and concept of the nurse practitioner. J Adv Nurs. 1998; 27 :1202–1211. [ PubMed ] [ Google Scholar ]
• Tsay M, Brady M. A case study of cooperative learning and communication pedagogy: does working in teams make a difference. J Scholarsh Teach Learn. 2010; 10 (2):78–89. [ Google Scholar ]
• Van Boxtel C. PhD Thesis. Utrecht, Netherlands: Utrecht University; 2000. Collaborative concept learning: collaborative learning tasks, student interaction and the learning of physics concepts. [ Google Scholar ]
• Van Boxtel C, Van Der Linden J, Kanselaar G. Collaborative learning tasks and the elaboration of conceptual knowledge. Learn Instr. 2000; 10 :311–330. [ Google Scholar ]
• Veenman S, Denessen E, Van den Akker A, Van der Rijt J. Effects of a cooperative learning program on the elaborations of students during help seeking and help giving. Am Educ Res J. 2005; 42 :114–151. [ Google Scholar ]
• Visschers-Pleijers AJSF, Dolmans DHJM, De Leng BA, Wolfhagen IHAP, Van Der Vleuten CPM. Analysis of verbal interactions in tutorial groups: a process study. Med Educ. 2006; 40 :129–137. [ PubMed ] [ Google Scholar ]
• Volet SE, Summers M, Thurman J. High-level co-regulation in collaborative learning: how does it emerge and how is it sustained. Learn Instr. 2009; 19 :128–143. [ Google Scholar ]
• Webb NM. The teacher’s role in promoting collaborative dialogue in the classroom. Br J Educ Psychol. 2009; 79 :1–28. [ PubMed ] [ Google Scholar ]
• Webb NM, Nemer KM, Zuniga S. Short circuits or superconductors? Effects of group composition on high-achieving students’ science assessment performance. Am Educ Res J. 2002; 39 :943–989. [ Google Scholar ]
• Webb NM, Palinscar AS. Group processes in the classroom. In: Berliner D, Calfee R, editors. Handbook of Educational Psychology. New York: Macmillan; 1996. pp. 841–873. [ Google Scholar ]
• Wiegant F, Boonstra J, Peeters A, Scager K. Team-based learning in honors science education: the benefit of complex writing assignments. J National Collegiate Honors Council. 2012; 13 :219–227. [ Google Scholar ]
• Wiegant F, Scager K, Boonstra J. An undergraduate course to bridge the gap between textbooks and scientific research. CBE Life Sci Educ. 2011; 10 :83–94. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wiegant FAC, Scager K, Peeters AJM, Boonstra J. The challenge of writing a PhD proposal in honors (undergraduate) education: a group project as significant learning experience. In: Wolfensberger MVC, Drayer L, Volker JJM, editors. Munster, Germany: Waxmann; 2014. pp. 77–84. Pursuit of Excellence in a Networked Society: Theoretical and Practical Approaches Coming from the Conference Excellence in Higher Education and Beyond. [ Google Scholar ]
• Wittrock MC. Generative processes of comprehension. Educ Psychol. 1990; 24 :345–376. [ Google Scholar ]
• Woolley AW, Aggarwal I, Malone TW. Collective intelligence and group performance. Curr Dir Psychol Sci. 2015; 24 :420–424. [ Google Scholar ]

• Research article
• Open access
• Published: 24 April 2023

Artificial intelligence in higher education: the state of the field

• Helen Crompton   ORCID: orcid.org/0000-0002-1775-8219 1 , 3 &
• Diane Burke 2  

International Journal of Educational Technology in Higher Education volume  20 , Article number:  22 ( 2023 ) Cite this article

94k Accesses

93 Citations

64 Altmetric

Metrics details

This systematic review provides unique findings with an up-to-date examination of artificial intelligence (AI) in higher education (HE) from 2016 to 2022. Using PRISMA principles and protocol, 138 articles were identified for a full examination. Using a priori, and grounded coding, the data from the 138 articles were extracted, analyzed, and coded. The findings of this study show that in 2021 and 2022, publications rose nearly two to three times the number of previous years. With this rapid rise in the number of AIEd HE publications, new trends have emerged. The findings show that research was conducted in six of the seven continents of the world. The trend has shifted from the US to China leading in the number of publications. Another new trend is in the researcher affiliation as prior studies showed a lack of researchers from departments of education. This has now changed to be the most dominant department. Undergraduate students were the most studied students at 72%. Similar to the findings of other studies, language learning was the most common subject domain. This included writing, reading, and vocabulary acquisition. In examination of who the AIEd was intended for 72% of the studies focused on students, 17% instructors, and 11% managers. In answering the overarching question of how AIEd was used in HE, grounded coding was used. Five usage codes emerged from the data: (1) Assessment/Evaluation, (2) Predicting, (3) AI Assistant, (4) Intelligent Tutoring System (ITS), and (5) Managing Student Learning. This systematic review revealed gaps in the literature to be used as a springboard for future researchers, including new tools, such as Chat GPT.

A systematic review examining AIEd in higher education (HE) up to the end of 2022.

Unique findings in the switch from US to China in the most studies published.

A two to threefold increase in studies published in 2021 and 2022 to prior years.

AIEd was used for: Assessment/Evaluation, Predicting, AI Assistant, Intelligent Tutoring System, and Managing Student Learning.

Introduction

The use of artificial intelligence (AI) in higher education (HE) has risen quickly in the last 5 years (Chu et al., 2022 ), with a concomitant proliferation of new AI tools available. Scholars (viz., Chen et al., 2020 ; Crompton et al., 2020 , 2021 ) report on the affordances of AI to both instructors and students in HE. These benefits include the use of AI in HE to adapt instruction to the needs of different types of learners (Verdú et al., 2017 ), in providing customized prompt feedback (Dever et al., 2020 ), in developing assessments (Baykasoğlu et al., 2018 ), and predict academic success (Çağataylı & Çelebi, 2022 ). These studies help to inform educators about how artificial intelligence in education (AIEd) can be used in higher education.

Nonetheless, a gap has been highlighted by scholars (viz., Hrastinski et al., 2019 ; Zawacki-Richter et al., 2019 ) regarding an understanding of the collective affordances provided through the use of AI in HE. Therefore, the purpose of this study is to examine extant research from 2016 to 2022 to provide an up-to-date systematic review of how AI is being used in the HE context.

Artificial intelligence has become pervasive in the lives of twenty-first century citizens and is being proclaimed as a tool that can be used to enhance and advance all sectors of our lives (Górriz et al., 2020 ). The application of AI has attracted great interest in HE which is highly influenced by the development of information and communication technologies (Alajmi et al., 2020 ). AI is a tool used across subject disciplines, including language education (Liang et al., 2021 ), engineering education (Shukla et al., 2019 ), mathematics education (Hwang & Tu, 2021 ) and medical education (Winkler-Schwartz et al., 2019 ),

Artificial intelligence

The term artificial intelligence is not new. It was coined in 1956 by McCarthy (Cristianini, 2016 ) who followed up on the work of Turing (e.g., Turing, 1937 , 1950 ). Turing described the existence of intelligent reasoning and thinking that could go into intelligent machines. The definition of AI has grown and changed since 1956, as there has been significant advancements in AI capabilities. A current definition of AI is “computing systems that are able to engage in human-like processes such as learning, adapting, synthesizing, self-correction and the use of data for complex processing tasks” (Popenici et al., 2017 , p. 2). The interdisciplinary interest from scholars from linguistics, psychology, education, and neuroscience who connect AI to nomenclature, perceptions and knowledge in their own disciplines could create a challenge when defining AI. This has created the need to create categories of AI within specific disciplinary areas. This paper focuses on the category of AI in Education (AIEd) and how AI is specifically used in higher educational contexts.

As the field of AIEd is growing and changing rapidly, there is a need to increase the academic understanding of AIEd. Scholars (viz., Hrastinski et al., 2019 ; Zawacki-Richter et al., 2019 ) have drawn attention to the need to increase the understanding of the power of AIEd in educational contexts. The following section provides a summary of the previous research regarding AIEd.

Extant systematic reviews

This growing interest in AIEd has led scholars to investigate the research on the use of artificial intelligence in education. Some scholars have conducted systematic reviews to focus on a specific subject domain. For example, Liang et. al. ( 2021 ) conducted a systematic review and bibliographic analysis the roles and research foci of AI in language education. Shukla et. al. ( 2019 ) focused their longitudinal bibliometric analysis on 30 years of using AI in Engineering. Hwang and Tu ( 2021 ) conducted a bibliometric mapping analysis on the roles and trends in the use of AI in mathematics education, and Winkler-Schwartz et. al. ( 2019 ) specifically examined the use of AI in medical education in looking for best practices in the use of machine learning to assess surgical expertise. These studies provide a specific focus on the use of AIEd in HE but do not provide an understanding of AI across HE.

On a broader view of AIEd in HE, Ouyang et. al. ( 2022 ) conducted a systematic review of AIEd in online higher education and investigated the literature regarding the use of AI from 2011 to 2020. The findings show that performance prediction, resource recommendation, automatic assessment, and improvement of learning experiences are the four main functions of AI applications in online higher education. Salas-Pilco and Yang ( 2022 ) focused on AI applications in Latin American higher education. The results revealed that the main AI applications in higher education in Latin America are: (1) predictive modeling, (2) intelligent analytics, (3) assistive technology, (4) automatic content analysis, and (5) image analytics. These studies provide valuable information for the online and Latin American context but not an overarching examination of AIEd in HE.

Studies have been conducted to examine HE. Hinojo-Lucena et. al. ( 2019 ) conducted a bibliometric study on the impact of AIEd in HE. They analyzed the scientific production of AIEd HE publications indexed in Web of Science and Scopus databases from 2007 to 2017. This study revealed that most of the published document types were proceedings papers. The United States had the highest number of publications, and the most cited articles were about implementing virtual tutoring to improve learning. Chu et. al. ( 2022 ) reviewed the top 50 most cited articles on AI in HE from 1996 to 2020, revealing that predictions of students’ learning status were most frequently discussed. AI technology was most frequently applied in engineering courses, and AI technologies most often had a role in profiling and prediction. Finally, Zawacki-Richter et. al. ( 2019 ) analyzed AIEd in HE from 2007 to 2018 to reveal four primary uses of AIEd: (1) profiling and prediction, (2) assessment and evaluation, (3) adaptive systems and personalization, and (4) intelligent tutoring systems. There do not appear to be any studies examining the last 2 years of AIEd in HE, and these authors describe the rapid speed of both AI development and the use of AIEd in HE and call for further research in this area.

Purpose of the study

The purpose of this study is in response to the appeal from scholars (viz., Chu et al., 2022 ; Hinojo-Lucena et al., 2019 ; Zawacki-Richter et al., 2019 ) to research to investigate the benefits and challenges of AIEd within HE settings. As the academic knowledge of AIEd HE finished with studies examining up to 2020, this study provides the most up-to-date analysis examining research through to the end of 2022.

The overarching question for this study is: what are the trends in HE research regarding the use of AIEd? The first two questions provide contextual information, such as where the studies occurred and the disciplines AI was used in. These contextual details are important for presenting the main findings of the third question of how AI is being used in HE.

In what geographical location was the AIEd research conducted, and how has the trend in the number of publications evolved across the years?

What departments were the first authors affiliated with, and what were the academic levels and subject domains in which AIEd research was being conducted?

Who are the intended users of the AI technologies and what are the applications of AI in higher education?

A PRISMA systematic review methodology was used to answer three questions guiding this study. PRISMA principles (Page et al., 2021 ) were used throughout the study. The PRISMA extension Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Protocols (PRISMA-P; Moher et al., 2015 ) were utilized in this study to provide an a priori roadmap to conduct a rigorous systematic review. Furthermore, the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA principles; Page et al., 2021 ) were used to search, identify, and select articles to be included in the research were used for searching, identifying, and selecting articles, then in how to read, extract, and manage the secondary data gathered from those studies (Moher et al., 2015 , PRISMA Statement, 2021 ). This systematic review approach supports an unbiased synthesis of the data in an impartial way (Hemingway & Brereton, 2009 ). Within the systematic review methodology, extracted data were aggregated and presented as whole numbers and percentages. A qualitative deductive and inductive coding methodology was also used to analyze extant data and generate new theories on the use of AI in HE (Gough et al., 2017 ).

The research begins with the search for the research articles to be included in the study. Based on the research question, the study parameters are defined including the search years, quality and types of publications to be included. Next, databases and journals are selected. A Boolean search is created and used for the search of those databases and journals. Once a set of publications are located from those searches, they are then examined against an inclusion and exclusion criteria to determine which studies will be included in the final study. The relevant data to match the research questions is then extracted from the final set of studies and coded. This method section is organized to describe each of these methods with full details to ensure transparency.

Search strategy

Only peer-reviewed journal articles were selected for examination in this systematic review. This ensured a level of confidence in the quality of the studies selected (Gough et al., 2017 ). The search parameters narrowed the search focus to include studies published in 2016 to 2022. This timeframe was selected to ensure the research was up to date, which is especially important with the rapid change in technology and AIEd.

The data retrieval protocol employed an electronic and a hand search. The electronic search included educational databases within EBSCOhost. Then an additional electronic search was conducted of Wiley Online Library, JSTOR, Science Direct, and Web of Science. Within each of these databases a full text search was conducted. Aligned to the research topic and questions, the Boolean search included terms related to AI, higher education, and learning. The Boolean search is listed in Table 1 . In the initial test search, the terms “machine learning” OR “intelligent support” OR “intelligent virtual reality” OR “chatbot” OR “automated tutor” OR “intelligent agent” OR “expert system” OR “neural network” OR “natural language processing” were used. These were removed as they were subcategories of terms found in Part 1 of the search. Furthermore, inclusion of these specific AI terms resulted in a large number of computer science courses that were focused on learning about AI and not the use of AI in learning.

Part 2 of the search ensured that articles involved formal university education. The terms higher education and tertiary were both used to recognize the different terms used in different countries. The final Boolean search was “Artificial intelligence” OR AI OR “smart technologies” OR “intelligent technologies” AND “higher education” OR tertiary OR graduate OR undergraduate. Scholars (viz., Ouyang et al., 2022 ) who conducted a systematic review on AIEd in HE up to 2020 noted that they missed relevant articles from their study, and other relevant journals should intentionally be examined. Therefore, a hand search was also conducted to include an examination of other journals relevant to AIEd that may not be included in the databases. This is important as the field of AIEd is still relatively new, and journals focused on this field may not yet be indexed in databases. The hand search included: The International Journal of Learning Analytics and Artificial Intelligence in Education, the International Journal of Artificial Intelligence in Education, and Computers & Education: Artificial Intelligence.

Electronic and hand searches resulted in 371 articles for possible inclusion. The search parameters within the electronic database search narrowed the search to articles published from 2016 to 2022, per-reviewed journal articles, and duplicates. Further screening was conducted manually, as each of the 138 articles were reviewed in full by two researchers to examine a match against the inclusion and exclusion criteria found in Table 2 .

The inter-rater reliability was calculated by percentage agreement (Belur et al., 2018 ). The researchers reached a 95% agreement for the coding. Further discussion of misaligned articles resulted in a 100% agreement. This screening process against inclusion and exclusion criteria resulted in the exclusion of 237 articles. This included the duplicates and those removed as part of the inclusion and exclusion criteria, see Fig.  1 . Leaving 138 articles for inclusion in this systematic review.

(From: Page et al., 2021 )

PRISMA flow chart of article identification and screening

The 138 articles were then coded to answer each of the research questions using deductive and inductive coding methods. Deductive coding involves examining data using a priori codes. A priori are pre-determined criteria and this process was used to code the countries, years, author affiliations, academic levels, and domains in the respective groups. Author affiliations were coded using the academic department of the first author of the study. First authors were chosen as that person is the primary researcher of the study and this follows past research practice (e.g., Zawacki-Richter et al., 2019 ). Who the AI was intended for was also coded using the a priori codes of Student, Instructor, Manager or Others. The Manager code was used for those who are involved in organizational tasks, e.g., tracking enrollment. Others was used for those not fitting the other three categories.

Inductive coding was used for the overarching question of this study in examining how the AI was being used in HE. Researchers of extant systematic reviews on AIEd in HE (viz., Chu et al., 2022 ; Zawacki-Richter et al., 2019 ) often used an a priori framework as researchers matched the use of AI to pre-existing frameworks. A grounded coding methodology (Strauss & Corbin, 1995 ) was selected for this study to allow findings of the trends on AIEd in HE to emerge from the data. This is important as it allows a direct understanding of how AI is being used rather than how researchers may think it is being used and fitting the data to pre-existing ideas.

Grounded coding process involved extracting how the AI was being used in HE from the articles. “In vivo” (Saldana, 2015 ) coding was also used alongside grounded coding. In vivo codes are when codes use language directly from the article to capture the primary authors’ language and ensure consistency with their findings. The grounded coding design used a constant comparative method. Researchers identified important text from articles related to the use of AI, and through an iterative process, initial codes led to axial codes with a constant comparison of uses of AI with uses of AI, then of uses of AI with codes, and codes with codes. Codes were deemed theoretically saturated when the majority of the data fit with one of the codes. For both the a priori and the grounded coding, two researchers coded and reached an inter-rater percentage agreement of 96%. After discussing misaligned articles, a 100% agreement was achieved.

Findings and discussion

The findings and discussion section are organized by the three questions guiding this study. The first two questions provide contextual information on the AIEd research, and the final question provides a rigorous investigation into how AI is being used in HE.

RQ1. In what geographical location was the AIEd research conducted, and how has the trend in the number of publications evolved across the years?

The 138 studies took place across 31 countries in six of seven continents of the world. Nonetheless, that distribution was not equal across continents. Asia had the largest number of AIEd studies in HE at 41%. Of the seven countries represented in Asia, 42 of the 58 studies were conducted in Taiwan and China. Europe, at 30%, was the second largest continent and had 15 countries ranging from one to eight studies a piece. North America, at 21% of the studies was the continent with the third largest number of studies, with the USA producing 21 of the 29 studies in that continent. The 21 studies from the USA places it second behind China. Only 1% of studies were conducted in South America and 2% in Africa. See Fig.  2 for a visual representation of study distribution across countries. Those continents with high numbers of studies are from high income countries and those with low numbers have a paucity of publications in low-income countries.

Geographical distribution of the AIEd HE studies

Data from Zawacki-Richter et. al.’s ( 2019 ) 2007–2018 systematic review examining countries found that the USA conducted the most studies across the globe at 43 out of 146, and China had the second largest at eleven of the 146 papers. Researchers have noted a rapid trend in Chinese researchers publishing more papers on AI and securing more patents than their US counterparts in a field that was originally led by the US (viz., Li et al., 2021 ). The data from this study corroborate this trend in China leading in the number of AIEd publications.

With the accelerated use of AI in society, gathering data to examine the use of AIEd in HE is useful in providing the scholarly community with specific information on that growth and if it is as prolific as anticipated by scholars (e.g., Chu et al., 2022 ). The analysis of data of the 138 studies shows that the trend towards the use of AIEd in HE has greatly increased. There is a drop in 2019, but then a great rise in 2021 and 2022; see Fig.  3 .

Chronological trend in AIEd in HE

Data on the rise in AIEd in HE is similar to the findings of Chu et. al. ( 2022 ) who noted an increase from 1996 to 2010 and 2011–2020. Nonetheless Chu’s parameters are across decades, and the rise is to be anticipated with a relatively new technology across a longitudinal review. Data from this study show a dramatic rise since 2020 with a 150% increase from the prior 2 years 2020–2019. The rise in 2021 and 2022 in HE could have been caused by the vast increase in HE faculty having to teach with technology during the pandemic lockdown. Faculty worldwide were using technologies, including AI, to explore how they could continue teaching and learning that was often face-to-face prior to lockdown. The disadvantage of this rapid adoption of technology is that there was little time to explore the possibilities of AI to transform learning, and AI may have been used to replicate past teaching practices, without considering new strategies previously inconceivable with the affordances of AI.

However, in a further examination of the research from 2021 to 2022, it appears that there are new strategies being considered. For example, Liu et. al.’s, 2022 study used AIEd to provide information on students’ interactions in an online environment and examine their cognitive effort. In Yao’s study in 2022, he examined the use of AI to determine student emotions while learning.

RQ2. What departments were the first authors affiliated with, and what were the academic levels and subject domains in which AIEd research was being conducted?

Department affiliations

Data from the AIEd HE studies show that of the first authors were most frequently from colleges of education (28%), followed by computer science (20%). Figure  4 presents the 15 academic affiliations of the authors found in the studies. The wide variety of affiliations demonstrate the variety of ways AI can be used in various educational disciplines, and how faculty in diverse areas, including tourism, music, and public affairs were interested in how AI can be used for educational purposes.

Research affiliations

In an extant AIED HE systematic review, Zawacki-Richter et. al.’s ( 2019 ) named their study Systematic review of research on artificial intelligence applications in higher education—where are the educators? In this study, the authors were keen to highlight that of the AIEd studies in HE, only six percent were written by researchers directly connected to the field of education, (i.e., from a college of education). The researchers found a great lack in pedagogical and ethical implications of implementing AI in HE and that there was a need for more educational perspectives on AI developments from educators conducting this work. It appears from our data that educators are now showing greater interest in leading these research endeavors, with the highest affiliated group belonging to education. This may again be due to the pandemic and those in the field of education needing to support faculty in other disciplines, and/or that they themselves needed to explore technologies for their own teaching during the lockdown. This may also be due to uptake in professors in education becoming familiar with AI tools also driven by a societal increased attention. As the focus of much research by education faculty is on teaching and learning, they are in an important position to be able to share their research with faculty in other disciplines regarding the potential affordances of AIEd.

Academic levels

The a priori coding of academic levels show that the majority of studies involved undergraduate students with 99 of the 138 (72%) focused on these students. This was in comparison to the 12 of 138 (9%) for graduate students. Some of the studies used AI for both academic levels: see Fig.  5

Academic level distribution by number of articles

This high percentage of studies focused on the undergraduate population was congruent with an earlier AIED HE systematic review (viz., Zawacki-Richter et al., 2019 ) who also reported student academic levels. This focus on undergraduate students may be due to the variety of affordances offered by AIEd, such as predictive analytics on dropouts and academic performance. These uses of AI may be less required for graduate students who already have a record of performance from their undergraduate years. Another reason for this demographic focus can also be convenience sampling, as researchers in HE typically has a much larger and accessible undergraduate population than graduates. This disparity between undergraduates and graduate populations is a concern, as AIEd has the potential to be valuable in both settings.

Subject domains

The studies were coded into 14 areas in HE; with 13 in a subject domain and one category of AIEd used in HE management of students; See Fig.  6 . There is not a wide difference in the percentages of top subject domains, with language learning at 17%, computer science at 16%, and engineering at 12%. The management of students category appeared third on the list at 14%. Prior studies have also found AIEd often used for language learning (viz., Crompton et al., 2021 ; Zawacki-Richter et al., 2019 ). These results are different, however, from Chu et. al.’s ( 2022 ) findings that show engineering dramatically leading with 20 of the 50 studies, with other subjects, such as language learning, appearing once or twice. This study appears to be an outlier that while the searches were conducted in similar databases, the studies only included 50 studies from 1996 to 2020.

Subject domains of AIEd in HE

Previous scholars primarily focusing on language learning using AI for writing, reading, and vocabulary acquisition used the affordances of natural language processing and intelligent tutoring systems (e.g., Liang et al., 2021 ). This is similar to the findings in studies with AI used for automated feedback of writing in a foreign language (Ayse et al., 2022 ), and AI translation support (Al-Tuwayrish, 2016 ). The large use of AI for managerial activities in this systematic review focused on making predictions (12 studies) and then admissions (three studies). This is positive to see this use of AI to look across multiple databases to see trends emerging from data that may not have been anticipated and cross referenced before (Crompton et al., 2022 ). For example, to examine dropouts, researchers may consider examining class attendance, and may not examine other factors that appear unrelated. AI analysis can examine all factors and may find that dropping out is due to factors beyond class attendance.

RQ3. Who are the intended users of the AI technologies and what are the applications of AI in higher education?

Intended user of AI

Of the 138 articles, the a priori coding shows that 72% of the studies focused on Students, followed by a focus on Instructors at 17%, and Managers at 11%, see Fig.  7 . The studies provided examples of AI being used to provide support to students, such as access to learning materials for inclusive learning (Gupta & Chen, 2022 ), provide immediate answers to student questions, self-testing opportunities (Yao, 2022 ), and instant personalized feedback (Mousavi et al., 2020 ).

Intended user

The data revealed a large emphasis on students in the use of AIEd in HE. This user focus is different from a recent systematic review on AIEd in K-12 that found that AIEd studies in K-12 settings prioritized teachers (Crompton et al., 2022 ). This may appear that HE uses AI to focus more on students than in K-12. However, this large number of student studies in HE may be due to the student population being more easily accessibility to HE researchers who may study their own students. The ethical review process is also typically much shorter in HE than in K-12. Therefore, the data on the intended focus should be reviewed while keeping in mind these other explanations. It was interesting that Managers were the lowest focus in K-12 and also in this study in HE. AI has great potential to collect, cross reference and examine data across large datasets that can allow data to be used for actionable insight. More focus on the use of AI by managers would tap into this potential.

How is AI used in HE

Using grounded coding, the use of AIEd from each of the 138 articles was examined and six major codes emerged from the data. These codes provide insight into how AI was used in HE. The five codes are: (1) Assessment/Evaluation, (2) Predicting, (3) AI Assistant, (4) Intelligent Tutoring System (ITS), and (5) Managing Student Learning. For each of these codes there are also axial codes, which are secondary codes as subcategories from the main category. Each code is delineated below with a figure of the codes with further descriptive information and examples.

Assessment/evaluation

Assessment and Evaluation was the most common use of AIEd in HE. Within this code there were six axial codes broken down into further codes; see Fig.  8 . Automatic assessment was most common, seen in 26 of the studies. It was interesting to see that this involved assessment of academic achievement, but also other factors, such as affect.

Codes and axial codes for assessment and evaluation

Automatic assessment was used to support a variety of learners in HE. As well as reducing the time it takes for instructors to grade (Rutner & Scott, 2022 ), automatic grading showed positive use for a variety of students with diverse needs. For example, Zhang and Xu ( 2022 ) used automatic assessment to improve academic writing skills of Uyghur ethnic minority students living in China. Writing has a variety of cultural nuances and in this study the students were shown to engage with the automatic assessment system behaviorally, cognitively, and affectively. This allowed the students to engage in self-regulated learning while improving their writing.

Feedback was a description often used in the studies, as students were given text and/or images as feedback as a formative evaluation. Mousavi et. al. ( 2020 ) developed a system to provide first year biology students with an automated personalized feedback system tailored to the students’ specific demographics, attributes, and academic status. With the unique feature of AIEd being able to analyze multiple data sets involving a variety of different students, AI was used to assess and provide feedback on students’ group work (viz., Ouatik et al., 2021 ).

AI also supports instructors in generating questions and creating multiple question tests (Yang et al., 2021 ). For example, (Lu et al., 2021 ) used natural language processing to create a system that automatically created tests. Following a Turing type test, researchers found that AI technologies can generate highly realistic short-answer questions. The ability for AI to develop multiple questions is a highly valuable affordance as tests can take a great deal of time to make. However, it would be important for instructors to always confirm questions provided by the AI to ensure they are correct and that they match the learning objectives for the class, especially in high value summative assessments.

The axial code within assessment and evaluation revealed that AI was used to review activities in the online space. This included evaluating student’s reflections, achievement goals, community identity, and higher order thinking (viz., Huang et al., 2021 ). Three studies used AIEd to evaluate educational materials. This included general resources and textbooks (viz., Koć‑Januchta et al., 2022 ). It is interesting to see the use of AI for the assessment of educational products, rather than educational artifacts developed by students. While this process may be very similar in nature, this shows researchers thinking beyond the traditional use of AI for assessment to provide other affordances.

Predicting was a common use of AIEd in HE with 21 studies focused specifically on the use of AI for forecasting trends in data. Ten axial codes emerged on the way AI was used to predict different topics, with nine focused on predictions regarding students and the other on predicting the future of higher education. See Fig.  9 .

Predicting axial codes

Extant systematic reviews on HE highlighted the use of AIEd for prediction (viz., Chu et al., 2022 ; Hinojo-Lucena et al., 2019 ; Ouyang et al., 2022 ; Zawacki-Richter et al., 2019 ). Ten of the articles in this study used AI for predicting academic performance. Many of the axial codes were often overlapping, such as predicting at risk students, and predicting dropouts; however, each provided distinct affordances. An example of this is the study by Qian et. al. ( 2021 ). These researchers examined students taking a MOOC course. MOOCs can be challenging environments to determine information on individual students with the vast number of students taking the course (Krause & Lowe, 2014 ). However, Qian et al., used AIEd to predict students’ future grades by inputting 17 different learning features, including past grades, into an artificial neural network. The findings were able to predict students’ grades and highlight students at risk of dropping out of the course.

In a systematic review on AIEd within the K-12 context (viz., Crompton et al., 2022 ), prediction was less pronounced in the findings. In the K-12 setting, there was a brief mention of the use of AI in predicting student academic performance. One of the studies mentioned students at risk of dropping out, but this was immediately followed by questions about privacy concerns and describing this as “sensitive”. The use of prediction from the data in this HE systematic review cover a wide range of AI predictive affordances. students Sensitivity is still important in a HE setting, but it is positive to see the valuable insight it provides that can be used to avoid students failing in their goals.

AI assistant

The studies evaluated in this review indicated that the AI Assistant used to support learners had a variety of different names. This code included nomenclature such as, virtual assistant, virtual agent, intelligent agent, intelligent tutor, and intelligent helper. Crompton et. al. ( 2022 ), described the difference in the terms to delineate the way that the AI appeared to the user. For example, if there was an anthropomorphic presence to the AI, such as an avatar, or if the AI appeared to support via other means, such as text prompt. The findings of this systematic review align to Crompton et. al.’s ( 2022 ) descriptive differences of the AI Assistant. Furthermore, this code included studies that provide assistance to students, but may not have specifically used the word assistance. These include the use of chatbots for student outreach, answering questions, and providing other assistance. See Fig.  10 for the axial codes for AI Assistant.

AI assistant axial codes

Many of these assistants offered multiple supports to students, such as Alex , the AI described as a virtual change agent in Kim and Bennekin’s ( 2016 ) study. Alex interacted with students in a college mathematics course by asking diagnostic questions and gave support depending on student needs. Alex’s support was organized into four stages: (1) goal initiation (“Want it”), (2) goal formation (“Plan for it”), (3) action control (“Do it”), and (4) emotion control (“Finish it”). Alex provided responses depending on which of these four areas students needed help. These messages supported students with the aim of encouraging persistence in pursuing their studies and degree programs and improving performance.

The role of AI in providing assistance connects back to the seminal work of Vygotsky ( 1978 ) and the Zone of Proximal Development (ZPD). ZPD highlights the degree to which students can rapidly develop when assisted. Vygotsky described this assistance often in the form of a person. However, with technological advancements, the use of AI assistants in these studies are providing that support for students. The affordances of AI can also ensure that the support is timely without waiting for a person to be available. Also, assistance can consider aspects on students’ academic ability, preferences, and best strategies for supporting. These features were evident in Kim and Bennekin’s ( 2016 ) study using Alex.

Intelligent tutoring system

The use of Intelligent Tutoring Systems (ITS) was revealed in the grounded coding. ITS systems are adaptive instructional systems that involve the use of AI techniques and educational methods. An ITS system customizes educational activities and strategies based on student’s characteristics and needs (Mousavinasab et al., 2021 ). While ITS may be an anticipated finding in AIED HE systematic reviews, it was interesting that extant reviews similar to this study did not always describe their use in HE. For example, Ouyang et. al. ( 2022 ), included “intelligent tutoring system” in search terms describing it as a common technique, yet ITS was not mentioned again in the paper. Zawacki-Richter et. al. ( 2019 ) on the other hand noted that ITS was in the four overarching findings of the use of AIEd in HE. Chu et. al. ( 2022 ) then used Zawacki-Richter’s four uses of AIEd for their recent systematic review.

In this systematic review, 18 studies specifically mentioned that they were using an ITS. The ITS code did not necessitate axial codes as they were performing the same type of function in HE, namely, in providing adaptive instruction to the students. For example, de Chiusole et. al. ( 2020 ) developed Stat-Knowlab, an ITS that provides the level of competence and best learning path for each student. Thus Stat-Knowlab personalizes students’ learning and provides only educational activities that the student is ready to learn. This ITS is able to monitor the evolution of the learning process as the student interacts with the system. In another study, Khalfallah and Slama ( 2018 ) built an ITS called LabTutor for engineering students. LabTutor served as an experienced instructor in enabling students to access and perform experiments on laboratory equipment while adapting to the profile of each student.

The student population in university classes can go into the hundreds and with the advent of MOOCS, class sizes can even go into the thousands. Even in small classes of 20 students, the instructor cannot physically provide immediate unique personalize questions to each student. Instructors need time to read and check answers and then take further time to provide feedback before determining what the next question should be. Working with the instructor, AIEd can provide that immediate instruction, guidance, feedback, and following questioning without delay or becoming tired. This appears to be an effective use of AIEd, especially within the HE context.

Managing student learning

Another code that emerged in the grounded coding was focused on the use of AI for managing student learning. AI is accessed to manage student learning by the administrator or instructor to provide information, organization, and data analysis. The axial codes reveal the trends in the use of AI in managing student learning; see Fig.  11 .

Learning analytics was an a priori term often found in studies which describes “the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs” (Long & Siemens, 2011 , p. 34). The studies investigated in this systematic review were across grades and subject areas and provided administrators and instructors different types of information to guide their work. One of those studies was conducted by Mavrikis et. al. ( 2019 ) who described learning analytics as teacher assistance tools. In their study, learning analytics were used in an exploratory learning environment with targeted visualizations supporting classroom orchestration. These visualizations, displayed as screenshots in the study, provided information such as the interactions between the students, goals achievements etc. These appear similar to infographics that are brightly colored and draw the eye quickly to pertinent information. AI is also used for other tasks, such as organizing the sequence of curriculum in pacing guides for future groups of students and also designing instruction. Zhang ( 2022 ) described how designing an AI teaching system of talent cultivation and using the digital affordances to establish a quality assurance system for practical teaching, provides new mechanisms for the design of university education systems. In developing such a system, Zhang found that the stability of the instructional design, overcame the drawbacks of traditional manual subjectivity in the instructional design.

Another trend that emerged from the studies was the use of AI to manage student big data to support learning. Ullah and Hafiz ( 2022 ) lament that using traditional methods, including non-AI digital techniques, asking the instructor to pay attention to every student’s learning progress is very difficult and that big data analysis techniques are needed. The ability to look across and within large data sets to inform instruction is a valuable affordance of AIEd in HE. While the use of AIEd to manage student learning emerged from the data, this study uncovered only 19 studies in 7 years (2016–2022) that focused on the use of AIEd to manage student data. This lack of the use was also noted in a recent study in the K-12 space (Crompton et al., 2022 ). In Chu et. al.’s ( 2022 ) study examining the top 50 most cited AIEd articles, they did not report the use of AIEd for managing student data in the top uses of AIEd HE. It would appear that more research should be conducted in this area to fully explore the possibilities of AI.

Gaps and future research

From this systematic review, six gaps emerged in the data providing opportunities for future studies to investigate and provide a fuller understanding of how AIEd can used in HE. (1) The majority of the research was conducted in high income countries revealing a paucity of research in developing countries. More research should be conducted in these developing countries to expand the level of understanding about how AI can enhance learning in under-resourced communities. (2) Almost 50% of the studies were conducted in the areas of language learning, computer science and engineering. Research conducted by members from multiple, different academic departments would help to advance the knowledge of the use of AI in more disciplines. (3) This study revealed that faculty affiliated with schools of education are taking an increasing role in researching the use of AIEd in HE. As this body of knowledge grows, faculty in Schools of Education should share their research regarding the pedagogical affordances of AI so that this knowledge can be applied by faculty across disciplines. (4) The vast majority of the research was conducted at the undergraduate level. More research needs to be done at the graduate student level, as AI provides many opportunities in this environment. (5) Little study was done regarding how AIEd can assist both instructors and managers in their roles in HE. The power of AI to assist both groups further research. (6) Finally, much of the research investigated in this systematic review revealed the use of AIEd in traditional ways that enhance or make more efficient current practices. More research needs to focus on the unexplored affordances of AIEd. As AI becomes more advanced and sophisticated, new opportunities will arise for AIEd. Researchers need to be on the forefront of these possible innovations.

In addition, empirical exploration is needed for new tools, such as ChatGPT that was available for public use at the end of 2022. With the time it takes for a peer review journal article to be published, ChatGPT did not appear in the articles for this study. What is interesting is that it could fit with a variety of the use codes found in this study, with students getting support in writing papers and instructors using Chat GPT to assess students work and with help writing emails or descriptions for students. It would be pertinent for researchers to explore Chat GPT.

Limitations

The findings of this study show a rapid increase in the number of AIEd studies published in HE. However, to ensure a level of credibility, this study only included peer review journal articles. These articles take months to publish. Therefore, conference proceedings and gray literature such as blogs and summaries may reveal further findings not explored in this study. In addition, the articles in this study were all published in English which excluded findings from research published in other languages.

In response to the call by Hinojo-Lucena et. al. ( 2019 ), Chu et. al. ( 2022 ), and Zawacki-Richter et. al. ( 2019 ), this study provides unique findings with an up-to-date examination of the use of AIEd in HE from 2016 to 2022. Past systematic reviews examined the research up to 2020. The findings of this study show that in 2021 and 2022, publications rose nearly two to three times the number of previous years. With this rapid rise in the number of AIEd HE publications, new trends have emerged.

The findings show that of the 138 studies examined, research was conducted in six of the seven continents of the world. In extant systematic reviews showed that the US led by a large margin in the number of studies published. This trend has now shifted to China. Another shift in AIEd HE is that while extant studies lamented the lack of focus on professors of education leading these studies, this systematic review found education to be the most common department affiliation with 28% and computer science coming in second at 20%. Undergraduate students were the most studied students at 72%. Similar to the findings of other studies, language learning was the most common subject domain. This included writing, reading, and vocabulary acquisition. In examination of who the AIEd was intended for, 72% of the studies focused on students, 17% instructors, and 11% managers.

Grounded coding was used to answer the overarching question of how AIEd was used in HE. Five usage codes emerged from the data: (1) Assessment/Evaluation, (2) Predicting, (3) AI Assistant, (4) Intelligent Tutoring System (ITS), and (5) Managing Student Learning. Assessment and evaluation had a wide variety of purposes, including assessing academic progress and student emotions towards learning, individual and group evaluations, and class based online community assessments. Predicting emerged as a code with ten axial codes, as AIEd predicted dropouts and at-risk students, innovative ability, and career decisions. AI Assistants were specific to supporting students in HE. These assistants included those with an anthropomorphic presence, such as virtual agents and persuasive intervention through digital programs. ITS systems were not always noted in extant systematic reviews but were specifically mentioned in 18 of the studies in this review. ITS systems in this study provided customized strategies and approaches to student’s characteristics and needs. The final code in this study highlighted the use of AI in managing student learning, including learning analytics, curriculum sequencing, instructional design, and clustering of students.

The findings of this study provide a springboard for future academics, practitioners, computer scientists, policymakers, and funders in understanding the state of the field in AIEd HE, how AI is used. It also provides actionable items to ameliorate gaps in the current understanding. As the use AIEd will only continue to grow this study can serve as a baseline for further research studies in the use of AIEd in HE.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Alajmi, Q., Al-Sharafi, M. A., & Abuali, A. (2020). Smart learning gateways for Omani HEIs towards educational technology: Benefits, challenges and solutions. International Journal of Information Technology and Language Studies, 4 (1), 12–17.

Google Scholar  

Al-Tuwayrish, R. K. (2016). An evaluative study of machine translation in the EFL scenario of Saudi Arabia. Advances in Language and Literary Studies, 7 (1), 5–10.

Ayse, T., & Nil, G. (2022). Automated feedback and teacher feedback: Writing achievement in learning English as a foreign language at a distance. The Turkish Online Journal of Distance Education, 23 (2), 120–139. https://doi.org/10.7575/aiac.alls.v.7n.1p.5

Article   Google Scholar  

Baykasoğlu, A., Özbel, B. K., Dudaklı, N., Subulan, K., & Şenol, M. E. (2018). Process mining based approach to performance evaluation in computer-aided examinations. Computer Applications in Engineering Education, 26 (5), 1841–1861. https://doi.org/10.1002/cae.21971

Belur, J., Tompson, L., Thornton, A., & Simon, M. (2018). Interrater reliability in systematic review methodology: Exploring variation in coder decision-making. Sociological Methods & Research, 13 (3), 004912411887999. https://doi.org/10.1177/0049124118799372

Çağataylı, M., & Çelebi, E. (2022). Estimating academic success in higher education using big five personality traits, a machine learning approach. Arab Journal Scientific Engineering, 47 , 1289–1298. https://doi.org/10.1007/s13369-021-05873-4

Chen, L., Chen, P., & Lin, Z. (2020). Artificial intelligence in education: A review. IEEE Access, 8 , 75264–75278. https://doi.org/10.1109/ACCESS.2020.2988510

Chu, H., Tu, Y., & Yang, K. (2022). Roles and research trends of artificial intelligence in higher education: A systematic review of the top 50 most-cited articles. Australasian Journal of Educational Technology, 38 (3), 22–42. https://doi.org/10.14742/ajet.7526

Cristianini, N. (2016). Intelligence reinvented. New Scientist, 232 (3097), 37–41. https://doi.org/10.1016/S0262-4079(16)31992-3

Crompton, H., Bernacki, M. L., & Greene, J. (2020). Psychological foundations of emerging technologies for teaching and learning in higher education. Current Opinion in Psychology, 36 , 101–105. https://doi.org/10.1016/j.copsyc.2020.04.011

Crompton, H., & Burke, D. (2022). Artificial intelligence in K-12 education. SN Social Sciences, 2 , 113. https://doi.org/10.1007/s43545-022-00425-5

Crompton, H., Jones, M., & Burke, D. (2022). Affordances and challenges of artificial intelligence in K-12 education: A systematic review. Journal of Research on Technology in Education . https://doi.org/10.1080/15391523.2022.2121344

Crompton, H., & Song, D. (2021). The potential of artificial intelligence in higher education. Revista Virtual Universidad Católica Del Norte, 62 , 1–4. https://doi.org/10.35575/rvuen.n62a1

de Chiusole, D., Stefanutti, L., Anselmi, P., & Robusto, E. (2020). Stat-Knowlab. Assessment and learning of statistics with competence-based knowledge space theory. International Journal of Artificial Intelligence in Education, 30 , 668–700. https://doi.org/10.1007/s40593-020-00223-1

Dever, D. A., Azevedo, R., Cloude, E. B., & Wiedbusch, M. (2020). The impact of autonomy and types of informational text presentations in game-based environments on learning: Converging multi-channel processes data and learning outcomes. International Journal of Artificial Intelligence in Education, 30 (4), 581–615. https://doi.org/10.1007/s40593-020-00215-1

Górriz, J. M., Ramírez, J., Ortíz, A., Martínez-Murcia, F. J., Segovia, F., Suckling, J., Leming, M., Zhang, Y. D., Álvarez-Sánchez, J. R., Bologna, G., Bonomini, P., Casado, F. E., Charte, D., Charte, F., Contreras, R., Cuesta-Infante, A., Duro, R. J., Fernández-Caballero, A., Fernández-Jover, E., … Ferrández, J. M. (2020). Artificial intelligence within the interplay between natural and artificial computation: Advances in data science, trends and applications. Neurocomputing, 410 , 237–270. https://doi.org/10.1016/j.neucom.2020.05.078

Gough, D., Oliver, S., & Thomas, J. (2017). An introduction to systematic reviews (2nd ed.). Sage.

Gupta, S., & Chen, Y. (2022). Supporting inclusive learning using chatbots? A chatbot-led interview study. Journal of Information Systems Education, 33 (1), 98–108.

Hemingway, P. & Brereton, N. (2009). In Hayward Medical Group (Ed.). What is a systematic review? Retrieved from http://www.medicine.ox.ac.uk/bandolier/painres/download/whatis/syst-review.pdf

Hinojo-Lucena, F., Arnaz-Diaz, I., Caceres-Reche, M., & Romero-Rodriguez, J. (2019). A bibliometric study on its impact the scientific literature. Education Science . https://doi.org/10.3390/educsci9010051

Hrastinski, S., Olofsson, A. D., Arkenback, C., Ekström, S., Ericsson, E., Fransson, G., Jaldemark, J., Ryberg, T., Öberg, L.-M., Fuentes, A., Gustafsson, U., Humble, N., Mozelius, P., Sundgren, M., & Utterberg, M. (2019). Critical imaginaries and reflections on artificial intelligence and robots in postdigital K-12 education. Postdigital Science and Education, 1 (2), 427–445. https://doi.org/10.1007/s42438-019-00046-x

Huang, C., Wu, X., Wang, X., He, T., Jiang, F., & Yu, J. (2021). Exploring the relationships between achievement goals, community identification and online collaborative reflection. Educational Technology & Society, 24 (3), 210–223.

Hwang, G. J., & Tu, Y. F. (2021). Roles and research trends of artificial intelligence in mathematics education: A bibliometric mapping analysis and systematic review. Mathematics, 9 (6), 584. https://doi.org/10.3390/math9060584

Khalfallah, J., & Slama, J. B. H. (2018). The effect of emotional analysis on the improvement of experimental e-learning systems. Computer Applications in Engineering Education, 27 (2), 303–318. https://doi.org/10.1002/cae.22075

Kim, C., & Bennekin, K. N. (2016). The effectiveness of volition support (VoS) in promoting students’ effort regulation and performance in an online mathematics course. Instructional Science, 44 , 359–377. https://doi.org/10.1007/s11251-015-9366-5

Koć-Januchta, M. M., Schönborn, K. J., Roehrig, C., Chaudhri, V. K., Tibell, L. A. E., & Heller, C. (2022). “Connecting concepts helps put main ideas together”: Cognitive load and usability in learning biology with an AI-enriched textbook. International Journal of Educational Technology in Higher Education, 19 (11), 11. https://doi.org/10.1186/s41239-021-00317-3

Krause, S. D., & Lowe, C. (2014). Invasion of the MOOCs: The promise and perils of massive open online courses . Parlor Press.

Li, D., Tong, T. W., & Xiao, Y. (2021). Is China emerging as the global leader in AI? Harvard Business Review. https://hbr.org/2021/02/is-china-emerging-as-the-global-leader-in-ai

Liang, J. C., Hwang, G. J., Chen, M. R. A., & Darmawansah, D. (2021). Roles and research foci of artificial intelligence in language education: An integrated bibliographic analysis and systematic review approach. Interactive Learning Environments . https://doi.org/10.1080/10494820.2021.1958348

Liu, S., Hu, T., Chai, H., Su, Z., & Peng, X. (2022). Learners’ interaction patterns in asynchronous online discussions: An integration of the social and cognitive interactions. British Journal of Educational Technology, 53 (1), 23–40. https://doi.org/10.1111/bjet.13147

Long, P., & Siemens, G. (2011). Penetrating the fog: Analytics in learning and education. Educause Review, 46 (5), 31–40.

Lu, O. H. T., Huang, A. Y. Q., Tsai, D. C. L., & Yang, S. J. H. (2021). Expert-authored and machine-generated short-answer questions for assessing students learning performance. Educational Technology & Society, 24 (3), 159–173.

Mavrikis, M., Geraniou, E., Santos, S. G., & Poulovassilis, A. (2019). Intelligent analysis and data visualization for teacher assistance tools: The case of exploratory learning. British Journal of Educational Technology, 50 (6), 2920–2942. https://doi.org/10.1111/bjet.12876

Moher, D., Shamseer, L., Clarke, M., Ghersi, D., Liberati, A., Petticrew, M., Shekelle, P., & Stewart, L. (2015). Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Systematic Reviews, 4 (1), 1–9. https://doi.org/10.1186/2046-4053-4-1

Mousavi, A., Schmidt, M., Squires, V., & Wilson, K. (2020). Assessing the effectiveness of student advice recommender agent (SARA): The case of automated personalized feedback. International Journal of Artificial Intelligence in Education, 31 (2), 603–621. https://doi.org/10.1007/s40593-020-00210-6

Mousavinasab, E., Zarifsanaiey, N., Kalhori, S. R. N., Rakhshan, M., Keikha, L., & Saeedi, M. G. (2021). Intelligent tutoring systems: A systematic review of characteristics, applications, and evaluation methods. Interactive Learning Environments, 29 (1), 142–163. https://doi.org/10.1080/10494820.2018.1558257

Ouatik, F., Ouatikb, F., Fadlic, H., Elgoraria, A., Mohadabb, M. E. L., Raoufia, M., et al. (2021). E-Learning & decision making system for automate students assessment using remote laboratory and machine learning. Journal of E-Learning and Knowledge Society, 17 (1), 90–100. https://doi.org/10.20368/1971-8829/1135285

Ouyang, F., Zheng, L., & Jiao, P. (2022). Artificial intelligence in online higher education: A systematic review of empirical research from 2011–2020. Education and Information Technologies, 27 , 7893–7925. https://doi.org/10.1007/s10639-022-10925-9

Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T., Mulrow, C., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. British Medical Journal . https://doi.org/10.1136/bmj.n71

Popenici, S. A. D., & Kerr, S. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research and Practice in Technology Enhanced Learning, 12 (22), 1–13. https://doi.org/10.1186/s41039-017-0062-8

PRISMA Statement. (2021). PRISMA endorsers. PRISMA statement website. http://www.prisma-statement.org/Endorsement/PRISMAEndorsers

Qian, Y., Li, C.-X., Zou, X.-G., Feng, X.-B., Xiao, M.-H., & Ding, Y.-Q. (2022). Research on predicting learning achievement in a flipped classroom based on MOOCs by big data analysis. Computer Applied Applications in Engineering Education, 30 , 222–234. https://doi.org/10.1002/cae.22452

Rutner, S. M., & Scott, R. A. (2022). Use of artificial intelligence to grade student discussion boards: An exploratory study. Information Systems Education Journal, 20 (4), 4–18.

Salas-Pilco, S., & Yang, Y. (2022). Artificial Intelligence application in Latin America higher education: A systematic review. International Journal of Educational Technology in Higher Education, 19 (21), 1–20. https://doi.org/10.1186/S41239-022-00326-w

Saldana, J. (2015). The coding manual for qualitative researchers (3rd ed.). Sage.

Shukla, A. K., Janmaijaya, M., Abraham, A., & Muhuri, P. K. (2019). Engineering applications of artificial intelligence: A bibliometric analysis of 30 years (1988–2018). Engineering Applications of Artificial Intelligence, 85 , 517–532. https://doi.org/10.1016/j.engappai.2019.06.010

Strauss, A., & Corbin, J. (1995). Grounded theory methodology: An overview. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 273–285). Sage.

Turing, A. M. (1937). On computable numbers, with an application to the Entscheidungs problem. Proceedings of the London Mathematical Society, 2 (1), 230–265.

Turing, A. M. (1950). Computing machinery and intelligence. Mind, 59 , 443–460.

MathSciNet   Google Scholar  

Ullah, H., & Hafiz, M. A. (2022). Exploring effective classroom management strategies in secondary schools of Punjab. Journal of the Research Society of Pakistan, 59 (1), 76.

Verdú, E., Regueras, L. M., Gal, E., et al. (2017). Integration of an intelligent tutoring system in a course of computer network design. Educational Technology Research and Development, 65 , 653–677. https://doi.org/10.1007/s11423-016-9503-0

Vygotsky, L. S. (1978). Mind and society: The development of higher psychological processes . Harvard University Press.

Winkler-Schwartz, A., Bissonnette, V., Mirchi, N., Ponnudurai, N., Yilmaz, R., Ledwos, N., Siyar, S., Azarnoush, H., Karlik, B., & Del Maestro, R. F. (2019). Artificial intelligence in medical education: Best practices using machine learning to assess surgical expertise in virtual reality simulation. Journal of Surgical Education, 76 (6), 1681–1690. https://doi.org/10.1016/j.jsurg.2019.05.015

Yang, A. C. M., Chen, I. Y. L., Flanagan, B., & Ogata, H. (2021). Automatic generation of cloze items for repeated testing to improve reading comprehension. Educational Technology & Society, 24 (3), 147–158.

Yao, X. (2022). Design and research of artificial intelligence in multimedia intelligent question answering system and self-test system. Advances in Multimedia . https://doi.org/10.1155/2022/2156111

Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education—Where are the educators? International Journal of Educational Technology in Higher Education, 16 (1), 1–27. https://doi.org/10.1186/s41239-019-0171-0

Zhang, F. (2022). Design and application of artificial intelligence technology-driven education and teaching system in universities. Computational and Mathematical Methods in Medicine . https://doi.org/10.1155/2022/8503239

Zhang, Z., & Xu, L. (2022). Student engagement with automated feedback on academic writing: A study on Uyghur ethnic minority students in China. Journal of Multilingual and Multicultural Development . https://doi.org/10.1080/01434632.2022.2102175

Download references

Acknowledgements

The authors would like to thank Mildred Jones, Katherina Nako, Yaser Sendi, and Ricardo Randall for data gathering and organization.

Author information

Authors and affiliations.

Department of Teaching and Learning, Old Dominion University, Norfolk, USA

Helen Crompton

ODUGlobal, Norfolk, USA

Diane Burke

RIDIL, ODUGlobal, Norfolk, USA

You can also search for this author in PubMed   Google Scholar

Contributions

HC: Conceptualization; Data curation; Project administration; Formal analysis; Methodology; Project administration; original draft; and review & editing. DB: Conceptualization; Data curation; Project administration; Formal analysis; Methodology; Project administration; original draft; and review & editing. Both authors read and approved this manuscript.

Corresponding author

Correspondence to Helen Crompton .

Ethics declarations

Competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

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

Rights and permissions

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

Reprints and permissions

About this article

Cite this article.

Crompton, H., Burke, D. Artificial intelligence in higher education: the state of the field. Int J Educ Technol High Educ 20 , 22 (2023). https://doi.org/10.1186/s41239-023-00392-8

Download citation

Received : 30 January 2023

Accepted : 23 March 2023

Published : 24 April 2023

DOI : https://doi.org/10.1186/s41239-023-00392-8

Share this article

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

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

Provided by the Springer Nature SharedIt content-sharing initiative

• Artificial Intelligence
• Systematic review
• Higher education

• Open access
• Published: 31 May 2024

Public involvement and engagement in scientific research and higher education: the only way is ethics?

• Claire Nollett 1 ,
• Matthias Eberl 2 , 3 ,
• Jim Fitzgibbon 4 ,
• Natalie Joseph-Williams 5 , 6 &
• Sarah Hatch 7  

Research Involvement and Engagement volume  10 , Article number:  50 ( 2024 ) Cite this article

294 Accesses

28 Altmetric

Metrics details

Involving and engaging the public in scientific research and higher education is slowly becoming the norm for academic institutions in the United Kingdom and elsewhere. Driven by a wide range of stakeholders including regulators, funders, research policymakers and charities public involvement and public engagement are increasingly seen as essential in delivering open and transparent activity that is relevant and positively impacts on our society. It is obvious that any activities involving and engaging members of the public should be conducted safely and ethically. However, it is not clear whether conducting activities ethically means they require ethical approval from a research ethics committee.

Although there is some guidance available from government organisations (e.g. the UK Health Research Authority) to suggest ethical approval is not required for such activities, requests from funders and publishers to have ethical approval in place is commonplace in the authors’ experience. We explore this using case studies from our own institution.

We conclude that any public-facing activity with the purpose to systemically investigate knowledge, attitudes and experiences of members of the public as research and as human participants requires prior approval from an ethics committee. In contrast, engaging and involving members of the public and drawing on lived experience to inform aspects of research and teaching does not. However, lack of clarity around this distinction often results in the academic community seeking ethical approval ‘just in case’, leading to wasted time and resources and erecting unnecessary barriers for public involvement and public engagement. Instead, ethical issues and risks should be appropriately considered and mitigated by the relevant staff within their professional roles, be it academic or a professional service. Often this can involve following published guidelines and conducting an activity risk assessment, or similar. Moving forward, it is critical that academic funders and publishers acknowledge the distinction and agree on an accepted approach to avoid further exacerbating the problem.

Plain English summary

Involving and engaging members of the public is recognised best practice in university research and teaching. Involvement and engagement activities (for instance, working with the public to design a research study) continue to increase in priority and are an important part of an academic’s role. However, there is often confusion amongst researchers and educators around whether involving the public in these activities requires prior ethical approval, similar to what would be the case when inviting members of the public to participate in a clinical research study, or to donate samples such as blood for experiments. As an example, sometimes researchers are asked for ethical approval by scientific journals when trying to publish the findings from their public involvement and engagement work, when in fact this is not needed. The ongoing uncertainty about the difference between actual research on one hand and public involvement and engagement on the other hand wastes precious time and resources, and is a barrier for scientists to working with the public. We have developed guidance for academic staff on when ethical approval is and is not required, using examples from our own experience. We wrote this article to bring awareness to this problem; share our views with the wider academic community; encourage discussion around the problem and possible solutions; and ultimately contribute to educating on when research ethics approval is needed, and when not.

Peer Review reports

Public involvement (PI) is ‘important, expected and possible in all types of health and social care research’ [ 1 ]. It is now commonly embedded and reported in health research papers in the UK, with approximately half mentioning public involvement activities [ 2 ]. Public engagement (PE) is also encouraged and recognised by funders and other stakeholders across the higher education sector to raise awareness, increase trust and transparency, share knowledge, foster learning and deliver positive impact to society [ 3 ].

In 2019, the UK Standards Partnership published the UK Standards for Public Involvement ‘to help researchers and organisations improve the quality and consistency of public involvement in health and care research’ [ 4 ], and a large knowledge base is developing around how to do public involvement well. However, PI is not without its challenges, as identified both in the literature e.g [ 5 ]. and through our own experience as academic researchers, professional services staff and members of several national public involvement committees. Key issues include how to efficiently pay and reimburse public contributors within organisations, how to effectively evaluate the impact, and how to provide inclusive opportunities and reach under-served groups to increase the diversity of those involved [ 6 ].

The Research Excellence Framework (REF) 2029, the UK’s national assessment of the quality of research produced by its higher education institutions held every 6–7 years, will see a 25% weighting of returns with respect to the social, economic and political influence of the research conducted. The 2029 round will in fact be the first REF assessment where impact will be measured as “ Engagement and Impact” (our emphasis), alongside an accompanying statement to evidence engagement and impact activity beyond case studies [ 7 ]. As with PI, researchers face challenges in delivering PE including achieving the inclusion of under-served communities [ 8 ] and how to evaluate impact [ 3 ].

With individual researchers and their host institutions increasingly embracing PI and PE as part of their research and scholarship activities, there is one issue that we have found particularly contentious with researchers, employers, funders and publishers across both involvement and engagement and that is the focus of this commentary: the role of ethical approval in PI and PE activity.

Public involvement, sometimes referred to as Patient & Public Involvement (PPI) in health and social care research, is defined as ‘research being carried out ‘with’ or ‘by’ members of the public, rather than ‘to’, ‘about’ or ‘for’ them’ [ 9 ]. PE, adopting the UK’s National Coordinating Centre for Public Engagement’s definition, is a ‘myriad of ways in which the activity and benefits of higher education and research can be shared with the public’ [ 10 ]. PE is by definition a two-way process, involving interaction and listening, with the goal of generating mutual benefit. Both PI and PE are distinct from human participation in research whereby a member of the public agrees via informed consent to be a participant in research, e.g. receiving a study intervention, donating samples or sharing lived experiences. Whilst health and social care research involving human participants requires approval from a research ethics committee (REC), PI and PE activities typically do not.

In the UK, ethical approval is granted by a REC under the auspices of the National Health Service (NHS) for research on patients or healthcare professionals, or a local review committee or panel for research that does not include NHS patients. In academic research, this would usually be a university or school REC (referred to here as an Institutional Review Board, IRB). Other countries may use different approaches but the general need for RECs to approve research with human participants is ubiquitous. With regard to public involvement, the UK Health Research Authority (HRA) that is responsible for all NHS RECs explicitly states that ‘You do not need to submit an application to a Research Ethics Committee in order to involve the public in the planning or the design stage of research, even if the people involved are patients’ [ 11 ]. This advice would also apply to university ethics committees. However, despite this clear distinction, we have encountered and become aware of situations in which investigators were asked to acquire ethical approval for activities with the public – including PI, PE and impact activities. This highlights a potential misunderstanding of the nature of PI and PE, and their role alongside research. Whilst either activity can raise ethical considerations for the individuals involved, the requests to acquire research ethics approval for PI and PE need to be challenged within the academic community to increase awareness, understanding of and best practice around these activities. Seeking unnecessary approval adds a heavy additional burden on researchers which effectively acts as a barrier to carrying out PI and PE; can significantly delay timely activities; and uses valuable resources.

We propose that the requests to gain ethical approval for PI and PE activities stem largely from three main issues.

Firstly, ‘grey’ areas, such as a blurring of the boundary between qualitative research and PI and PE activities, including confusion amongst the research community over the differences between research involvement, engagement and participation.

Secondly, a perception amongst the research community that it is best to seek ethical approval ‘just in case’ or to ‘be on the safe side’, e.g. if asked by journal editors when trying to publish, rather than complete appropriate risk assessments to address any ethical considerations when carrying out PI and PE.

And finally, lack of knowledge of an alternative recognised process on how to evidence that PI and PE activities with the public have been conducted in an ethical manner, if not approved by an NHS REC or local IRB.

Despite guidance indicating other ways to address ethical concerns in PI and PE [ 12 , 13 , 14 , 15 , 16 , 17 , 18 ], researchers, funders and publishers appear to be turning increasingly to university IRBs as the (perceived) ultimate arbiters of deciding ethical issues related to PI and PE activities. We see the need to highlight this as a growing problem and suggest ways the issues above can be overcome. We will firstly explore in more detail the distinction between qualitative research and PI and PE activities before outlining examples from our own experience around the three issues identified, and then proceeding to make recommendations for moving forwards.

Public involvement and engagement vs. qualitative research

Distinguishing between whether activities with members of the public constitute PI and PE or qualitative research (and therefore require ethical approval) is a particularly ‘grey’ area [ 19 ]. This is especially true when consulting with a number of people at one time in what is usually referred to as a ‘focus group’. Going forward, it may be helpful to distinguish between ‘focus groups’, which are used for research, and ‘discussion groups’ used for PI and PE [ 20 ].

Several authors and organisations have described the difference between the two activities and developed useful side-by-side comparisons [ 19 , 20 ]. In focus groups which are part of research, people attending are research participants who receive a standard Participant Information Sheet and provide informed consent. Their input will usually be recorded via an audio device, transcribed verbatim, treated as ‘data’, and systematically be analysed to answer a research question. For this, ethical approval is usually required. On the other hand, the contributions of people attending PI discussion groups will be recorded only as key points (e.g. a list of key themes emerging or key priorities discussed by the group in relation to a specific topic) to help shape and guide the research itself, such as agreeing which research outcome measures to use, helping to shape the intervention or the development of data collection materials like participant information sheets or interview guides. PI discussion groups do not require ethical approval but should be conducted in an ethical manner. Those involved should still be provided with information about the activity up front to ensure they are clear what their involvement will entail, and they may be asked to provide agreement or consent, but not in the formally documented way required for research. This is discussed in more detail in the recommendations.

Another grey area concerns whether direct quotes gathered from people in a discussion group can be used in a publication. Whilst ethical approval is not required for this, we do advise gaining documented agreement if you wish to do this, e.g. an email from the group member agreeing to quotes being used in a publication to illustrate the key points identified (not as data). In some cases, researchers will need to combine PI activities with a qualitative research approach and there may be confusion regarding which activities require approval. For example, an investigator may wish to interview new mothers as research participants to get their views on motherhood (research participation). This would require ethical approval. But prior to interviews, they may want to involve a separate group of new mothers in a discussion to help shape the topic guide for the interviews (PI). This would not need ethical approval [ 21 ].

The extent of the problem - examples from our own experience

Through requesting examples from colleagues on their experiences, we uncovered many different situations within our own institution highlighting a difference of opinion on whether research ethics should be sought for PI and PE activity. We here outline three examples, giving the background to the project, the activity undertaken and the issues encountered.

Writing a training program with charity service users and staff – request from charity and publication to seek ethical approval from the university IRB for the project .

This project involved service users and charity staff in writing a mental health training curriculum for staff to identify depression in service users. Staff and service user input was sought through online meetings and email feedback. The attendees gave their opinions (based on their lived experience) on what should be included in the curriculum, and the key points were summarised to inform curriculum development. The information they gave was not treated as data to answer a research question and was not systematically analysed using qualitative methods. In this respect, HRA state that ’if you are collecting opinions rather than study data, your activity is likely an involvement activity’ [ 22 ].

Regardless of the above considerations, the project lead was asked by third sector organisations to seek university IRB approval, to ensure the service users would be treated in an ethical manner. An academic colleague agreed this was a good idea ‘just in case’ it was questioned by others, in particular by a journal editor when seeking to publish (which indeed it was). However, we view this as unnecessary given the activity was not classed as research and therefore not in the remit of the IRB. The IRB provided written agreement that ethical review was not required for this project and the project team agreed a standard engagement risk assessment would consider and address any ethical issues.

Co-producing an educational online resource for school children – request from publication to seek ethical approval for the project .

This co-production project working with researchers, a PI and PE professional, school teachers and web designers aimed to develop an educational online resource for school age children and their teachers. This interdisciplinary team of experts were involved in four online workshops to support the delivery and development of a website that would support teachers and enhance learning. All individuals involved fully signed up to the coproduction focus of the project and provided verbal agreement to take part in the workshops and off-line discussions. However, when trying to publish the co-production process, the journal editor stressed that according to journal policy ‘research involving human subjects, human material or human data must have been approved by an appropriate ethics committee’.

The authors explained that the project did not involve human subjects, human material or human data (as it was not research) and therefore in their opinion did not require ethical approval. The journal editor disagreed, arguing that the project was a research study that collected and analysed data, and that the teachers and web designers involved in this project were human participants of the study and data had been generated of their opinions. The editor recommended seeking either retrospective ethical approval or else removing all human data. The team saw no alternative but to withdraw their original manuscript and submit the work elsewhere.

Co-production project involving people from minority ethnic backgrounds in discussion about inclusive health research – project investigators not comfortable including quotes from public contributors due to lack of informed consent.

This project involving researchers, an artist, charity project workers serving the most ethnically diverse ward in Wales and local residents aimed to answer the question: ‘How can people from minority ethnic backgrounds influence health research in terms of both what and how this research is done?’ Eight co-production workshops drawing on the participatory democracy approach were held and delivered a set of recommendations for the health research community. In advance of these workshops, a university IRB Chair helped to clarify that ethical approval was not needed.

When publishing this work, researchers did not include quotes obtained from the workshops as informed consent had not been sought (as it was not research) [ 23 ]. On reflection, the authors would like to have gained agreement for the residents’ quotes to be used, in the absence of the requirement for documented informed consent.

Identified exceptions

Whilst PI and PE activities do not generally require ethical approval, there are at least two example scenarios where approval is required. Firstly, for example, when systematically comparing two methods of involvement and/or engagement to understand which is better i.e. answering a research question about PI/PE to produce generalisable or transferable findings. Secondly, when public members come into direct contact with study participants or their data e.g. if assisting with conducting research interviews or analysing the transcripts. In this situation, ethical approval is required because human participants are involved in the research.

Recommendations for moving forwards

We encourage the research community, including researchers, publishers, reviewers, funders and ethics committees to better appreciate the difference between PI and PE and research involving human participants; to recognise that all involved stakeholders operate within professional boundaries; and to work together to agree an alternative accepted approach when the PI and PE activity raises ethical considerations (e.g. when working with vulnerable groups or publishing of public contributor quotes). The responsibility of determining whether research ethic approval is required falls on the individuals/team planning the activity. We understand that it is tempting to seek research ethical approval for PI and PE activity ‘just in case’ or ‘to be on the safe side’, but we do believe this is detrimental for several reasons including:

Sustains the confusion between qualitative research and PI and PE activity, and the different purposes of each.

Wastes valuable researcher and committee time and resources.

Undermines the importance of the research ethics approval process.

Delays PI and PE activities in the research process, potentially leading to missing out on the benefits of earlier involvement.

Undermines coproduction principles such as equality and shared responsibility between researchers and members of the public. The process of acquiring ethical approval itself asserts a hierarchy whereby a researcher is identified as Chief/Principal investigator, and other members of the team are listed below an identified leader.

Acts as an additional barrier and disincentive to researchers carrying out PI and PE activity.

Simple flow diagram to support researchers to decide on the need for research ethical approval via an IRB

There is a need to address this growing problem, via education and generating solutions acceptable to the community as a whole, providing confidence in decisions made and assurances that the health and safety and any risks associated with the proposed PI and PE activity have been carefully considered and approved. Here we present key recommendations for those conducting public involvement and engagement activities based on our internal guidance (Appendix 1) for alternative courses of action moving forwards when faced with these challenges.

Purpose - Consider the purpose of the activity. Is it to answer a scientific or clinical question (research) or help shape, guide or disseminate the research (PI/PE)? If you are unsure if your project is research, you can consult the UK Health Research Authority’s ‘Is my study research’ decision tool. Following response to three questions, (1. Are the participants in your study randomised to different groups? 2. Does your study protocol demand changing treatment/care/services from accepted standards for any of the patients/service users involved? 3. Is your study designed to produce generalisable or transferable findings? ) The tool confirms if your study would be considered as research. This result can be downloaded and further advice can be sought [ 24 ]. The HRA table ‘Defining Research’ can also help provide clarification [ 25 ].

Internally, a simple flow diagram (Fig.  1 ) has been created to support researchers in making a decision on the need for research ethics approval when carrying out public involvement activity.

Risk assessment – To ensure PI and PE activities are conducted in a safe and ethical manner, particularly when engaging and/or involving ‘vulnerable’ groups, refer to published guidance on conducting ethical PI&E [ 12 , 13 , 14 , 15 , 16 ], consider completing a specifically designed PI and PE risk assessment (See Appendix 2 for an example) or using the PIRIT tool [ 26 ]to assess your planned activities and undertake adequate training (See Appendix 2 for an example). Use the same considerations as you might for research or teaching e.g. what to do if an individual becomes upset in a discussion group, how to support them, where to refer them. Also consider safety, protection of anonymity and confidentiality of personal data. Use the UK Standards on Public Involvement [ 4 ] to guide your thinking around accessibility and inclusivity when completing the assessment. If possible, involve a public contributor and have this signed off by a senior academic/responsible member of staff in your organisation.

Adequate information and agreement to take part – Ensure that public members being invited to take part in PI and PE activity agree for you to use their anonymous quotes in any output. But understand that standard Participant Information Sheets and Informed Consent Forms are not required as formal consent is not required.

Language – To avoid confusion for reviewers and publishers, think carefully about the language you use to describe your PI and PE activities. For example, use the term ‘discussion group’ rather than ‘focus group’; refer to members of the public as ‘attendees’ not ‘participants’ and input as ‘contributions’ rather than ‘data’; and ‘summarising key points or themes’ as opposed to ‘thematic analysis’ when describing your activities (if that is indeed what you are doing).

Written confirmation – Some institutions have established infrastructure to support researchers through a self-assessment process for governance and ethics, providing a confirmatory statement as to whether ethical approval is required if challenged by funders and publishers [ 27 ]. However, not all institutions have this facility and until this area of contention is resolved, some individuals may wish to seek written confirmation from their local IRB. In our experience, a letter confirming approval is not required is acceptable by journal editors. Liaise with your local IRB to determine if this is within their remit.

Training – The development and inclusion of training for researchers and support staff is required on when to seek ethical approval and how to effectively manage ethical, risks, and health and safety aspects of PI and PE in a considered, widely accepted and non-burdensome way.

Conclusions

Our experience suggests that ambiguity remains in the academic community about whether ethical approval is needed for PI and PE activities. We believe this stems from (1) the grey area between qualitative research and PI and PE activities; (2) seeking approval ‘just in case’ they are requested by funders, publishers or authorities (based on previous experience) (3) funders, publishers and authorities not being clear in the distinction and equally asking for approval ‘just in case’ and (4) a lack of an alternative recognised way to evidence that ethical issues have been considered and mitigated against. We have used real world examples to demonstrate the issues encountered in a single institution and make several recommendations aimed at researchers for addressing this area of contention going forward. We appreciate that our views may be framed by our experience of conducting PI&E in a healthcare context and in the UK, and the experiences of researchers in other disciplines and countries may vary significantly.

We hope this commentary triggers debate in the community to highlight, educate and clarify the position surrounding research ethics and PI and PE activity amongst researchers, funders and journal editors. Our experience shows that this issue is effectively acting as a barrier to researchers conducting PI and PE activity and publishing PI and PE learning. An alternative recognised process needs to be established by the community to resolve this growing detrimental development.

Data availability

Not applicable.

Abbreviations

Health Research Authority

Institutional review board

• Public engagement
• Public involvement

Research Ethics Committee

Research Excellence Framework

Health Research Authority. Putting people first - embedding public involvement in health and social care research 2023 [ https://www.hra.nhs.uk/planning-and-improving-research/best-practice/public-involvement/putting-people-first-embedding-public-involvement-health-and-social-care-research/

Lang I, King A, Jenkins G, Boddy K, Khan Z, Liabo K. How common is patient and public involvement (PPI)? Cross-sectional analysis of frequency of PPI reporting in health research papers and associations with methods, funding sources and other factors. BMJ Open. 2022;12:e063356.

Article   PubMed   PubMed Central   Google Scholar  

Eberl M, Joseph-Williams N, Nollett C, Fitzgibbon J, Hatch S. Overcoming the disconnect between scientific research and the public. Immunol Cell Biol. 2023;101(7):590.

Article   PubMed   Google Scholar  

UK Public Involvement Standards Development Partnership. UK Standards for Public Involvement 2019 [ https://drive.google.com/file/d/1U-IJNJCfFepaAOruEhzz1TdLvAcHTt2Q/view

Staniszewska S, Denegri S, Matthews R, Minogue V. Reviewing progress in public involvement in NIHR research: developing and implementing a new vision for the future. BMJ Open. 2018;8(7):e017124.

Hatch S, Fitzgibbon J, Tonks AJ, Forty L. Diversity in patient and public involvement in healthcare research and education—realising the potential. Health Expect. 2023;27(1):e13896.

Research Excellence Framework. Research Excellence Framework 2029 2024 [ https://www.ref.ac.uk/

Nguyen Thanh H, Cheah P, Chambers M. Identifying ‘hard-to-reach’ groups and strategies to engage them in biomedical research: perspectives from engagement practitioners in Southeast Asia. Wellcome Open Res. 2019;4(102).

National Institute for Health and Care Research. Briefing notes for researchers-public involvement in NHS, health and social care research. 2021.

National Co-ordinating Centre for Public Engagement. Introducing Public Engagament 2024 [ https://www.publicengagement.ac.uk/introducing-public-engagement

Health Research Authority. What do I need to do? 2020 [ https://www.hra.nhs.uk/planning-and-improving-research/best-practice/public-involvement/what-do-i-need-do/

Pandya-Wood R, Barron D, Elliott J. A framework for public involvement at the design stage of NHS health and social care research: time to develop ethically conscious standards. Res Involv Engagem. 2017;3(3).

Canadian Institutes of Health Research. Ethics Guidance for Developing Partnerships with Patients and Researchers 2020 [ https://cihr-irsc.gc.ca/e/51910.html#2

Hersh D, Israel M. C. S. The ethics of patient and public involvement across the research process: towards partnership with people with aphasia. Aphaisology. 2021.

Abma T, Groot B, Widdershoven B. The Ethics of Public and Service User Involvement in Health Research: the need for participatory reflection on everyday ethical issues. Am J Bioeth. 2019;19:23–5.

Groot B, Abma T. Ethics framework for citizen science and public and patient participation in research. BMC Med Ethics. 2022;23.

National Institute for Health and Care Research. Ethical dimensions of community engagement and involvement in global health research 2021 [ https://www.nihr.ac.uk/documents/ethical-dimensions-of-community-engagement-and-involvement-in-global-health-research/28258#references

Martineau JT, Minyaoui A, Boivin A. Partnering with patients in healthcare research: a scoping review of ethical issues, challenges, and recommendations for practice. BMC Med Ethics. 2020;21.

Hanley B, Staley K, Stewart D, Barber R. Qualitative research and patient and public involvement in health and social care research: What are the key differences? 2019 [ https://www.learningforinvolvement.org.uk/content/resource/qualitative-research-and-patient-and-public-involvement-in-health-and-social-care-research-what-are-the-key-differences/

Doria N, Condran B, Curtis Maillet LB, Dowling D, Levy L. A. Sharpening the focus: differentiating between focus groups for patient engagement vs. Qualitative Res Res Involv Engagem. 2018;4.

Morgan H, Thomson G, Crossland N, Dykes F, Hoddinott P, ‘BIBS’ study team. Combining PPI with qualitative research to engage ‘harder-to-reach’ populations: service user groups as co-applicants on a platform study for a trial. Res Involv Engagem. 2016;24(7).

Health Research Authority. Accessing study support and advice services 2024 [ https://www.hra.nhs.uk/planning-and-improving-research/research-planning/access-study-support-advice-services/

Bridges S, Lamont-Robinson C, Herbert A, Din M, Smith C, Ahmed N et al. Talking trials: an arts-based exploration of attitudes to clinical trials amongst minority ethnic members of the South Riverside Community of Cardiff Health expectations. 2023(26):3.

Health Research Authority. Is my study research? 2024 [ https://www.hra-decisiontools.org.uk/research/

Health Research Authority. Defining Research Table 2022 [ https://www.hra-decisiontools.org.uk/research/docs/DefiningResearchTable_Oct2022.pdf

Marie Curie Research Centre. Public Involvement in Research Impact Toolkit (PIRIT) [ https://www.cardiff.ac.uk/marie-curie-research-centre/patient-and-public-involvement/public-involvement-in-research-impact-toolkit-pirit

University of Surrey. Ethics Guide 2022 [ https://issuu.com/universityofsurrey/docs/surrey_ethics_cop_v20_single?fr=sNDM5NzUwNzU0OTI

Download references

Acknowledgements

We are grateful to our colleagues Martina Svobodoba, Sarah Bridges and Dr Vicky Shepherd for providing useful insights and resources from their experiences and to Dr Emma Yhnell for her helpful review and comment on the first draft.

Author information

Authors and affiliations.

Centre for Trials Research, Cardiff University, 7th Floor, Neuadd Meirionnydd, Heath Park Campus, Cardiff, UK

Claire Nollett

Division of Infection and Immunity, School of Medicine, Cardiff University, Cardiff, UK

Matthias Eberl

Systems Immunity Research Institute, Cardiff University, Cardiff, UK

School of Medicine, Lead Public Contributor, Cardiff University, Cardiff, UK

Jim Fitzgibbon

Division of Population Medicine, School of Medicine, Cardiff University, Cardiff, UK

Natalie Joseph-Williams

Health and Care Research Wales Evidence Centre, Cardiff, UK

Public Involvement and Engagement Team, School of Medicine, Cardiff University, Cardiff, UK

Sarah Hatch

You can also search for this author in PubMed   Google Scholar

Contributions

CN and SH drafted the first version; CN, SH, ME and NJW added case studies; ME, NJW and JF contributed to revised versions and all authors read and approved the final manuscript.

Corresponding authors

Correspondence to Claire Nollett or Sarah Hatch .

Ethics declarations

Ethics approval and consent to participate.

Not applicable – no participants involved.

Consent for publication

Competing interests.

The authors declare that they have no competing interests.

Authors information

CN is Academic Lead for Public Involvement and Engagement in the Centre for Trials Research, Cardiff University. SH is the Public Involvement and Engagement Manager for the School of Medicine, Cardiff University, alongside researchers ME and NJW who are the Joint Academic Leads for Public Involvement and Engagement in the School of Medicine, Cardiff University. ME is also the Engagement Lead for the Systems Immunity Research Institute at Cardiff University and the Engagement Secretary for the British Society for Immunology. JF was the Lead Public Contributor in the School of Medicine at the time of writing.

Additional information

Publisher’s note.

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

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary material 2, rights and permissions.

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

Reprints and permissions

About this article

Cite this article.

Nollett, C., Eberl, M., Fitzgibbon, J. et al. Public involvement and engagement in scientific research and higher education: the only way is ethics?. Res Involv Engagem 10 , 50 (2024). https://doi.org/10.1186/s40900-024-00587-x

Download citation

Received : 15 February 2024

Accepted : 24 May 2024

Published : 31 May 2024

DOI : https://doi.org/10.1186/s40900-024-00587-x

Share this article

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

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

Provided by the Springer Nature SharedIt content-sharing initiative

• Research ethics committee
• Ethical approval

Research Involvement and Engagement

ISSN: 2056-7529

• General enquiries: [email protected]

State laws threaten to erode academic freedom in US higher education

Associate Professor of Political Science, Trinity College

Disclosure statement

Isaac Kamola is the Director of the Center for the Defense of Academic Freedom housed at the American Association of University Professors, which seeks to promote and defend academic freedom within higher education. The center is funded by the Mellon Foundation.

Trinity College provides funding as a member of The Conversation US.

View all partners

Over the past few years, Republican state lawmakers have introduced more than 150 bills in 35 states that seek to curb academic freedom on campus. Twenty-one of these bills have been signed into law.

This legislation is detailed in a new white paper published by the Center for the Defense of Academic Freedom , a project established by the American Association of University Professors, or AAUP. Taken together, this legislative onslaught has undermined academic freedom and institutional autonomy in five distinct and overlapping ways.

1. Academic gag orders

As detailed in the report, state legislators introduced 99 academic gag orders during legislative sessions in 2021, 2022 and 2023. All of the 10 gag orders signed into law were done so by Republican governors. These bills assert that teaching about structural racism, gender identity or unvarnished accounts of American history harm students.

These gag orders are widely known as “divisive concept” or “anti-CRT” bills. CRT is an acronym for critical race theory, an academic framework that holds racism as deeply embedded in America’s legal and political systems. The partisan activists, such as Christopher Rufo , have used this term to generate a “ moral panic ” as part of a political response to the 2020 Black Lives Matter protests.

For example, in April 2022, Florida Gov. Ron DeSantis signed House Bill 7, the “ Stop Woke Act .” The law defines a “divisive concept” as any of eight vague claims. They include claims that “Such virtues as merit, excellence, hard work, fairness, neutrality, objectivity, and racial colorblindness are racist or sexist.”

U.S. District Judge Mark Walker described this law as “ positively dystopian .” He noted that the government’s own lawyers admitted that the law would likely make any classroom discussion concerning the merits of affirmative action illegal. The vague wording of these gag orders has a chilling effect , leaving many faculty unsure about what they can and cannot legally discuss in the classroom.

2. Bans on DEI programs

The expansion of diversity, equity and inclusion – or DEI – services on campus was a major outcome of the racial justice protests in 2020. By 2023, however, the legislative backlash was in full swing. Forty bills restricting DEI efforts were introduced during the 2023 legislative cycle, with seven signed into law.

For example, Texas’ Senate Bill 17 drew directly from model policy language developed by Rufo and published by the Manhattan Institute , a right-wing think tank. SB 17 banned diversity statements and considerations in hiring. It also restricted campus diversity training and defunded campus DEI offices at Texas’ public universities.

As detailed in the AAUP white paper, only a handful of people testified in favor of SB 17, and almost all had stated or unstated affiliations with right-wing think tanks. In contrast, more than a hundred educators and citizens testified, or registered to testify, against the bill. Since its passage, Texas public universities have seen the closing of DEI programs and reduced campus services for students from minority populations. For example, after the Legislature accused the University of Texas-Austin of violating SB 17, the school was forced to shut down its DEI office. This involved laying off 40 employees .

3. Weakening tenure

Tenure was developed to shield faculty members from external political pressure. The protections of tenure make it possible for faculty to teach, research and speak publicly without fear of losing their jobs because their speech angers those in power. As detailed in the report, however, during the 2021, 2022 and 2023 legislative sessions, 20 bills were introduced, with two bills weakening tenure protections signed into law in Florida and another in Texas .

In Florida, for example, SB 7044 created a system of post-tenure review, empowering administrators to review tenured faculty every five years. The law further empowers administrators to dismiss those whose performance is deemed unsatisfactory. The law also requires that faculty post course content in a public and searchable database.

The AAUP criticized the law , noting that SB 7044 has “substantially weakened tenure in the Florida State University System and, if fully implemented as written,” would effectively “eliminate tenure protections.” Now even tenured faculty have reason to fear that what they teach might be construed as a “divisive concept,” as CRT, or as promoting DEI.

4. Mandating content

Lawmakers in several states have also passed legislation mandating viewpoint diversity, establishing new academic programs and centers to teach conservative content and shifting curricular decision-making away from the faculty.

For example, Florida’s Senate Bill 266 expanded the Hamilton Center for Classical and Civic Education at the University of Florida, without faculty input or oversight. The original proposal for the Hamilton Center stated that the center’s goal was to advance “ a conservative agenda ” within the curriculum.

SB 266 also gave the governing boards overseeing the university and college systems the authority to decide which classes count toward the core curriculum. This power was exercised in November 2023 after Manny Diaz, the education commissioner in Florida, requested that the boards remove an introduction to sociology course . He stated on social media that the discipline had been “ hijacked by left-wing activists and no longer serves its intended purpose as a general knowledge course for students.”

5. Weakening accreditation

The accreditation process is an obscure area of academic governance whereby colleges and universities regularly subject themselves to external peer review. Nonprofit accrediting agencies conduct these institutional performance reviews.

As detailed in the report, during the 2021-23 legislative cycles, six bills were introduced – three of them were passed into law – weakening the accreditation process, thereby making it easier for political interests to shape university policy.

For example, University of North Carolina-Chapel Hill’s accreditor, the Southern Association of Colleges and Schools Commission on Colleges, warned the school’s board of trustees that establishing the School of Civic Life and Leadership without faculty oversight and consultation raised serious concerns about institutional independence. The Legislature responded with Senate Bill 680 , which would require that North Carolina public universities choose a different accrediting agency each accreditation cycle. Eventually passed as part of the omnibus House Bill 8, this policy allows schools to “shop” for an accrediting agency less likely to object to such political interference in the curriculum.

These five overlapping and reinforcing attacks on academic freedom and institutional autonomy threaten to radically transform public higher education in ways that serve the partisan interests of those in power.

• Academic freedom
• US higher education
• Education law
• Higher ed attainment
• Diversity, equity and inclusion

Head of School, School of Arts & Social Sciences, Monash University Malaysia

Chief Operating Officer (COO)

Clinical Teaching Fellow

Data Manager

Director, Social Policy

E-SEARCH Opens Doors to Higher Education

The new research-focused mentorship program unlocks the potential to pursue academic careers in engineering for undergraduates.

• by Jessica Heath
• June 05, 2024

In the spring quarter of 2023, over three dozen engineering students gathered in a UC Davis conference room with anticipation. Graduate student researchers presented their project ideas while undergraduates milled about, looking for a project that sparked their interests. These first participants in the new program, E-SEARCH , were all hoping to have a research meet-cute and find their perfect mentor-mentee match. 

Initiated as part of the College of Engineering’s Next Level strategic vision for education , E-SEARCH pairs undergraduate engineering students with graduate student researchers to work on a research project of their choosing over the course of a quarter, presenting their findings at an end-of-term showcase. Through this partnership and hands-on research experience, the program aims to equip students with the tools they will need for graduate school or careers in engineering.

Thus far, E-SEARCH has held three cycles (spring and fall of 2023 and spring of 2024), with a fourth scheduled for this summer. In total, the program has supported 53 teams of students across all engineering disciplines conducting research on topics that include converting carbon dioxide on Mars into methanol, an efficient fuel source; using machine learning to detect calcium deficiencies in hydroponically grown lettuce; and training a neural network to separate the lung-specific data from X-ray images using a collection of chest X-rays and their associated lung masks. 

“It’s a low-stakes, high-reward opportunity to get your foot in the door and explore research,” said Patrick Cunningham, a Ph.D. candidate in civil and environmental engineering who served as a mentor in the fall of 2023. “It offers the freedom to craft your own research plan and explore a topic of shared interest. It’s really exciting to have that flexibility in a ‘free-to-fail’ environment.”  

The response has been immensely positive. Mentors said they learned how important preparation, planning, communication and listening are to having a successful research relationship. 

Mentees, on the other hand, stated that participating in the program helped them make more informed decisions about pursuing graduate school, learn how to feel comfortable asking questions and become more adaptable as obstacles arise. 

Isidro Valdez-Lopez, a second-year civil and environmental engineering major, was a mentee in the initial cohort last spring. He worked with civil and environmental engineering Ph.D. candidate Mandeep Singh Basson on research into Sacramento area levees, using performance-based analysis to improve their efficiency and identify how climate change-induced water fluctuations have affected their failure probability in the past. 

Valdez-Lopez joined the program to connect with people more academically advanced then himself. His weekly meetings and chats with Singh Basson gave him new skills in experimenting and a better outlook on math. Additionally, as an undocumented student, Valdez-Lopez has not had many chances to participate in research.  

“It gave me an opportunity to expand as a UC Davis student, enabling me to take part in programs not usually offered to undocumented students,” he said. “It made me want to strive for more.” 

The program is part of the college’s commitment to create opportunities for undergraduate engineering students, no matter their major or discipline, for hands-on research early in their education. 

“We continually strive to prepare and empower our students to become future leaders in the field. They will use the foundational skills they learned at UC Davis to engineer a better planet, a better community for all of us,” said Dean Richard Corsi. “Learning how to research early on and knowing that a path to academia is open to them is a huge part of that.” 

E-SEARCH, similar to other College of Engineering programs like LEADR and AvenueE , creates a path for undergraduates to explore the possibility of engineering research as a career by connecting with others with shared experiences. As one mentee reflected:

“One of the things I’ve learned is how much alike we are with people from higher levels of education. I think my talks with my mentor about what was going on in our lives is what really ended up motivating me to aim higher, not just in school but in life as a whole.”  

This article was originally featured in the Spring 2024 Engineering Progress Magazine .

Primary category, secondary categories.

Advertisement

Leadership development programming in higher education: an exploration of perceptions of transformational leadership across gender and role types

• Original Paper
• Published: 11 October 2021
• Volume 27 , pages 297–312, ( 2021 )

Cite this article

• Kevan W. Lamm   ORCID: orcid.org/0000-0001-5238-8842 1 ,
• L. Rochelle Sapp 2 ,
• Nekeisha L. Randall 3 &
• Alexa J. Lamm 1  

6565 Accesses

5 Citations

2 Altmetric

Explore all metrics

Transformational leadership, a type of leadership commonly promoted within higher education, has been shown to positively affect performance, collaborative behavior, and goal accomplishment. Such skills may correlate with the level of job responsibility one has been given and the technical, human, and conceptual skills needed for one to be successful. This study sought to bridge a research gap by exploring correlations between transformational leadership and skills-approach leadership with an exploration of the role of gender within perceptions. An unexpected result based on gender was found: As females achieve higher roles within the Land-Grant University System, the perception of their transformational leadership decreases while that of males increases. Transformational leadership and skills-approach leadership is discussed within the context of gender.

Similar content being viewed by others

Authoritarian leadership styles and performance: a systematic literature review and research agenda

Leadership in healthcare education

What Is Leadership?

Avoid common mistakes on your manuscript.

Introduction

Higher education has always been faced with challenges and opportunities that prompt the field to seek progressive, mission-critical ways to move forward. Worldwide, colleges and universities have faced challenges such as competition for funding and students (Vieira da Motta & Bolan, 2008 ). Recent global events such as the COVID-19 pandemic and demands for racial equality and inclusion have required higher education to be more flexible and innovative than ever before. Institution closures and re-openings along with heightened reliance on technology for online instruction and communication require faculty and staff to adapt and “[be] ready for anything” (Major, 2020 p. 266). Challenges that existed prior to recent events, such as those related to institutional finances, public scrutiny, and the impact of federal decisions and economic issues, have magnified. In a time when vision, direction, and trust is needed, the case for strong leadership within higher education administration continues to be vital. In fact, “the selection and training of good administrators is widely recognized as one of American industry’s most pressing problems” (Katz, 1955 p. 33). Though this observation was made in 1955, its sentiment still rings true today and for every country.

Leadership is predicted to remain the top human resource challenge through the year 2025 at minimum (Society for Human Resource Management, 2015 ). Though there are a plethora of individuals working in higher education, administration and leadership challenges exists because there is a dearth of individuals who specifically possess enhanced leadership skills and competencies that twenty-first century educational institutions demand and need. Transformational leadership is a theory and practice known for helping fill the gap between leadership pipeline issues and well-qualified future leaders (e.g. Lamm et al., 2016 ). It is a well-studied, multi-dimensional theory consisting of attitudes and behaviors that inspire followers to reach improved levels of determination and commitment for the betterment of the whole, leading to overall improved performance (Ayman & Korabik, 2010 ; Bass & Avolio, 1993 ). Perceived as a bureaucratic and professional organization (Nica, 2013 ), higher education may benefit from having more employees emanate this type of leadership (Lamm et al., 2016 ). However, not all administrative positions have the same span of control nor expectations associated with them. Therefore, a more nuanced approach may be warranted. Specifically, Katz ( 1955 ) proposed the skills approach model whereby differing levels of leadership responsibility may require different areas of focus. Leadership development researchers and practitioners have an obligation to continue exploring how skill development should be facilitated and at what stage in an individual’s career the development of certain competencies should be encouraged, which can aid in one’s development of transformational leadership.

Within the higher education leadership literature, the role of gender, and gender experiences has been established (e.g. Dunn et al., 2014 ), yet, “very little research examines gender differences in [higher education] leadership styles in any systematic way” (Madden, 2011 p.63). Furthermore, there remains a need to more empirically examine higher educator leader perceptions of leadership from a gender-based perspective. For example, Dunn et al. ( 2014 ) state “Future research is needed to systematically compare the experiences of female leaders in various types of academic institutions to inform how gender impacts leadership experiences” (p. 17).

The study at hand investigates the perception higher education leaders have of their own transformational leadership capacity and whether that perception is related to other characteristics, specifically gender and/or administrative level held at their respective college or university. Podsakoff et al. ( 1990 ) model of transformational leadership is applied to this topic and Katz’s ( 1955 ) model of skills-based leadership is used to organize and analyze data to explore new leadership insights.

Conceptual framework

Training leaders through leadership development programming has been found to increase transformational leadership capacity in faculty within the Land-Grant University System (Lamm et al., 2016 ). The Land-Grant University System is a group of higher education institutions in the United States federally established by law beginning in 1862 to provide education for citizens in each state (Association of Public & Land-Grant Universities, 2016 ). Congruent with the goals of this study, exploring how transformational leadership aligns with the skills-based approach can aid leadership theory and development efforts for the benefit of current and future leaders in higher education.

• Transformational leadership

Podsakoff et al.’s ( 1990 ) transformational leadership conceptualization is helpful in identifying leaders who, “transform or change the basic values, beliefs, and attitudes of followers so that they are willing to perform beyond the minimum levels specified by the organization” (p. 108). More specifically, these types of leaders were considered through the lens of the Transformational Leadership Inventory (TLI). This multi-dimensional inventory measures transformational leadership capacity within leadership development program participants (Lamm et al., 2016 ) through four dimensions of transformational leadership: (1) core transformational leadership behaviors, (2) individualized support, (3) intellectual stimulation, and (4) high performance expectations (Podsakoff et al., 1990 ).

By acknowledging that each follower is unique (Bass & Riggio, 2006 ), transformational leaders show respect (Podsakoff et al., 1990 ) by providing individualized support catered to each individual’s growth and goals (Bass & Riggio, 2006 ; Bono & Judge, 2004 ). Intellectual stimulation fosters creativity and new ideas (Bono & Judge, 2004 ) when followers are encouraged to think outside the box and challenge their own assumptions about how to accomplish work goals (Podsakoff et al., 1990 ). Additionally, setting high expectations raises the standard of excellence and can generate enthusiasm among followers (Bass & Riggio, 2006 ; Bono & Judge, 2004 ; Podsakoff et al., 1990 ), which helps them exceed beyond what they originally thought could be accomplished.

Skills-based leadership

In his seminal Harvard Business Review (HBR) article, Katz ( 1955 ) proposed that effective administration rests on three basic skills which vary in importance depending upon one’s level of administrative responsibility within an organization. Levels of responsibility are categorized as low, middle, and top. Due to the popularity and influence of Katz’s ( 1955 ) work, the article was reprinted in 1974 and 1986 (Peterson & Van Fleet, 2004 ). Katz’s definition of administrator can be akin to that of a leader: “…one who (a) directs the activities of other persons and (b) undertakes the responsibility for achieving certain objectives through these efforts” (Katz, 1955 p. 34). Thus, for the purposes of this study, academic administrators will be referred to synonymously as administrators and leaders. The three foundational skills of an administrator, or leader, are categorized as: technical, human, and conceptual and are recognized as being simultaneously valuable as independent and interdependent, as each complements the others (Katz, 1955 ).

Technical skills relate to specific, field-related work-tasks and an understanding of the processes and methods underlying such tasks (Katz, 1955 ). Due to technical skills being necessary for an organized entity, such as a university, to produce the products and services it has been created for (Northouse, 2013 ), these skills are the most concrete and identifiable; they involve, “specialized knowledge, analytical ability within that specialty, and facility in the use of the tools and techniques of the specific discipline” (Katz, 1955 p. 34).

Human skills are seen as important for employees on all foundational, middle, and top levels of an organization (Katz, 1955 ; Northouse, 2013 ) and refer to the ability to communicate and interact with others effectively and cooperatively, which includes resolving conflict and being a contributing member of a team (Peterson & Van Fleet, 2004 ). Human skill is connected to how an individual interacts with everyone, no matter how the person they are interacting with is categorized in the hierarchical structure.

When an individual reaches an administrative role in an organization that requires competencies beyond technical and human skills, “conceptual skill becomes increasingly more important with the need for policy decisions and broad-scale action.” (Katz, 1955 p. 37). Thus, conceptual skills require one to envision how sub-sets and functions of organizations are interdependent and how the organization or institution as a whole fit within larger environmental contexts such as an industry, a community, and a society (Katz, 1955 ).

Benefits of Katz’s ( 1955 ) theory are the recognition that anyone can be a leader and that taking an inventory of a person’s skills helps with the selection and placement of leaders (Katz, 1955 ). Northouse ( 2013 ) notes the alignment of the skills approach with the majority of leadership education curricula and the usefulness of how it frames what is taught in leadership development programs.

Literature review

Transformational leadership has been promoted for leadership development initiatives specifically oriented toward higher education leaders (e.g. Turnbull & Edwards, 2005 ). Additionally, a more in-depth look at career development needs for different levels of leadership and gender-specific experiences have also been promoted as necessary avenues of continued research (Dopson et al., 2019 ; Turnbull & Edwards, 2005 ).

The effectiveness of transformational leadership has been specifically studied in the field of higher education through topics such as diversity management perception (Brown et al., 2019 ), organizational culture and performance (Hambali & Idris, 2020 ), workplace engagement and spirituality (Arokiasamy & Tat, 2020 ), academic research (Hung et al., 2019 ), and quality management (Argia & Ismail, 2013 ). Regarding administrative role responsibility in particular, research advises higher education leaders to: establish a vision for their units, departments, and/or institutions; treat those who report to them with fairness and inclusivity; promote shared leadership and collaboration; and steward organizational values (Berson et al., 2016 ; Gigliotti, 2017 ; Pearce et al., 2018 ). Such characteristics relate to Podsakoff et al.’s ( 1990 ) description of transformational leadership’s prioritization of innovation and relationships. Higher education literature suggests that there is not only a shift to this type of leadership, but that leadership development efforts should be reconceptualized to include these competencies (Dopson et al., 2019 ).

Those who demonstrate transformational leadership behavior have been found to be rated as the most effective employees by both subordinates and superiors (Burke & Collins, 2001 ). Additionally, transformational leadership has been found to transcend different cultures (Carless, 1998 ) and aid in the transformation of higher education’s academic cultures (Thomas et al., 2015 ). Transformational leaders benefit organizations by helping with adaption to change (Kearns et al., 2015 ) and introducing new ideas and improving existing ones (Anthony & Schwartz, 2017 ). These transformational leaders “communicate powerful narratives about the future,” and “develop a road map before disruption takes hold” (Anthony & Schwartz, 2017 para. 25–29). Research shows that transformational leaders: have and exercise self-awareness; can work independently and across silos; build positive cultures; are willing to collaborate and see situations from different perspectives; build trust and can be trusted; are humble and ask for help when it is needed; make decisions with timeliness and purpose; and challenge, inspire, and empower others (Anthony & Schwartz, 2017 ; Thompson, 2012 ). Moreover, “[t]ransformational leaders articulate a vision, use lateral or nontraditional thinking, encourage individual development, give regular feedback, use participative decision-making, and promote a cooperative and trusting work environment” (Carless, 1998 p. 888). It has also been found that employees benefit from such characteristics by experiencing enhanced performance, well-being, and motivation (Fernet et al., 2015 ; Limsila & Ogunlana, 2008 ). While both transactional and transformational leadership have been measured in leadership behavior (e.g. Megheirkouni et al., 2018 ), “leaders who are more satisfying to their followers and who are more effective as leaders are more transformational and less transactional” (Bass, 1999 p. 11). Thus, even in higher education, it is suggested that leadership development efforts should be designed to emphasize transformational, rather than transactional, leadership models (Turnbull & Edwards, 2005 ).

Connecting transformational leadership and skills-based leadership

Higher education leadership influences every level of a college or university, including the strategic direction of the institution, the culture shared by faculty and staff, and the academic success of students (Nica, 2013 ). As highlighted in a higher education leadership development program design literature review by Dopson et al. ( 2019 ) found “[institutional] leadership is a contextual, processual, relational, social, political and temporal phenomenon” (p. 225). Therefore, it is also suggested that attention is paid to the unique needs leaders experience at different levels of seniority in the higher education hierarchy (Dopson et al., 2019 ; Turnbull & Edwards, 2005 ). Furthermore, promotion to a new level based on technical skills rather than human and conceptual skills has been observed in higher education as faculty are sometimes promoted to administrator levels based solely on their teaching or research experience; it is assumed they will be good leaders and will know how to self-correct their leadership methods (Vieira da Motta & Bolan, 2008 ). Promoting employees that have yet to acquire certain skills can be costly (Benson et al., 2018 ), with consequences of the leader’s impact possibly not observed until after the transition to the new administrative role (Dopson et al., 2019 ).

Though the Katz’s ( 1955 ) skills-based leadership theory has been used to study specific skills of ground-level leaders in manufacturing companies (Petkevičiūtė & Giedraitis, 2013 ) and the adequate preparation of on-line learning instructors (Muldrow, 2014 ), limited empirical efforts have been made to directly connect Katz’s ( 1955 ) theory with transformational leadership. Based on Podsakoff et al.’s ( 1990 ) description of transformational leadership, the present study conceptually links transformational leadership with the human and conceptual components of skills-based leadership. In the Chronicle of Higher Education , Maimon ( 2018 ) highlights this notion by explaining that:

Transformative leadership is more focused on relationships, open to multiple interpretations, adaptable to new situations, and more flexible in adjusting to new environments. The transformative leader is readier to multitask and capable of paying attention both to goals and to the process for achieving them. (para. 3)

Katz ( 1955 ) highlights the importance of human and conceptual components on the middle and top management leadership levels. The study at hand answers the call for leadership research within higher education and the Land-Grant University System to focus more on human and conceptual components rather than just technical skills (Lamm et al., 2016 ).

Leadership and gender

In addition to requiring specific competencies, transformational leaders must understand the value of diversity and how, in “promot[ing] diversity in academic leadership, the college or university should be a microcosm of the total society” (Nica, 2013 p. 192). The relationship between the diversity aspect of gender and leadership has received considerable analysis (e.g. Anim & Shotte, 2020 ; Bass, 2008 ). Leadership styles and experiences of women as well as development programs specifically designed for women have been studied in the context of higher education (see Dopson et al., 2019 ). Additionally, in the U.S. alone, it was observed that in 1972 only 17% of leadership positions were held by women (Bass, 2008 ) and in 2015 39.2% of such positions were held by women, including 65.7% of education administrators (U.S. Bureau of Labor Statistics, 2016 ). Despite observable shifts of gender representation within the workforce, recent research still indicates persistent differences between genders (Badura et al., 2018 ). In higher education, women remain underrepresented in leadership roles (Anim & Shotte, 2020 ; Nica, 2013 ) despite making up nearly half of all faculty roles and obtaining more than half of the undergraduate degrees earned in the U.S. (Judson et al., 2019 ). These trends are represented in countries outside the U.S. as well (e.g. Anim & Shotte, 2020 ). Moreover, though the significance was small, Baker et al. ( 2019 ) found that women were less likely to seek department chair roles, positions that are springboards into higher level administrative positions within higher education.

In a recent meta-analysis of predictors and moderators of motivation to lead, the relationship between gender and leadership outcomes was again observed (Badura et al., 2019 ). The authors suggested, “[p]art of succession planning is being able to identify future leaders so that they can receive additional training and experiences before they are called upon to fill leadership vacancies” (Badura et al., 2019 p. 17). It has been documented that for females in higher education there may exist barriers limiting access to networking, developmental opportunities, and subsequent recruitment for high level administrative positions (Bagilhole & White, 2008 ). There remains a gap in the literature specifically identifying what characteristics are valued within perceptions of leadership emergence therefore providing an entry into further developmental opportunities (Badura et al., 2018 ).

Katz’s ( 1955 ) skills-based approach has been applied to gender in a number of empirical studies with many using the Style Inventory Survey (Northouse, 2010 ) to measure technical, human, and conceptual skill. Research findings of such studies found that samples of female leaders in India scored very high on technical and human skill while their male counterparts scored high on conceptual skills (Kaifi & Mujtaba, 2010 ). A similar study involving Afghan leaders showed opposite results; males had high technical and human skills while females had higher conceptual skills (Mujtaba & Kaifi, 2011 ). Despite conflicting findings, a mixture of all three skills are needed at each level of administrative responsibility (Katz, 1955 ; Peterson & Fleet, 2004 ).

Research objectives

The purpose of this study was to explore relationships between academic leaders’ perceived transformational leadership capacity, corresponding administrative level, and self-reported gender. The study was guided by the following objectives:

Describe the self-perceived transformational leadership capacity of faculty members and administrators participating in a leadership development program for higher education.

Determine whether administrative role categories were statistically significantly related to self-perceived transformational leadership capacity.

Determine whether self-reported gender categories were statistically significantly related to self-perceived transformational leadership capacity.

Determine whether administrative role categories, by gender, were statistically significantly related to self-perceived transformational leadership capacity.

The following information expounds upon this study’s sample and data analysis process.

The sample for this study were participants in the LEAD21 leadership development program, which focuses on capacity-building in the areas of communication, conflict management, collaboration, and leading change (LEAD21, n.d. ). Participants are associated with the Land-Grant University System and its affiliated organizations such as the U.S. Department of Agriculture (USDA) or Non-Land-Grant Agricultural and Renewable Resources Universities, as well as strategic partners who work alongside institutions to connect academic research with outreach efforts and the needs of the general public (LEAD21, n.d. ). Representing various U.S. states and territories, participants were nominated for the program by their institution based on their leadership potential within the Land-Grant University System, with role titles ranging from assistant professor to dean. LEAD21 caters to emerging and top Land-Grant University System administrators and uses adult learning theory to help participants connect program content to their past work experience, which also aligns with what Katz ( 1955 ) believed was necessary for effective skill development. The program consisted of three seminars, ranging from four to six days, conducted over the course of nine months. LEAD21 participants also participate in periodic check-in calls and activities between seminars. All data were collected prior to the start of the program to serve as a pre-program baseline value.

The study at hand expounds upon Lamm et al.’s ( 2016 ) finding that LEAD21 develops self-assessed transformational leadership capacity within participants by an average of 7%. It further analyzes transformational leadership by answering the call to study multiple classes of LEAD21 participants rather than just one class (Lamm et al., 2016 ). Four classes of LEAD21 participants were included, creating an overall convenience sample of 340 respondents. The sample consisted of 84 members from the 2015 to 2016 cohort, 85 members from the 2016 to 2017 cohort, 80 members from the 2017 to 2018 cohort, and 91 members from the 2018 to 2019 cohort. Individuals were asked to self-report their gender; 195 individuals identified as male, 143 individuals identified as female, and two individuals did not provide a response.

Data analysis

Respondents were also asked to provide their professional appointment percentages amongst the following categories: academic (teaching), research, Extension (outreach), and administration. For the purposes of this study administrative appointment percentage served as a proxy for Katz’s ( 1955 ) role typologies. There were 11 participants that did not respond to the administrative appointment question, therefore these individuals were not included in subsequent analysis. Of the remaining 329 respondents, administrative appointments ranged from 0 to 100% with a mean of 41.4% ( SD  = 36.2). To establish administrative categories, z-scores were calculated where the score represented the number of standard deviations away from the observed mean score (Lewis-Beck et al., 2003 ). A negative z-score was associated with the low category (low) ( n  = 45), a z-score of zero was associated with the middle category (middle) ( n  = 189), and a positive z-score was associated with the top category (top) ( n  = 76). Additional details are presented in Table 1 .

Transformational leadership scores were calculated using the scoring key related to the TLI, which consists of 14 Likert-type items using a five point scale (1 =  strongly disagree , 2 =  disagree , 3 =  neither agree nor disagree , 4 =  agree , 5 =  strongly agree ). The scale had an observed Cronbach Alpha of 0.77 (Podsakoff et al., 1990 ). Statistical analysis was conducted using the Statistical Package for the Social Sciences (SPSS) version 25.

According to the first study objective, overall TLI descriptive statistics were calculated. The observed data resulted in a minimum transformational leadership score of 2.79 and a maximum score of 5.00 ( M  = 3.78, SD  = 0.38). Next, an ANOVA test between cohort groups was conducted, no statistically significant differences were observed when transformational leadership was analyzed by cohort group [ F (3, 294) = 1.22, p  = 0.30]. As no statistically significant relationships between classes were observed, data were considered to be statistically equivalent for subsequent analysis and considered as a single group to improve analytical power.

Next, to address the second research objective TLI scores were analyzed relative to administrative groupings. The low group had an observed mean score of 3.73 ( SD  = 0.35), the middle group had an observed mean score of 3.79 ( SD  = 0.40), and the top group had an observed mean score of 3.78 ( SD  = 0.33). An ANOVA test between administrative groups was conducted, no statistically significant differences were observed when transformational leadership was analyzed by administrative group [ F (2, 288) = 0.55, p  = 0.58].

To analyze research objective three, TLI scores were then analyzed based on self-reported gender. Overall females reported a higher self-reported level of transformational leadership ( M  = 3.82, SD  = 0.40) than did men ( M  = 3.76, SD  = 0.37). However, when the observed differences were analyzed using an ANOVA test, no statistically significant differences between groups were observed [ F (1, 295) = 1.74, p  = 0.19].

For the fourth and final research objective, TLI observations were compared between gender groups at each of the three administrative levels. At the low and middle administrative levels females reported a higher mean TLI score than did men; however, at the top administrative level men reported a higher mean TLI score. Additional details are presented in Table 2 . The three pairs of data were further analyzed using an ANOVA test. For the low administrative level, a statistically significant 0.25-point difference between gender mean TLI scores was observed [ F (1, 38) = 5.89, p  = 0.02]. At the middle administrative level, the 0.06-point difference was not found to be statistically significantly different [ F (1, 174) = 0.98, p  = 0.32]. Finally, at the top administrative level, the 0.06-point difference was not found to be statistically significantly different [ F (1, 72) = 0.54, p  = 0.47]. A graphical representation of the data is provided in Fig.  1 to visually represent the trends observed.

Gender, administrative appointment, & transformational leadership

Gender differences were the main findings among males and females who perceive themselves differently depending on the level of administrative responsibility they have achieved. The study contributes to literature on gender and leadership by showing that progressions of increased or decreased transformational leadership perception may occur as leaders are promoted. Findings not only align with past studies where females rated themselves higher than males on the use of transformational leadership (e.g. Burke & Collins, 2001 ; Carless, 1998 ), but the data also expands the discourse by illuminating how perceptions can shift over time based on administrative responsibility and leadership status. The acknowledgement of possible perception changes, based on time passed and position acquired, adds to the process of learning the idiosyncrasies of higher education leadership in general and gendered transformational leadership factors in particular.

Despite the novel nature of the findings, there are a number of limitations which should be acknowledged. First, though personal perceptions and self-reported data can be impactful in guiding researcher understanding of a phenomenon, it is important to note that it does not reflect actual leadership behavior and that responses may potentially be inflated by participants (Burke & Collins, 2001 ). Therefore, the use and interpretation of the data should be done with care. Specifically, self-reported transformational leadership data were used in the study. A recommendation would be to replicate the present study using a more objective measure of transformational leadership, such as external reviewer (subordinate, peer, supervisor) data. A second limitation is the context in which the study was conducted. As participants in a leadership development program, it is likely the individuals involved are not necessarily representative of all higher education faculty or administrators. The individuals participating in the program were generally identified based on their leadership potential. Therefore, a second recommendation would be replicate the study with a random sample of faculty within higher education more generally. Nevertheless, the current study is intended to provide a foundation upon which future research may benchmark and expand upon present findings.

Although participants of leadership development programs are generally already competent and successful in their respective job functions, leadership educators can help them move from mastering technical skills to developing human and conceptual competencies (Lamm et al., 2016 ). While the purpose of this study was not to assess whether LEAD21 helped participants develop human and conceptual skills, participants did perceive themselves to have these competencies in some capacity. In addition, participation in the program is based on nomination procedures, indicating that participants’ supervisors believed in the leadership capacity of the participant. Such information indicates that one’s perception of their own capacity to be a transformational leader can shift even if they, and others, initially believe their level of capacity is already high.

Results indicating that a decrease can occur in females’ confidence of their transformational leadership capacity are also interesting given that feminine leadership qualities (e.g. collaboration) have been associated with transformational leadership, leading scholars to posit that, “females and males may differ in their use of certain transformational leadership behaviors” (Carless, 1998 p. 890). A synthesis of research indicates the following female leadership attributes are more closely aligned with transformational leadership: a focus on interpersonal versus task success, empathy (rather than sole evaluation) while helping others, and group dynamics and harmony (as compared to solely reviewing individual performance) (Bass, 2008 ). Such sentiments referring to effective female leadership attributes have been noted in higher education-specific literature and dialogues (e.g. Nica, 2013 ). Transformational leadership has been “associated with a pattern of personality including high levels of pragmatism, nurturance, feminine attributes, and self-confidence, and low levels of criticalness and aggressiveness” (Bass, 1999 p. 28).

Previous research indicates women are perceived as more transformational leaders and men are perceived as more transactional leaders; even in women’s transactional leadership style, there is more compassion in situational and corrective circumstances (Bass, 2008 ). Bass ( 1999 ) points out that findings on gender differences could be attributed to the competencies females have to show more of to reach the same levels of leadership as men. For example, the glass ceiling concept, indicating barriers to advancement for females, does not only occur in corporate settings, but in higher education as well (Gunluk-Senesen, 2009 ). Bagilhole and White ( 2008 ) acknowledge that females remain excluded from top leadership positions at colleges and universities due to issues relating to “career mobility, experience outside academia, selection processes, and gender stereotyping” (abstract). This is a global phenomenon and higher education institutions are aware of such barriers, but have yet to fully address systematic structures and organizational cultures that continue this trend (Özkanlı et al., 2009 ). Although prior research has indicated females may receive more direct leadership development assistance than males (Burke & Collins, 2001 ), results from this study imply that higher education can do more (e.g. mentoring, promotion of educational associations, gender-specific leadership development initiatives) to provide specific support in helping females maintain their perception of their own transformational leadership effectiveness once they reach top administrative levels at their institution. With awareness and continued study, leadership development efforts can assist with this task. For example, in their review of leadership development literature, Dopson et al. ( 2019 ) point to the potential effectiveness of leadership development programming regarding promotion challenges faced by women by noting: “formalised leadership and skill‐based programmes may be more helpful in unblocking…unconscious gendered views rather than experiential methods which do not shift these gendered notions” (p. 223).

Factors such as socialized gender roles, gendered tasks, societal expectations, and challenges females face in leadership and promotion processes (Ayman & Korabik, 2010 ; Badura et al., 2018 ; Baker et al., 2019 ; Bass, 1999 ; Eagly & Karau, 1991 ; Judson et al., 2019 ; Nica, 2013 ) could possibly affect the decrease in female perceptions of their transformational leadership capacity, but more research should be done to gain a better understanding of the source of change and should include other factors such as organizational aspects. Also in future studies, inviting subordinate and/or superiors to rate their perception of leaders’ behavior could offset the biases and limitations self-reporting can create. Furthermore, the study at hand only looks at relationships between Katz’s ( 1955 ) administration levels (low, middle, and high) and the perception of transformational leadership skills; past studies review Katz’s ( 1955 ) technical, human, and conceptual skill levels as they relate to gender. Future research is recommended to study the perception of transformational leadership skills against Katz’s ( 1955 ) skill categories. Although the literature would indicate the human skill component of Katz’s ( 1955 ) theory best aligns with transformational leadership, future research could further examine the relationship combining the three administration levels, the three skills, transformational leadership, and/or gender factors into a single study. Also, due to there being a dearth of studies connecting Katz’s ( 1955 ) theory and gender future research may provide additional societal and cultural context (as they relate to gender) to complement the findings in the study at hand, particularly as it relates to leadership and administration within higher education around the globe. Lastly, few studies explore the long-term impact and outcomes of higher education leadership development programs (Dopson et al., 2019 ). Therefore, specific leadership development factors possibly contributing to any observed changes is a worthwhile response to a call in the literature for more outcome- and longitudinal-based empirical research (Dopson et al., 2019 ).

Contributions to the literature

The results of the study and subsequent discussion provide an overview of the study and the relationship to the previous research. Nevertheless, from a tactical perspective, the current study provides a series of contributions to the higher education leadership literature. First, the study provides a framework for considering different higher education role types based on administrative appointment.

Second, from an application perspective the preliminary results, before applying the role level and gender variables, are representatives of the risk associated with aggregating groups without appreciating the nuance and unique characteristics associated with different role types and individuals within the roles. Although the transformational leadership results provide a baseline and set of average scores among a sample of higher education leaders, the utility and practical value of the results is somewhat limited. Perhaps the use of better attuned models appears warranted.

Third, the results of the study indicate the role of gender within higher education leadership may not be limited to simply demographic identification. Instead, the results imply the need to consider interaction effects between demographics variables, such as gender, and less proximal role related variables, such as level of leadership role. These contributions specifically address the need for systemic analysis of the role of gender within higher education leadership.

Contributions to practice

The results also contribute to practice and discourse related to higher education leadership. Specifically, the results of the transformational leadership analysis are simultaneously expected and unexpected from a gender and role perspective. From a gender point of view the higher mean scores associated with individuals who self-identified as female is consistent with expectations. Similarly, the increase in mean scores moving from the low administrative group to the middle and top groups is also expected. However, the divergence in observations when overlaying gender and role administrative group is where specific implications for practice may emerge. Specifically, within the low administrative group, self-reported females had statistically significantly higher levels of transformational leadership than their male counterparts.

The results may indicate that initial leadership emergence identification may be associated with individuals fulfilling expected gender roles. However, when transitioning to higher levels of administrative responsibility, the gap between males and females narrows considerably at the middle levels of administrative responsibility as males take on the characteristics of transformational leaders. Concurrently, females appear to adapt their transformational leadership style to match expectations of top-level administrators by modulating aspects of transformational leadership.

Based on these findings, the potential for gender role expectations to influence perceptions of leadership capacity should be noted during the initial stages of leadership identification. Previous researchers have identified that, “women and men who are effective leaders are expected to demonstrate different behaviors and leadership styles” (Dunn et al., 2014 , p. 10). Therefore, leadership potential across a range of criteria, both observed and potentially developed, should be considered to ensure individuals are provided opportunities to develop accordingly.

Second, throughout the leadership development process, it is important to acknowledge the role of context and to provide a variety of examples of successful leadership from both males and females, with varying levels of transformational leadership. As previous scholars have found, “The underrepresentation of women in academic administration suggests that masculine practices and leadership norms function to exclude women” (Dunn et al., 2014 , p. 9). Finding opportunities and exemplars of success across genders and leadership styles has the potential to inspire more individuals, of both genders, to pursue higher education leadership roles.

The development of leadership skills continues to be one of the most important investments an organization, including higher education, can make in its employees (Badura et al., 2019 ). Transformational leadership, in particular, has been shown to be an effective and necessary response to meet present and future challenges (Kezar et al., 2019 ). Therefore, studying the transformational capacity of leaders at varying levels of leadership in higher education domains is a worthwhile venture. The study at hand sought to explore transformational leadership as it relates to administrative roles, possibly being the first to do so and to also connect gender to both topics in the same study. The TLI, an instrument not typically applied to skills-based studies, was used and the use of four LEAD21 leadership development classes gives the study the benefit of comparing multi-year information as well as improved statistical power within which to analyze trends. Also, the study itself provides updated information that can influence the direction of future research surrounding leadership in general and gendered transformational leadership in particular within higher education contexts. Leaders must be educated on how to lead effectively within the sphere of higher education (Dopson et al., 2019 ; Pearce et al., 2018 ); findings from this study can be included in such critical development efforts.

Data availability

The datasets analyzed during the current study are available from the corresponding author on reasonable request pending confidentiality requirements associated with Institutional Review Board approval of the research.

Code availability

Not applicable.

Anim, M. N., & Shotte, G. (2020). Leadership roles, women and higher education: Lessons from the University of Buea, Cameroon. Bulgarian Comparative Education Society, 18 , 148–155.

Google Scholar  

Anthony, S. D. & Schwartz, E. I. (2017). What the best transformational leaders do . Harvard Business Review. Retrieved May 8, 2017, from https://hbr.org/2017/05/what-the-best-transformational-leaders-do

Argia, H. A. A., & Ismail, A. (2013). The influence of transformational leadership on the level of TQM implementation in the higher education sector. Higher Education Studies, 3 (1), 136–146. https://doi.org/10.5539/hes.v3n1p136

Article   Google Scholar  

Arokiasamy, A. R. A., & Tat, H. H. (2020). Exploring the influence of transformational leadership on work engagement and workplace spirituality of academic employees in the private higher education institutions in Malaysia. Management Science Letters, 10 (4), 855–864. https://doi.org/10.5267/j.msl.2019.10.011

Association of Public and Land-Grant Universities. (2016). Land-Grant University FAQ . Retrieved April 22, 2016, from https://www.aplu.org/about-us/history-of-aplu/what-is-a-land-grant-university/#:~:text=A%20land%2Dgrant%20college%20or,1862%2C%201890%2C%20and%201994.&text=The%20first%20Morrill%20Act%20provided,federal%20lands%20to%20each%20state

Ayman, R., & Korabik, K. (2010). Leadership: Why gender and culture matter. American Psychologist, 65 (3), 157–170. https://doi.org/10.1037/a0018806

Badura, K. L., Grijalva, E., Galvin, B. M., Owens, B. P., & Joseph, D. L. (2019). Motivation to lead: A meta-analysis and distal-proximal model of motivation and leadership. Journal of Applied Psychology . Advance online publication. https://doi.org/10.1037/apl0000439

Badura, K. L., Grijalva, E., Newman, D. A., Yan, T. T., & Jeon, G. (2018). Gender and leadership emergence: A meta-analysis and explanatory model. Personnel Psychology, 71 (3), 335–367. https://doi.org/10.1111/peps.12266

Bagilhole, B., & White, K. (2008). Towards a gendered skills analysis of senior management positions in UK and Australian universities. Tertiary Education Management, 14 , 1–12. https://doi.org/10.1080/13583880701814124

Baker, V. L., Lunsford, L. G., & Pifer, M. J. (2019). Patching up the “leaking leadership pipeline”: Fostering mid-career faculty succession management. Research in Higher Education, 60 (6), 823–843. https://doi.org/10.1007/s11162-018-9528-9

Bass, B. M. (1999). Two decades of research and development in transformational leadership. European Journal of Work and Organizational Psychology, 8 (1), 9–32. https://doi.org/10.1080/135943299398410

Bass, B. M. (2008). The Bass handbook of leadership: Theory, research, and managerial applications (4th ed.). Free Press.

Bass, B. M., & Avolio, B. J. (1993). Transformational leadership: A response to critiques. In M. M. Chemers & R. Ayman (Eds.), Leadership theory and research: Perspectives and directions (pp. 49–80). Academic Press.

Bass, B. M., & Riggio, R. E. (2006). Transformational leadership . Psychology Press.

Book   Google Scholar  

Benson, A., Li, D., & Shue, K. (2018). Promotions and the peter principle . The National Bureau of Economic Research. http://www.nber.org/papers/w24343

Berson, Y., Waldman, D. A., & Pearce, C. L. (2016). Enhancing our understanding of vision in organizations: Toward an integration of leader and follower processes. Organizational Psychology Review, 6 (2), 171–191. https://doi.org/10.1177/2041386615583736

Bono, J. E., & Judge, T. A. (2004). Personality and transformational and transactional leadership: A meta-analysis. The Journal of Applied Psychology . https://doi.org/10.1037/0021-9010.89.5.901

Brown, M., Brown, R. S., & Nandedkar, A. (2019). Transformational leadership theory and exploring the perceptions of diversity management in higher education. Journal of Higher Education Theory & Practice, 19 (7), 11–21. https://doi.org/10.33423/jhetp.v19i7.2527

Burke, S., & Collins, K. M. (2001). Gender differences in leadership styles and management skills. Women in Management Review, 16 (5), 244–257. https://doi.org/10.1108/09649420110395728

Carless, S. A. (1998). Gender differences in transformational leadership: An examination of superior, leader, and subordinate perspectives. Sex Roles, 39 (11–12), 887. https://doi.org/10.1023/A:1018880706172

Dopson, S., Ferlie, E., McGivern, G., Fischer, M. D., Mitra, M., Ledger, J., & Behrens, S. (2019). Leadership development in higher education: A literature review and implications for programme redesign. Higher Education Quarterly, 73 (2), 218–234. https://doi.org/10.1111/hequ.12194

Dunn, D., Gerlach, J. M., & Hyle, A. E. (2014). Gender and leadership: Reflections of women in higher education administration.  International Journal of Leadership and Change ,  2 (1), 2, 9–18. http://digitalcommons.wku.edu/ijlc/vol2/iss1/2

Eagly, A. H., & Karau, S. J. (1991). Gender and the emergence of leaders: A meta-analysis. Journal of Personality and Social Psychology, 60 , 685–710. https://doi.org/10.1037/0022-3514.60.5.685

Fernet, C., Trépanier, S.-G., Austin, S., Gagné, M., & Forest, J. (2015). Transformational leadership and optimal functioning at work: On the mediating role of employees’ perceived job characteristics and motivation. Work & Stress, 29 (1), 11–31. https://doi.org/10.1080/02678373.2014.1003998

Gigliotti, R. A. (2017). An exploratory study of academic leadership education within the association of American Universities. Journal of Applied Research in Higher Education, 9 (2), 196–210. https://doi.org/10.1108/JARHE-11-2015-0080

Gunluk-Senesen, G. (2009). Glass ceiling in academic administration in Turkey: 1990s versus 2000s. Tertiary Education Management, 15 , 305–322. https://doi.org/10.1080/13583880903335480

Hambali, M., & Idris, I. (2020). Transformational leadership, organizational culture, quality assurance, and organizational performance: Case study in Islamic higher education institutions (IHEIS). Jurnal Aplikasi Manajemen, 18 (3), 572–587. https://doi.org/10.21776/ub.jam.2020.018.03.18

Hung Vu, T., Thu Vu, M., & Hoang, V. N. (2019). The impact of transformational leadership on promoting academic research in higher educational system in Vietnam. Management Science Letters, 10 (3), 585–592. https://doi.org/10.5267/j.msl.2019.9.022

Judson, E., Ross, L., & Glassmeyer, K. (2019). How research, teaching, and leadership roles are recommended to male and female engineering faculty differently. Research in Higher Education, 60 (7), 1025–1047. https://doi.org/10.1007/s11162-018-09542-8

Kaifi, B. A., & Mujtaba, B. G. (2010). A study of management skills with Indian respondents: Comparing their technical, human and conceptual scores based on gender. Journal of Applied Business & Economics, 11 (2), 129–138.

Katz, R. L. (1955). Skills of an effective administrator. Harvard Business Review, 33 (1), 33–42.

Kearns, K. P., Livingston, J., Scherer, S., & McShane, L. (2015). Leadership skills as construed by nonprofit chief executives. Leadership & Organization Development Journal, 36 (6), 712–727. https://doi.org/10.1108/LODJ-11-2013-0143

Kezar, A., Dizon, J. P. M., & Scott, D. (2019). Senior leadership teams in higher education: What we know and what we need to know. Innovative Higher Education, 45 , 103–120. https://doi.org/10.1007/s10755-019-09491-9

Lamm, K. W., Sapp, L. R., & Lamm, A. J. (2016). Leadership programming: Exploring a path to faculty engagement in transformational leadership. Journal of Agricultural Education, 57 (1), 106–120. https://doi.org/10.5032/jae.2016.01106

LEAD21. (n.d.). Program overview . http://lead21.org/program-overview/

Lewis-Beck, M., Bryman, A. E., & Liao, T. F. (2003). The Sage encyclopedia of social science research methods . Sage Publications.

Limsila, K. & Ogunlana, S. O. (2008). Linking personal competencies with transformational leadership style evidence from the construction industry in Thailand. Journal of Construction in Developing Countries, 13( 1), 27–50. http://web.usm.my/jcdc/vol13_1_2008/2_Kedsuda%20Limsila%20(p.%2027-50).pdf

Madden, M. (2011). Gender stereotypes of leaders: Do they influence leadership in higher education? Wagadu: A Journal of Transnational Women’s and Gender Studies, 9 , 55–88.

Maimon, E. P. (2018). A Checklist for Transformative Leaders . The Chronicle of Higher Education. Retrieved January 7, 2018, from https://www.chronicle.com/article/A-Checklist-for-Transformative/242167

Major, C. (2020). Innovations in teaching and learning during a time of crisis. Innovative Higher Education, 45 (4), 265–266. https://doi.org/10.1007/s10755-020-09514-w

Megheirkouni, M., Amaugo, A., & Jallo, S. (2018). Transformational and transactional leadership and skills approach: Insights on stadium management. International Journal of Public Leadership, 14 (4), 245–259. https://doi.org/10.1108/IJPL-06-2018-0029

Mujtaba, B., & Kaifi, B. (2011). Management skills of Afghan respondents: A comparison of technical, human and conceptual differences based on gender. Journal of International Business and Cultural Studies, 4 (1), 1–14.

Muldrow, M. (2014). Online learning and post-secondary expectations: Bridging the gap from academic leaders to instructors. Academic Leadership Journal in Student Research, 2 , 1–12.

Nica, E. (2013). The importance of leadership development within higher education. Contemporary Readings in Law & Social Justice, 5 (2), 189–194.

Northouse, P. G. (2010). Leadership: Theory and practice (5th ed.). Sage Publications.

Northouse, P. G. (2013). Leadership: Theory and practice (6th ed.). Sage Publications.

Özkanlı, Ö., de Lourdes Machado, M., White, K., O’Connor, P., Riordan, S., & Neale, J. (2009). Gender and management in HEIs: Changing organisational and management structures. Tertiary Education Management, 15 , 241–257. https://doi.org/10.1080/13583880903073008

Pearce, C. L., Wood, B. G., & Wassenaar, C. L. (2018). The future of leadership in public universities: Is shared leadership the answer? Public Administration Review, 78 (4), 640–644. https://doi.org/10.1111/puar.12938

Peterson, T., & Van Fleet, D. (2004). The ongoing legacy of R.L. Katz. Management Decision, 42 (10), 1297–1308. https://doi.org/10.1108/00251740410568980

Petkevičiūtė, N., & Giedraitis, A. (2013). Leadership skills formation in workgroup of the first level managers in manufacturing companies. Management of Organizations: Systematic Research, 67 , 69–82. https://doi.org/10.7720/MOSR.1392-1142.2013.67.5

Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. The Leadership Quarterly, 1 (2), 107–142. https://doi.org/10.1016/1048-9843(90)90009-7

Society for Human Resource Management. (2015). SHRM research overview: Leadership development.  https://www.shrm.org/hr-today/trends-and-forecasting/special-reports-and-expert-views/Documents/17-0396%20Research%20Overview%20Leadership%20Development%20FNL.pdf

Thomas, N., Bystydzienski, J., & Desai, A. (2015). Changing institutional culture through peer mentoring of women STEM faculty. Innovative Higher Education, 40 , 143–157. https://doi.org/10.1007/s10755-014-9300-9

Thompson, J. (2012). Transformational leadership can improve workforce competencies. Nursing Management, 18 (10), 21–24.

Turnbull, S., & Edwards, G. (2005). Leadership development for organizational change in a new UK university. Advances in Developing Human Resources, 7 (3), 396–413. https://doi.org/10.1177/1523422305277178

U.S. Bureau of Labor Statistics. (2016). 39 percent of managers in 2015 were women . Retrieved August 1, 2016, from https://www.bls.gov/opub/ted/2016/39-percent-of-managers-in-2015-were-women.htm

Vieira da Motta, M., & Bolan, V. (2008). Academic and managerial skills of academic deans: A self-assessment perspective. Tertiary Education Management, 14 , 303–316. https://doi.org/10.1080/13583880802481740

Download references

This work was supported by the LEAD21 organization.

Author information

Authors and affiliations.

Department of Agricultural Leadership, Education and Communication, University of Georgia, 141 Four Towers Building, Athens, GA, 30602, USA

Kevan W. Lamm & Alexa J. Lamm

Director LEAD21 and University of Georgia Extension, University of Georgia, Athens, USA

L. Rochelle Sapp

Department of Lifelong Education, Administration, and Policy, University of Georgia, Athens, USA

Nekeisha L. Randall

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Kevan W. Lamm .

Ethics declarations

Conflict of interest.

This research is sponsored by the LEAD21 organization in which we have a professional relationship. We have disclosed those interests fully to Springer, and have in place an approved plan for managing any potential conflicts arising from this arrangement.

Additional information

Publisher's note.

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

Rights and permissions

Reprints and permissions

About this article

Lamm, K.W., Sapp, L.R., Randall, N.L. et al. Leadership development programming in higher education: an exploration of perceptions of transformational leadership across gender and role types. Tert Educ Manag 27 , 297–312 (2021). https://doi.org/10.1007/s11233-021-09076-2

Download citation

Received : 17 July 2020

Accepted : 24 September 2021

Published : 11 October 2021

Issue Date : December 2021

DOI : https://doi.org/10.1007/s11233-021-09076-2

Share this article

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

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

Provided by the Springer Nature SharedIt content-sharing initiative

• Higher education
• Skills approach
• Find a journal
• Publish with us
• Track your research