Promoting Interdisciplinary Research Collaboration: A Systematic Review, a Critical Literature Review, and a Pathway Forward

ABSTRACT

Interdisciplinary research has been a topic of interest for many decades – perhaps longer. And yet, even now, there is still much we do not understand about how to stimulate collaboration across research disciplines. This article reports the results of a systematic review of the academic literature on strategies for promoting new interdisciplinary research collaborations, which returned only a very small number of empirical studies. A broader review of the scholarship in this area reveals a literature that is highly theorized, but not adequately operationalized for empirical investigation of how to enable new collaborations. A shift toward a better formulation of research questions, with a view toward generating more empirical data on practical strategies for connecting researchers and encouraging them to work together across disciplinary boundaries, will be required to further the research agenda on interdisciplinary collaboration.

KEYWORDS:

• Interdisciplinary research
• multidisciplinary research
• transdisciplinary research
• research collaboration

Introduction

Interdisciplinary research has been a topic of interest among scientists, scholars, university administrators, and policy makers for many decades (Klein 1990). According to Lattuca (2001, 4–10), the question of how to join different strands of knowledge within a learning environment is as old as academic scholarship itself. Interest in interdisciplinarity stems from a belief in its various perceived benefits: many observers (e.g. Lattuca 2001, 51; Szostak, Gnoli, and Lopez-Huertas 2016, 13;Winowiecki et al. 2011, 78–79) claim that interdisciplinary collaboration, by its very nature, promotes creativity, innovation, and outside-the-box thinking. Moreover, there is increasing consensus that real-world policy problems are inherently interdisciplinary and cannot be addressed with knowledge from only a single scientific or academic discipline (e.g. Aboelela et al. 2007, 330; Bark, Kragt, and Robson 2016, 1449; Siedlok and Hibbert 2014, 194). And yet, even now, in an era of advanced communications technology, instantaneous access to information, and increasingly insistent demands for academics to demonstrate research impact, and despite continued research efforts in this area, there is still much we do not understand about how to stimulate collaboration across research disciplines.

In an effort to synthesize understanding on strategies for promoting new interdisciplinary research collaborations, I conducted a systematic review of the academic literature, which I present below. However, the review yielded only a very small number of studies, and even within this small number, differences in methodology and weaknesses in data quality make it difficult to form any unified conclusions.

If interdisciplinary research is in demand, why is there so little empirical scholarship on how to get researchers to collaborate across disciplines? A broader, more critical review of the literature reveals that research on interdisciplinarity has largely not been operationalized in a way that would help address empirical questions related to forming new interdisciplinary teams. Research objectives and agendas are unclear; definitions of key terms and concepts are absent or are not widely accepted; general rules are frequently derived from unique case studies; and untested hypotheses, in some cases bordering on mythology, are often accepted as fact and left unquestioned.

There are many aspects of interdisciplinary research and interdisciplinarity more generally that have been studied widely, but the question of how to motivate researchers to conduct interdisciplinary research in the first place is of primary importance, and it is a question that has not received adequate attention. As will be demonstrated below, a shift in priorities toward empirical investigation, practical strategies, and above all, taking bold risks to trial new methods, is required if we are to learn how to bring together academics from different disciplines and empower them to collaborate on research projects that combine diverse bases of expertise.

A Systematic Review of the Literature

I conducted a systematic review of the academic literature on strategies for enabling discipline-based researchers to initiate new collaborations with one another across their disciplinary boundaries. The review was conducted in July 2021. Peer reviewed journal articles, academic books, and academic book chapters, written in English, either fully published or published online in early view, that present empirical research results on enabling researchers to come together across disciplines for interdisciplinary research collaboration, were included in the review. Commentary, theoretical or conceptual development, papers with no data (including those that outline a method but do not present or analyze any data), and items based on personal reflection were excluded. Works that report on how to enable ‘successful’ interdisciplinary collaboration by examining existing interdisciplinary teams of researchers (i.e. how to maintain interdisciplinary teams, rather than how to encourage the formation of interdisciplinary teams), and papers that focus on interdisciplinary training of undergraduate or postgraduate students, even if that includes research, were likewise excluded. The base year was 2000, since work conducted prior to that point would not be able to take into consideration changes in technology associated with the information revolution that have radically altered the capacity for people to interact and share knowledge and ideas.

I searched Google Scholar for keywords such as ‘interdisciplinary’, ‘multidisciplinary’, ‘transdisciplinary’, crossdisciplinary“, ‘research team/teams’, ‘research partnership/partnerships’, and ‘research collaboration/collaborations’. My search strategy used cascading configurations of keywords so that, for example, interdisciplinary research collaboration (with no quotation marks) was followed by multidisciplinary research collaboration -interdisciplinary (also with no quotation marks) in order to narrow down the search results and avoid duplication. It should also be noted that Google Scholar’s search function returns results with and without hyphens (for example, papers with the keywords ‘multi-disciplinary’ and ‘multidisciplinary’ are all returned in a single search for ‘multidisciplinary’).

In total, 25 searches were conducted. For each search, I scanned titles and abstracts for papers that met the inclusion criteria for the review, with the search continuing until 10 pages (i.e. 100 items) had passed with no items meeting the criteria. Items that met the inclusion criteria on the basis of title and abstract were downloaded for closer reading. After scanning a total of 3140 items, 48 were downloaded for close reading. Of these 48, 34 still did not meet the inclusion criteria for the review, leaving a final total of 14 items for inclusion.

The 14 studies included in the systematic review represent mainly a Western, Anglo-centric perspective. 7 studies were conducted in the United States, 2 in the United Kingdom, 1 in Sweden, 1 across the European Union, 1 in New Zealand, 1 in Australia, and 1 that straddled Australia and the United Kingdom. A variety of research methods were used across the included studies, such as surveys, interviews, focus groups, website user data, and network analysis. Nine studies describe an intervention, applied either by the researchers themselves or by the institutions they observed, and 5 studies had no intervention, relying instead on the general perceptions of survey or interview respondents. Two of the studies that described an intervention did not actually use the intervention in the study’s analysis and are therefore more similar to the studies that used cross-sectional survey or interview data. Included studies and their characteristics are noted in Appendix 1.

The 7 studies analyzing the results of an intervention all reported an increase in interdisciplinary collaboration. Three studies found that internal grant funding schemes were successful in increasing interdisciplinary collaboration at the institutions where they were trialled. One study reported increased interdisciplinary contact among researchers who had attended a workshop specifically organized to connect researchers from different academic backgrounds. One study presented the results of a trial of a website application designed to create an online research community, and reported that the website increased interdisciplinary connections. Two studies reviewed large, well-funded, independent interdisciplinary research institutes and found that these institutes improved interdisciplinary collaboration. However, while overall these studies suggest that organized activity can improve interdisciplinary interaction, the small number of studies and the variety of interventions described make it difficult to draw more specific conclusions.

Eleven studies used surveys, interviews, and focus groups to collect data, and some points repeat across more than one of them. For instance, in 5 studies, respondents remarked that career incentives, such as hiring, promotion, and tenure processes, discourage interdisciplinary research. In four studies, respondents claimed that physical distance between researchers was a barrier to collaboration. Three studies presented respondents who found it hard to obtain grant funding for interdisciplinary projects. In another three studies, respondents said that the increased bureaucracy faced by researchers trying to collaborate across internal units or departments was a barrier to interdisciplinary research.

Unfortunately, the results of these studies do not converge on any unifying message, and again, compelling conclusions are elusive. For example, even among the 5 studies that discussed career incentives, in two of those studies it was a minority of respondents who found this to be a problem for interdisciplinary collaboration. In one of the studies in which respondents stated that it was harder to obtain grant funding for interdisciplinary projects than for discipline-based projects, only 60% of respondents agreed with this sentiment, meaning that a sizeable minority of 40% did not agree. These disagreements are compounded further by the fact that numerous points appear only once or twice across the 11 studies, suggesting considerable divergence among the studies’ conclusions. A list of these points is given in Table 1.

Table 1. Barriers to interdisciplinary research collaboration, as reported by studies based on surveys and interviews.

Barriers to Interdisciplinary Research Collaboration	Number of Studies Reporting
Career incentives (e.g. hiring, promotion, and tenure processes) discourage interdisciplinary collaboration	5
Physical distance between researchers discourages collaboration	4
Grant funding for interdisciplinary projects is more difficult to obtain than for projects based in a single discipline	3
Working across departments creates excess bureaucracy	3
Difficulty demonstrating one’s contribution to an interdisciplinary project is a disincentive to collaboration	2
Interdisciplinary research takes more time than discipline-based research	2
Communicating across disciplines is difficult	2
Interdisciplinary research is considered lower quality by peers	2
Interdisciplinary research is harder to publish than discipline-based research	1
There is a lack of dedicated funding for interdisciplinary research	1

A further problem is that many of these studies present serious methodological weaknesses. For instance, of the 11 studies that report the results of surveys or interviews, 7 rely on very small numbers of respondents. Boden and Borrego (2011), for example, report the results of interviews with 9 researchers; Kandiko and Blackmore (2008) report results of 10 interviews; Lau and Pasquini (2004) report results from 14 interviews; and Rekers and Hansen (2015) report results from 15 interviews. Lin (2008) relies on data from a survey of 26 researchers, all based in one research area at a single university.

Other methodological issues are also evident. For example, Adams et al. (2008) describe an internal grant funding scheme at an American university and, after performing a network analysis comparing co-authored publications that were produced before and after the funding scheme was initiated, conclude that the program helped increase interdisciplinary collaboration. However, there is no discussion of causality – that is, evidence that the intervention actually contributed to the observed increase in collaboration – anywhere in the study. In two studies that describe an intervention, there was no direct follow-up with the intervention group and so it is impossible to know if the intervention promoted interdisciplinary collaboration at all.

In sum, the most conspicuous lesson revealed by this systematic review is that there is not very much empirical scholarship on strategies to encourage researchers to collaborate across disciplinary boundaries. The studies that do exist use a variety of methodologies and present varied and somewhat contradictory findings. Many of these studies have significant methodological weaknesses that raise questions about the generalizability of their conclusions. It may, therefore, not be possible to use these studies to derive any kind of combined message on promoting interdisciplinary research at all.

Critically Assessing the Broader Literature

Despite the small number of empirical studies that investigate how to promote new interdisciplinary collaborations, the broader literature on interdisciplinary research is voluminous. But if so little attention is being paid to strategies to bring researchers together, what are scholars focused on instead? In what follows, I present a critical review of the wider literature on interdisciplinary research, beyond the small number of items resulting from the systematic search discussed above. This wider review is not meant to be systematic or comprehensive, but instead to address what I see as conceptual and methodological gaps that can be identified from more extensive reading in this area.

Scholarship on interdisciplinarity can generally be divided into five themes: 1. Defining Interdisciplinarity; 2. Working Across Disciplines; 3. Unique Case Studies; 4. Measuring Interdisciplinarity; and 5. Barriers to (and Enablers of) Interdisciplinary Collaboration. At present, these themes represent quite diverse areas of investigation. However, as will be discussed below, each of these themes presents significant deficiencies when it comes to contributing to a discourse on promoting new interdisciplinary collaborations.

Theme 1: Defining Interdisciplinarity

A substantial body of work deals with the construct of disciplines of knowledge and the meaning of the word ‘interdisciplinary’. Lattuca (2001), for instance, in a landmark book on the subject, allocates three of her eight chapters to the history of academic disciplines, defining interdisciplinarity, and parsing out different forms of interdisciplinary research. Other works (e.g. Aram 2004; Lau and Pasquini 2004) dig deeper in an effort to produce ever more precise definitions. Several handbooks and textbooks have also been published, such as the Handbook of Transdisciplinary Research (Hadorn et al. 2008) and The Oxford Handbook of Interdisciplinarity (Frodeman 2017), each with significant space devoted to defining the concept of interdisciplinarity and categorizing its various forms. A report by the National Academies from 2005 provides probably the most thorough treatment of the concept of interdisciplinarity, and remains a standard reference on the subject (Institute of Medicine 2005).

Many authors take this task a step further by proposing typologies of different modes of interdisciplinarity. There are numerous versions (e.g. Lattuca 2001, 114; Klein 2017), but by far the most prevalent is a typology that distinguishes between ‘multi-disciplinary’, in which knowledge from commonly recognized disciplinary bases are used in parallel, ‘interdisciplinary’, in which multiple knowledge bases are used together to inform new research, and ‘transdisciplinary’, in which knowledge production transcends traditional disciplines altogether, in effect creating a completely novel brand of scientific inquiry. Most authors use this typology without attribution (e.g. Bark, Kragt, and Robson 2016, 1450), but Austin, Park, and Goble (2008, 557) trace its origin back to Nissani (1995).

A great many books, chapters, and articles – far too many to cite here – include the multi-/inter-/trans- typology in their definition of interdisciplinarity. However, the highly subjective nature of these terms makes it difficult to see the difference between the categories (Bammer 2016, 2), with the result that they are often used interchangeably. Furthermore, not all definitions are universally agreed upon: some authors (e.g. Frodeman 2011; Neuhauser 2018) use the term ‘transdisciplinary’ to refer specifically to research that directly engages with stakeholders or end-users, resulting in further definitional confusion. Pedersen (2016), for example, uses ‘transdisciplinary’ to mean both transcending disciplines and engagement with stakeholders at the same time, a usage that does not align with other definitions of the word. Admittedly, this is part of an ongoing discourse in which attempts have been made to systematize the terms of the debate (e.g. Klein 2017; Osborne 2015), and clarification may yet be achieved. But for now, there is still evident disagreement among scholars with respect to some of the vocabulary used for fundamental concepts.

Most importantly, despite the effort to achieve clarity, there is a noticeable confusion in the literature about whether interdisciplinary research is necessarily a team effort between individuals with different specializations, or whether a research project can be interdisciplinary on its own, regardless of the number of investigators and irrespective of their backgrounds. Like Aboelela et al. (2007), many authors assume from the start that interdisciplinarity requires teams of researchers from different disciplines (e.g. Beers and Bots 2009; Bellanca 2009; Pedersen 2016). On the other hand, textbooks on how to conduct interdisciplinary research, such as Menken and Keestra’s An Introduction to Interdisciplinary Research: Theory and Practice (2016), provide advice on how to ‘do’ interdisciplinary research, including developing an interdisciplinary research question, integrating different research perspectives, and mixed-methods approaches to collecting and analyzing data – all of which can be equally applied to individuals or teams. Some scholarship is specifically aimed at cultivating ‘interdisciplinary’ researchers and is only meaningful when talking about individuals (e.g. Millar 2013), whereas other work mixes analysis of interdisciplinary research by individuals and collaboration among teams of researchers without making any distinction between them (e.g. McBee and Leahey 2016).

Ultimately, while a lot of work has been done to define interdisciplinarity, the focus has largely been on slicing up definitions in ways that many authors find confusing, and which has perhaps resulted in less clarity and consensus than would be desired. Meanwhile, much less attention has been paid to the distinction between interdisciplinary research and interdisciplinary teams, which has considerable practical application: for instance, when investigating the employability of recent individual PhD graduates whose thesis research straddles multiple disciplines (Millar 2013), the term ‘interdisciplinary’ must be understood as applying to the research, not to the people. For a research question about antagonism between science researchers and arts researchers, and the difficulties in their collaborating on a single research project (Gardner 2013), the notion of interdisciplinarity must be applied to the teams. For understanding how to initiate and support new interdisciplinary collaborations, this is the definitional precision that is more often required, but is not spelled out often enough.

Therefore, when thinking about how to promote new interdisciplinary research collaborations, it may be more useful to imagine the concept of interdisciplinarity as operating along two dimensions. First, research projects can be less or more interdisciplinary, depending on the degree to which they bring together knowledge and scholarship from different academic disciplines. Second, research teams can be less or more interdisciplinary, depending on how ‘distal’ are the backgrounds of the team members (Gibson et al. 2019). These dimensions are illustrated in Figure 1:

Figure 1. Two dimensions of interdisciplinarity.

In the top left quadrant, one or more researchers all based in the same research discipline work on a project that is also sited comfortably within a single traditional discipline, for example mechanical engineers working on the design of an airfoil to reduce aerodynamic drag, or a political scientist studying voter turnout in first-past-the-post electoral systems. In the lower left quadrant, one or more researchers all based in the same research discipline undertake a project that crosses knowledge bases, such as a project on the philosophy of memory in which all investigators are psychologists. In the top right quadrant, which I have termed, ‘Integrated Collaboration’, a research team may include members from different disciplines working on essentially a single-discipline project, such as when a computer scientist might collaborate with a team of sociologists on the effects of artificial intelligence on the use of social media, which is essentially a sociology research project (notwithstanding the participation of the computer scientist). And finally, in the lower right quadrant, team members of various backgrounds unite for a project that undeniably crosses traditional disciplinary boundaries, for example when sports scientists work with geneticists to examine the effects of exercise on the process of aging, which is a project that is neither solely a sports science project nor solely a genetics project – it uses knowledge from both disciplines and involves the participation of researchers from both disciplinary backgrounds. This last example would constitute truly interdisciplinary collaboration.

It is important to note that the quadrants in Figure 1 are not meant as discrete categories, but more to illustrate a spectrum operating in two dimensions. Projects and teams can be more or less interdisciplinary by degrees. While this is obviously hard to quantify, one might agree that sociology and criminology are ‘closer’ disciplines than are sociology and chemical engineering. Abramo, D’angelo, and Zhang (2018, 1184) have proposed using variety (how many different disciplines are included), balance (how much each discipline is represented relative to the others), and disparity (how different are the disciplines from each other) to further position different interdisciplinary efforts along this scale.

Theme 2: Working Across Disciplines

A significant amount of scholarship on interdisciplinarity concerns the capacity of researchers to learn and to share knowledge across disciplines. The most common sentiment by far is that communication and collaboration across disciplinary boundaries is difficult, and can only be accomplished by working through a number of inherent challenges.

Many authors take the position that communication is the biggest challenge for interdisciplinary collaboration. Academics have highly specialized professional training, and each disciplinary tradition comes with its own language, fundamental concepts, and common terminology (Austin, Park, and Goble 2008, 562–563; Lattuca 2001, 29–31). Some observers assert that academic disciplines have particular methodologies (Lach 2014), or varying standards of rigour related to methodologies (Brister 2016, 87), which may present significant barriers to collaboration. Many authors argue that different disciplines are linked to fundamentally differing epistemologies – in other words, their entire basis of acquiring knowledge and understanding is unique (Jacobs and Frickel 2009, 47–48;Pedersen 2016). Severe epistemological differences would make sharing knowledge exceedingly difficult.

But even if shared understanding of research were possible, researchers working in different academic disciplines might have a hard time agreeing on common research objectives that would motivate an interdisciplinary research project. Robotics engineers working on an electromechanical prosthetic limb, for example, might be focused on a technical aspect of the device, such as range of motion, whereas an occupational therapist might have a more patient-oriented concern, such as how the device allows the user to carry out activities that they were not able to do previously. An interdisciplinary research project with team members from different disciplines would need to find a way to integrate the interests of the various researchers. Brister (2016) offers a vivid example of how discipline-based research objectives might even seriously conflict with one another: conservation biologists might conduct a research project and conclude that an exclusionary zone (a protected area in which people are not allowed to enter) is necessary to protect a species from extinction. Social scientists working in the same location might conclude that an indigenous group’s traditional practices must be preserved to maintain the economic viability of the community – which an exclusionary zone would disrupt significantly. In this case, different research objectives, as well as diverging values, would present a considerable obstacle to interdisciplinary research collaboration.

From here, it is easy to conclude – as many do – that various academic disciplines might engage in forms of tribalism. Lattuca (2001, 36), for example, argues that disciplines give researchers a sense of identity, membership in a community, and a feeling of security that comes with operating within a defined territory. Members would therefore be expected to show a sense of loyalty to that community. Siedlok and Hibbert (2014, 204) write that tribalism could be so apparent in some cases that researchers who engage in interdisciplinary research might be looked down upon by their co-disciplinary peers as ‘less competent’. Disciplines might engage in protectionism as well, hiring within discipline in order to develop and protect job markets for their PhD graduates, which would further enforce tribal identities and loyalties (Jacobs and Frickel 2009, 54–55).

The most frequently cited tribal separation in academia, of course, is between science and arts. Beyond questions of methodology and epistemology, observers often describe a tension between physical sciences, health disciplines, and engineering, on the one hand, and arts, humanities, and social sciences on the other (Campbell 2005; Lach 2014; Pedersen 2016). In some cases, this tension manifests as a perception that science-based disciplines are inherently superior or more valuable than arts disciplines (Gardner 2013, 243; McCoy and Gardner 2012, 46). Such perceptions can lead to ‘power imbalances’ when interdisciplinary collaborations are attempted (Campbell 2005, 276;Gardner 2013), in which some researchers participate only in ‘service’ to the research objectives of other members of the team (Barry, Born, and Weszkalnys 2008).

Unfortunately, empirical data supporting claims made under this theme are scant, especially since the year 2000. Gibson et al. (2019, 58), in examining a program to support interdisciplinary research collaboration at the University of Wollongong in Australia, report some instances where researchers found language barriers to be a problem when working across disciplines, but also report other instances where participants felt the scheme enabled them to access ideas and knowledge that was new to them. Lattuca interviewed 38 American researchers who had recently engaged in interdisciplinary collaborations and found that some respondents felt that academic disciplines were ‘arbitrary’, ‘social artifacts’, and even ‘products of deliberate racism on the part of Europeans’ (Lattuca 2001, 107–109), rather than representing methodological or epistemological barriers to sharing knowledge. Some of Lattuca’s other respondents did acknowledge differences in methods or terminology between disciplines, but noted that these challenges were not insurmountable (Lattuca 2001, 156–166). Working in the United Kingdom, O’Cathain, Murphy, and Nicholl (2008, 1580–1581) interviewed 20 researchers who had worked on interdisciplinary teams, and while some reported encountering friction between researchers over various research methods in the past (in particular, the perceived superiority of quantitative over qualitative methods), none of the interviewees actually held these views themselves.

This is not to say that challenges to interdisciplinary collaboration are completely unsubstantiated. A 2016 report to the British government, in which 2358 academics across the country were surveyed about their attitudes toward interdisciplinary research, showed that over half of the survey’s respondents thought that communication in interdisciplinary teams was a challenge, and about 40% of respondents reported that interdisciplinary research was looked down upon by discipline-based peers (Technopolis Group 2016, 27). However, this single report does not really provide enough data to support the wide-ranging narrative described by many authors, which is that disciplines are composed of staunch, monolingual loyalists who live in impenetrable silos of jargon to which only the initiated have access, and who are suspicious of and hostile to residents of neighbouring disciplines. More importantly, empirical data in this area of research is rare in general – the British report just described, for instance, is one of only a very small number of large statistical studies investigating interdisciplinary research since 2000. In the absence of more compelling evidence, claims of disparate communities, severe epistemological barriers, and disciplinary tribalism may be more myth than fact.

Theme 3: Unique Case Studies

A third subset of the academic scholarship on interdisciplinary collaboration examines individual case studies. These are overwhelmingly dominated by studies of giant research consortia, often composed of multiple university partners from wide-ranging international locations, government stakeholders, and participants from private industry. The cases described in these reports are usually well funded, long term, institutionalized, and staffed by numerous active researchers. Here are a few examples:

• Austin, Park, and Goble (2008) investigate an interdisciplinary research project on medical ethics, funded by the Social Sciences and Humanities Research Council of Canada. The project had 18 research partners, including ‘clinicians (from medicine, midwifery, nursing, psychology, physical therapy, social work, and pastoral care) and scholars (from anthropology, law, philosophy, psychology, and theology)’ (2008, 560). The project’s funding ‘permitted research assistants as well as a half-time coordinator, and the Centre [the University of Alberta’s Research Ethics Centre] provided a home base’, so there was significant institutional support and administrative infrastructure available to the project’s leaders (2008, 561).

• Brennan and Rondón-Sulbarán (2019) report on a ‘project consortium’ of 45 partners including ‘core scientists, scientific experts, professional practice experts, strategic case actors and advisory board members’, representing ‘a range of different disciplines and fields of expertise including: environmental engineering, electronics, microbiology, environmental sciences, biomedical sciences, public health, psychology, behavioural sciences, and management. Practice experts worked in public agencies or NGOs in a range of country settings’ (2019, 485).

• Gardner (2013) examines a ’[USD] $20 million, 5-year, federally funded project focused on ecological sustainability’. The project ‘spanned over 25 distinct disciplines, including those from the biophysical sciences, social sciences, humanities, and professional fields’ (2013, 246). 42 researchers were affiliated with the project.

These are meant to be only a few illustrative examples, but the literature is replete with similar studies.

Case study research can yield a variety of productive outcomes. Single case studies, in particular, can demonstrate the existence of a process or a phenomenon that previous scholarship has suggested might not be possible or might not currently exist. They can also provide depth of detail on a particular issue that other studies have not been able to achieve because, in trying to capture a bigger picture, they have been focused on a more superficial level.

Case studies on large, well-funded, institutionalized interdisciplinary research projects, such as the examples related above, are uniformly positive about the ability of these projects to bring researchers together for successful interdisciplinary collaboration. As single case studies, they are demonstrating that interdisciplinary research collaboration is possible, and they are also providing a level of depth on the specific activities and interactions that might be required to achieve success in this area. Viewed as a whole, the key message from this literature appears to be that centrally managed, funded, independent research institutes are the best way – perhaps, the only way – to achieve meaningful collaboration between researchers across disciplines.

However, these case studies have limited potential to contribute to the greater discourse. Large research institutes are expensive, and they require dedicated resources, institutional support, administrative support, and infrastructure. They are risky for budget-conscious universities, and often rely on scarce external partners for funding. The model that these institutes rely on cannot be scaled up to include large numbers of researchers. Even in the universities where these funded research consortia exist, they are still sporadic, because of their resource requirements.

While these case studies offer useful examples of how central management, ample funding, and institutional support can promote interdisciplinary research, ultimately they have little capacity for generalizability, especially with respect to questions of how to enable new interdisciplinary collaborations. The authors of these studies are describing exceptions, rather than data points that may be used inductively to form rules.

Theme 4: Measuring Interdisciplinarity

A fourth theme includes studies that attempt to measure aspects of interdisciplinarity. A number of difficulties in evaluating interdisciplinary research have been suggested, such as finding peer reviewers with the right disciplinary expertise to be able to assess the quality of research funding proposals (Bammer 2016), or the problem of properly identifying interdisciplinary research after it has been published (Abramo, D’angelo, and Zhang 2018).

Studies under this theme tend to focus on three aspects of measurement. Most commonly, studies measure the success of interdisciplinary research through publication metrics, such as the quantity of research outputs for a given research project, the number of times a publication is cited by other publications, or the ranking or impact factor of the journals in which the research gets published. The obvious problem with these metrics is that quantity is a poor substitute for quality, and, in any case, number of citations and journal impact factors are only indirect indicators of quality, at best. However, these metrics may be the best available, since subjective evaluations of quality, such as ‘contribution to knowledge’, are difficult to assess. In any case, studies that evaluate publication metrics have produced mixed results, rendering the question of whether interdisciplinary research results in better quality publications than discipline-based research inconclusive (e.g.Abramo, D’angelo, and Di Costa 2017; Judge, Weber, and Muller-Kahle 2012). Not surprisingly, no consensus has emerged on how to evaluate interdisciplinary research output effectively (Mansilla 2006).

Another direction of inquiry is to evaluate the success of an interdisciplinary collaboration in terms of cohesion of the team or sustainability of the project, by looking at how long an interdisciplinary collaboration lasted, or how many partners were involved, or other similar metrics. Klein (2008) for instance, recommends evaluating interdisciplinary research using indicators of team integration, interaction between research partners (including ability to manage and resolve conflicts), and organizational factors such as leadership and management.

A third group of studies measure the degree of interdisciplinarity itself, as in, is research project X or publication X ‘more’ interdisciplinary than research project Y or publication Y? This can be accomplished, for instance, by analyzing publication metrics to evaluate the disciplinary diversity of co-authors or of citations in a reference list (Abramo, D’angelo, and Zhang 2018).

Considered together, the underlying objective of the studies under this theme is not entirely clear. Lyall et al. (2015, 137) point out that ‘Improved evaluation criteria and processes are the key to achieving a more stable and consistent role for interdisciplinary initiatives of various kinds within academic and research-based organizations’, implying that if we can evaluate interdisciplinary research, we might be able to generate advice on how to execute it successfully. But studies that attempt to measure or evaluate interdisciplinary research are not contributing to this goal. Evaluating the publication success of interdisciplinary research, as compared to research based within a single discipline, for instance, could demonstrate that interdisciplinary research is desirable (if it can be shown to result in better publication outputs) or a waste of effort for researchers motivated by career advancement (if it tends to result in weaker publication outputs). But these conclusions are not very relevant, since other compelling reasons to pursue interdisciplinary research already exist, such as the potential for addressing real-world problems, as noted earlier. One might be interested in investigating the factors that contribute to cohesion or integration in interdisciplinary research teams, in order to provide advice to teams on how to work together successfully. But indicators that measure how cohesive or integrated a team is, as Klein (2008) proposes, do not accomplish this – they can assess whether one team is more cohesive or integrated than another, but not how they got there. And it is not immediately clear at all why it might be useful to know whether a research project is more interdisciplinary or less interdisciplinary relative to another project – there is no obvious connection between degree of interdisciplinarity and effective strategies for conducting interdisciplinary research. Some studies attempt to measure all three of these aspects (i.e. publication success, cohesion and integration, and degree of interdisciplinarity) at the same time, and in doing so, provide an effective demonstration of how difficult it is to articulate a research objective around evaluating interdisciplinarity (e.g. Bark, Kragt, and Robson 2016; Brennan and Rondón-Sulbarán 2019).

Theme 5: Barriers to (and Enablers of) Interdisciplinary Collaboration

Studies that deal with barriers and enablers of interdisciplinary collaboration make up the final theme in this literature. The barriers discussed under this theme mainly relate to perceptions that interdisciplinary research is harder to produce, and yields lighter rewards, than discipline-based research. The general implication is that researchers are unwilling to engage in interdisciplinary collaboration because they are afraid that the costs of collaborating with colleagues from other disciplines will outweigh the benefits.

Several different types of cost appear in these studies. Some observers have argued that interdisciplinary research requires more time and resources than discipline-based research, in part because researchers have to learn a new language and new skills, but also because of an additional general effort of working with people not based in the same home discipline (McCoy and Gardner 2012, 46; Technopolis Group 2016, 7–8). It may also be more difficult to secure grant funding for interdisciplinary research (Kandiko and Blackmore 2008, 91), perhaps because, as has been mentioned, there are no appropriate peer reviewers for interdisciplinary grant applications (Bammer 2016; Lamont, Mallard, and Guetzkow 2006; Laudel 2006). This is more of an opportunity cost, since time lost while working on an unattainable interdisciplinary grant application might have been better spent on a discipline-based application instead.

Different rewards are also frequently discussed, with the general sentiment that rewards for interdisciplinary research are less advantageous than rewards for discipline-based research. Lattuca (2001, 37–38) argues that it can be harder to get interdisciplinary research outputs published in good journals (or even published at all), and that when successful, the resulting publications are given less weight toward tenure and promotion by a researcher’s own departmental committees. Campbell (2005, 574–575) writes that navigating the peer review process is more difficult for interdisciplinary research, because, as with grant applications, reviewers are seldom knowledgeable in the multiple discipline areas drawn on for the research. Interdisciplinary research publications can also have trouble attracting citations, since they lie outside the mainstream debates (Campbell 2005, 574). Perhaps due to these factors, there is a perception that it takes longer for the outputs of interdisciplinary research to materialize than for discipline-based research (Brown, Deletic, and Wong 2015, 317).

These points contribute to an overall sense that interdisciplinary research is career-limiting, and should be avoided by career-minded research professionals (Lattuca 2001, 34–43; Rhoten 2004, 9;Technopolis Group 2016, 7–8). Junior scholars, in particular, are believed to be less interested in interdisciplinary collaboration, because the risks involved in not producing high-impact publications, or not producing them quickly enough, are too great for emerging academics trying to build a reputation (Kandiko and Blackmore 2008, 91; McCoy and Gardner 2012, 47).

One major failing of this branch of the literature is that empirical studies nearly always rely on the perceptions of human respondents rather than on the actual measurement of the phenomena or processes being described. For example, Kandiko and Blackmore (2008) form their conclusions about junior researchers’ reluctance to engage in interdisciplinary research from interviews with ten senior research leaders, without attempting to observe the number of junior scholars who participate in interdisciplinary research as a proportion of the overall junior scholar population. Bridle et al. (2013) use a survey of 26 participants in an early career researcher networking event to inform their views on what constitutes a successful initiative to promote research connections across disciplines – but they do not report on the actual interdisciplinary collaborations these people may have participated in after the networking event was concluded. Townsend, Pisapia, and Razzaq (2015, 671) report that ‘Leadership is important at both the university and college levels if interdisciplinarity is to thrive’, but they base this claim on the results of 127 survey responses and 25 interviews, not on any direct measurement of how leadership might improve interdisciplinarity. And I do not mean to suggest that these studies are not large enough or not rigorous enough to produce believable results, or that quantitative data are in any way more valid than qualitative data. The Technopolis report to the British government, cited several times above, involved mixed methods, including focus groups with 45 participants, 31 interviews, and a survey with over 2,000 respondents, and is a model of methodological rigour, but it too relies entirely on the perceptions of human respondents (Technopolis Group 2016).

The problem here is that human perceptions are subjective. Respondents in the studies cited above may believe that interdisciplinary research takes longer to publish, or is less likely to appear in higher quality journals, or does not attract as much grant funding as single-discipline research. But these beliefs may be wrong, even if many or most people believe them. And without more direct observation, it would be difficult to know one way or another.

Of course, positivist epistemology is only one way to approach questions about barriers and enablers of interdisciplinary research. Other approaches may be just as valid. However, most of the studies under this theme actually do explicitly take a positivist perspective: they ask, ‘What factors present barriers to conducting interdisciplinary research?’, and then they ask academics, administrators, or university leaders, through surveys and interviews, to offer their own perceptions of what these factors may be. The flaw in these studies is that they then present these human perceptions as fact, when they should more accurately be presented as belief. In a more robust body of knowledge, human perceptions would be presented alongside direct observation, as a method of triangulation. However, in this literature, the triangle used for triangulation is highly skewed toward human perception, and the imbalance is not acknowledged.

In the absence of studies reporting more direct observation, it is not surprising that nearly all of the assertions listed above are contested. For instance, Brown, Deletic, and Wong (2015, 317) argue that it is actually easier to obtain grant funding for interdisciplinary research projects, because interdisciplinarity is in high demand by funding agencies, and because in turn there are now more grants that can only be awarded to interdisciplinary projects. Millar (2013) finds that interdisciplinary research is not career limiting, and that interdisciplinary researchers are more likely to get a job after obtaining their PhD than researchers based in a single discipline and are also likely to have more publications. Rhoten and Parker (2004, 2046) argue that early career researchers are more likely, not less likely, to engage in interdisciplinary research collaboration, because they are eager to build their careers quickly and are therefore more willing to take risks, and also because they might experience pressure not to decline requests for collaboration from senior colleagues. In general, none of the arguments and counter-arguments in this space have been adequately tested through direct observation, and assertions are frequently presented as facts when they are actually beliefs.

Discussion and Conclusion

When it comes to informing strategies for promoting new interdisciplinary research collaborations, the scholarly literature has a problem with operationalization. This problem is evident when the literature is viewed as a whole, as well as across all five themes within the corpus. In many cases, the aims of the scholarship in this area are not entirely clear, and as a consequence, studies tend to be poorly designed and research methods often appear to be chosen inappropriately. While there is vast potential for real growth in our understanding of interdisciplinary research collaboration, existing studies are not contributing as efficiently as they could be to the primary question of how to get researchers from different disciplines to work together productively.

Earlier, I argued that studies that attempt to define interdisciplinarity have not focused on the definitions that are most useful for empirical research on how to promote new collaborations. This problem is common across the other four themes as well. For instance, the overall research objective of studies within the ‘Working Across Disciplines’ theme is, ostensibly, to determine how teams of collaborators, based in different academic disciplines, can work together to produce interdisciplinary research. But key concepts that would facilitate analysis have not been defined. How might one define success in interdisciplinary research collaboration? Are we interested in sustained teamwork, or one-off collaboration, or both? Is collaboration itself the end goal, or are we concerned with the quantity or quality of research outputs that the project produces? Studies under this theme are so preoccupied with the unproven assumption that discipline-based researchers do not work well together because of language barriers or clashing cultures and epistemologies, that they have neglected to define the basic terms that are necessary for informing primary empirical research questions.

In the ‘Unique Case Studies’ theme, practically all studies select along the dependent variable, that is to say, they examine cases where interdisciplinary collaboration has happened, but never cases where it failed to emerge or where interdisciplinary teams fell apart or experienced major conflict. As a result, studies under this theme generally report that large blocks of funding, critical mass, central organization, institutional support, and administrative independence are all contributing factors to successful interdisciplinary research collaboration. Perhaps this is true, but it should not be surprising that healthy financial support and organized resources should lead to success. What these studies cannot do is explain how new interdisciplinary collaborations might form under circumstances experienced by the vast majority of research academics, which is without significant funding or dedicated resources, and in parallel with the other responsibilities of academic work, including standard discipline-based research, service to the university and the discipline, and teaching duties.

Studies that aim to measure interdisciplinarity have the greatest challenge in articulating a research objective, as has already been discussed. It is not clear what they want to measure, why they want to measure it, or what these measurements will contribute to our understanding of how to form new interdisciplinary collaborations.

Studies under the fifth theme, which examine barriers and enablers of interdisciplinary collaboration, rely entirely on human perceptions acquired through survey, interview, and focus group research. These methods are excellent for investigating feelings, emotions, and lived experience – which, of course, can be very useful information – but they are not the best way to measure actual outcomes.

Furthermore – or, perhaps, as a result of the methodological issues described above – there is simply not enough empirical data available to support many of the claims put forward in this literature. Is there really a tension between researchers based in arts disciplines and their colleagues based in science, and if so, is this tension a significant barrier to collaboration? What works to incentivize interdisciplinary collaboration? What institutional or organizational changes can university departments, think tanks, and government partners make to support or enable this collaboration? This last area is especially fertile ground for empirical investigation. There is tremendous potential for trialling creative strategies for enabling interdisciplinary collaboration, including networking events, funding competitions, or changes to professional performance standards, but, as demonstrated by the systematic review reported above, very little formal testing has been done. Experimentation in this area is, of course, a challenge, in that it would require the cooperation of institutional leadership who would need to be willing to take risks with money and other resources. But with so little empirical data in existence, any attempt to move the needle even slightly would represent a significant contribution to knowledge.

It should be noted that none of this discussion is meant to imply that interdisciplinary research is in any way superior to discipline-based research. In all likelihood, research disciplines will continue to be the primary locus of the development of knowledge. Research disciplines give individual researchers a community in which they can find support from peers, creative competition, and critical mass – as well as an important sense of belonging. Discipline-based research allows researchers to develop a mastery of a subject to a level that they might not be able to attain if they were to spread their time and resources across diverse knowledge areas. Disciplines are also convenient administrative divisions that universities and other research institutions can use for management purposes.

That being said, interdisciplinary collaboration is desirable because of its ability to promote creativity and innovation, and also for its potential to help academic research contribute to addressing real-world challenges. In addition to discipline-based research, interdisciplinary research can add value that is greater than the sum of its parts. There is, therefore, understandable interest in how interdisciplinary research can be increased, without diminishing discipline-based research. However, the scholarship in this area is far too focused on the precision of taxonomies, unique cases with little potential for generalization, and the technicalities of publication metrics, and almost not at all concerned with empirical demonstration of how to get researchers to collaborate across disciplines. If this literature is to be advanced, scholars working in this area will need to pivot toward a research agenda that prioritizes empirical data collection to test the corpus of theoretical scholarship that already exists.

Disclosure Statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Joshua Newman

Joshua Newman is Associate Professor in the Faculty of Arts at Monash University in Melbourne, Australia. His research examines the politics of public-sector decision making, including research utilization, evidence-based policy making, and government-industry interaction.

Appendix Appendix 1.

Systematic Review - Included Studies

Table

Citation	Location	Year of Data Collection	Intervention	Research Method
Adams, S.M., Carter, N.C., Hadlock, C.R., Haughton, D.M. and Sirbu, G., 2008. “Proactive Encouragement of Interdisciplinary Research Teams in a Business School Environment: Strategy and Results.” Journal of Higher Education Policy and Management 30(2), 153–164.	US	2001–2004	internal grant funding scheme	network analysis
Boden, D. and Borrego, M., 2011. “Academic Departments and Related Organizational Barriers to Interdisciplinary Research.” Higher Education in Review 8, 41–64.	US	2007–2008	none	interviews
Bridle, H., A. Vrieling, M. Cardillo, Y. Araya, and L. Hinojosa. 2013. “Preparing for an Interdisciplinary Future: A Perspective from Early-Career Researchers.” Futures 53, 22–32.	EU	2012	workshop	survey
Butson, R., Hendrick, P., Kidd, M., Brannstrom, M. and Hedberg, M., 2012. “Developing a Virtual Interdisciplinary Research Community in Clinical Education: Enticing People to the ‘Tea-Room’.” Journal of Research in Interprofessional Practice and Education 2(3), 320–338.	NZ	2010	website application	website user data
Gibson, C., T. Stutchbury, V. Ikutegbe, and N. Michielin. 2019. “Challenge-led Interdisciplinary Research in Practice: Program Design, Early Career Research, and A Dialogic Approach to Building Unlikely Collaborations.” Research Evaluation, 28(1), 51–62.	Australia	2017	internal grant funding scheme	survey, focus groups, and interviews
Kandiko, C. B., and P. Blackmore. 2008. “Institutionalising Interdisciplinary Work in Australia and the UK.” Journal of Institutional Research 14(1), 87–95.	Australia and UK	unknown	none	interviews
Lau, L. and Pasquini, M. W., 2004. “Meeting Grounds: Perceiving and Defining Interdisciplinarity Across the Arts, Social Sciences and Sciences.” Interdisciplinary Science Reviews 29(1), 49–64.	UK	2003	none	interviews
Lin, H., 2008. “Opportunities and Challenges for Interdisciplinary Research and Education.” Journal of Natural Resources and Life Sciences Education 37, 83–91.	US	unknown	none	survey
Milman, A., Marston, J. M., S. E. Godsey, J. Bolson, H. P. Jones, and C. S. Weiler. 2017. “Scholarly Motivations to Conduct Interdisciplinary Climate Change Research.” Journal of Environmental Studies and Sciences 7(2), 239–250.	US	2013	none	survey
Ravid, K., F. Seta, D. Center, G. Waters, and D. Coleman. 2017. “Catalyzing Interdisciplinary Research and Training: Initial Outcomes and Evolution of the Affinity Research Collaboratives Model.” Academic Medicine 92(10), 1399–1405.	US	2007–2013	internal grant funding scheme	network analysis
Rekers, J. V., and T. Hansen., 2015. “Interdisciplinary Research and Geography: Overcoming Barriers through Proximity.” Science and Public Policy 42(2), 242–254.	Sweden	2010–2011	independent interdisciplinary research institute	interviews
Rhoten, D., 2004. “Interdisciplinary Research: Trend or Transition?” Items and Issues 5(1–2), 6–11.	US	2002–2003	independent interdisciplinary research institute	survey and interviews
Salazar, M., T. Lant, and A. Kane. 2011. “To Join or not to Join: An Investigation of Individual Facilitators and Inhibitors of Medical Faculty Participation in Interdisciplinary Research Teams.” Clinical and Translational Science 4(4), 274–278.	US	unknown	internal grant funding scheme	survey
Townsend, T., J. Pisapia, and J. Razzaq. 2015. “Fostering Interdisciplinary Research in Universities: A Case Study of Leadership, Alignment and Support.” Studies in Higher Education 40(4), 658–675.	UK	2011–2012	university restructure	survey and interviews