Unique Challenges of Multidisciplinary Clinical Science: Perspectives From a Multidisciplinary Team

Challenge 1: Disciplines-Specific Biases

Researchers tend to hold discipline-specific biases about the best ways to gain knowledge. Indeed, this is one of the advantages of multidisciplinary work! In theory, multidisciplinary collaborations involve exploring diverse viewpoints from researchers with unique backgrounds to produce novel insights. In practice, however, this multiplicity of viewpoints can lead to conflict and in extreme cases, intellectual impasses. In other words, although some discipline-specific biases are desirable and unavoidable, disciplinary dogmatism (i.e., strongly held disciplinary biases) can impede collaborative efforts.

Disciplinary biases can be categorized into methodological and epistemic biases. An example of methodological bias in clinical psychology is the emphasis placed on large sample sizes. Large sample sizes (e.g., n > 100 per cell) are often considered paramount for impactful and replicable clinical-psychological studies. Although there are certainly clinical psychologists that work with smaller sample sizes, including qualitative and mixed-methods researchers, in our experience, this work is often perceived by mainstream researchers as useful for preliminary hypothesis generation rather than making definitive empirical claims. This bias toward large sample sizes can lead clinical psychologists to assume that other disciplines place a similar emphasis on sample size. However, in basic vision-science research, there are foundational, highly replicable studies that consist of fewer than five people and often include the authors themselves! For example, foundational contrast-masking work by Legge and Foley (1980) has been cited more than 1,500 times, and the sample consisted of three individuals, one of which was Foley. Thus, when clinical psychologists collaborate with vision scientists, there may be understandable disagreement with respect to study design. Basic vision scientists may perceive clinical psychologists as being overly concerned with study recruitment, whereas the clinical psychologist may perceive the vision scientists as being overly optimistic with respect to strength of effect sizes and degree of heterogeneity in a clinical population of interest.

Disciplines also often differ with respect to what constitutes a substantial contribution to the literature. A linguist might conceive of an “impactful” study as involving the manipulation of at least two variables in each group of participants (e.g., manipulating the complexity of the task and the complexity of the linguistic materials). Ideally, at least one of the conditions created by this manipulation would be unique to the new study. In contrast, a clinical psychologist might find a simpler study—such as comparing the effect size of a single manipulation across two groups—as a significant contribution to the clinical literature. These disciplinary differences may not only lead to disagreement regarding task design but also may have long-lasting consequences for participating scientists. When the dossier of the collaborating linguist is evaluated, the linguist’s collaborative work may be deemed too simplistic by the standards of the linguistic field.

In addition to such methodological biases, higher-level epistemic biases are also embedded, often implicitly, in disciplines. For example, neuroscientists and psychologists often disagree about the best level at which to explain human behavior (Krakauer et al., 2017). Neuroscientists, by definition, are largely interested in brain-level explanations, whereas psychologists tend to be interested in psyche-level and social-level explanations. Note that this does not mean that neuroscientists are uninterested in psyche-level and social-level phenomena but rather, that they tend to focus on brain-level explanations of such phenomena. This mismatch in epistemic biases has led to significant controversy with respect to psychopathology in particular. Various forms of psychopathology have long been hypothesized to be best studied and explained at the level of the brain (Insel & Quirion, 2005). Indeed, many neuroscientists and a good deal of neuroscientifically leaning clinical psychologists believe that the altered neural circuitry responsible for psychopathology will someday be identified. However, other authors have argued that higher-level psychological- and social-level explanations may be more empirically tractable (Sharp & Miller, 2019) or that multilevel approaches are necessary (Cacioppo & Berntson, 1992). The importance of such multilevel approaches is especially apparent when identifying brain-based markers of behaviors: Similar behavioral profiles may result from distinct underlying neural profiles and vice versa. Such disagreements about the “best” level at which to study a given phenomena can permeate almost all aspects of multidisciplinary collaborations and can be especially insidious because of their often implicit nature.

Another example of epistemic bias within and outside of clinical-psychological science concerns the degree to which standpoint epistemology is emphasized. In the context of scientific research, standpoint epistemology holds that a researcher’s life experiences, in part, shape how the researcher gains and produces knowledge. According to proponents of standpoint epistemology, marginalized voices should be sought out and elevated because they can provide unique and novel insights to scientific endeavors (Friesen & Goldstein, 2022). Indeed, the rise of participatory research in the social and health sciences can be seen as an outgrowth of the core tenets of standpoint epistemology (Shalowitz et al., 2009; Tekin, 2022). However, disciplines seem to vary in the degree to which they seek out marginalized viewpoints. In particular, we have observed social and health scientists (including clinical psychologists) engage more actively with standpoint epistemology relative to biological scientists. This emphasis on incorporating underrepresented voices in the social and health sciences is somewhat unsurprising: An individual’s “standpoint” is heavily socially mediated. Thus, when social and health scientists collaborate with more biologically motivated scientists, there can be disagreement with respect to the importance of seeking out and considering marginalized viewpoints. Such conflict can be especially thorny because standpoint epistemology may be seen as having not only epistemic value but also moral value. Thus, disagreements about standpoint epistemology can be emotionally and politically charged and lead to ruptures in a multidisciplinary team if not navigated wisely.

Potential Solutions

Many of the foregoing methodological and epistemic biases become problematic only when disciplinary dogmatism is present. At the individual level, we recommend researchers critically examine their own discipline-specific biases and identify the provenance of these beliefs. In much the same way that clinicians are taught to practice cultural humility, we recommend that researchers practice disciplinary humility. Part of this humility involves acknowledging the limits of one’s own knowledge and being open to learning from collaborators. At the institutional level, training programs should facilitate examination of disciplinary biases (Aftab et al., 2024). One way to do this would be to offer history or philosophy of science courses in graduate programs to highlight the multiplicity of scientific epistemologies and to encourage students to critically examine and articulate their own methodological and epistemic biases. Understanding these biases will help trainees become better scientists and collaborators.

Challenge 2: Multidisciplinary Fluency

Communicating between disciplines is challenging because each discipline has idiosyncratic terminologies, methodologies, and theoretical frameworks. In particular, discipline-specific jargon can be a barrier to collaborative efforts. Despite the multidisciplinary challenges that accompany its use, jargon is often genuinely useful for concise communication in a discipline (Hirst, 2003). If every article about schizophrenia, for example, was forced to delineate the complex and controversial history of the construct (Jablensky, 2010; Kyziridis, 2005), authors could easily spend their lives writing the introduction to a single article. Thus, one must accept that some terms and concepts that are fundamental to a discipline may not be immediately accessible to the nonexpert. In other words, some basic knowledge of a given discipline is requisite for effective and efficient cross-disciplinary communication. At the same time, it is unreasonable and impractical to expect researchers to become experts in multiple fields before undertaking multidisciplinary collaborations. Therefore, we propose that a middle ground must be struck between expertise and ignorance. We refer to this intermediate knowledge level as “multidisciplinary fluency.”

Our research team recently experienced a breakdown in communication due to poor multidisciplinary fluency. We were interested in characterizing speech patterns of individuals at risk for developing psychosis. As part of our clinical assessment, interviewers rated a number of interviewees as having atypical prosody. However, when we presented recordings of these interviews to a linguist collaborator (M. Goldrick), he was perplexed by the prosody ratings. Specifically, he deemed many aspects of the interviewee’s prosody to be well within normal limits. This disagreement arose because in clinical-science contexts, “prosody” often simply refers to degree of monotony in speech, whereas from a linguistics perspective (Grice et al., 2023), prosody is a multidimensional construct (e.g., signaling grammatical structure, pragmatic properties, and emotional content). This example highlights a particularly troublesome challenge to multidisciplinary collaboration: Sometimes, the same term can mean different things depending on disciplinary context. This phenomenon can even occur within the same discipline: As Barack et al. (2022) pointed out, neuroscientists often disagree about the precise meaning and implications of the term “causality.”

The “curse of knowledge” may further exacerbate the challenge of communicating across disciplines. The curse of knowledge is a well-studied psychological phenomenon in which individuals with knowledge on a given topic assume, often incorrectly, that others have similar levels of knowledge (Birch & Bloom, 2007; Birch et al., 2017; Camerer et al., 1989; Xiong et al., 2020). This curse is evident in a variety of situations. In computer programming, programmers may fail to write code that is readable by others because the purpose and logic of the code seem obvious to the person writing it. It is only when another programmer looks over the code that the curse of knowledge becomes apparent. In multidisciplinary collaborations, the curse of knowledge may occur when an expert in a given area assumes that other team members have or should have a similar level of expertise in that area. V. Pokorny was recently at a multidisciplinary conference during which an animal neurophysiologist expressed frustration at the ignorance of psychiatrists and clinical psychologists with respect to brain circuits involved in eye movements. Many psychiatrists and clinical psychologists. It is likely that the animal neurophysiologist was experiencing the curse of knowledge. As this example highlights, the curse of knowledge can engender frustration in both the knowledge holders and knowledge seekers in multidisciplinary teams.

Developing multidisciplinary fluency may be especially challenging for clinical psychologists because training programs and academic institutions often incentivize specialization. Prospective clinical-psychology graduate students are frequently expected to articulate specialized research programs to gain admission to PhD programs. Upon matriculation, the numerous roles that a clinical-psychology trainee must juggle (e.g., clinician, data scientist, statistician) are such that there might not be enough time in the day to engage with other disciplines. Given these obstacles, it is perhaps unsurprising that clinical-science trainees may struggle to develop multidisciplinary fluency.

Potential Solutions

Institutions may enhance multidisciplinary fluency by placing less emphasis on specialization during graduate training, faculty selection, and promotion. One way this could be achieved is by requiring students to be mentored by at least one faculty outside of the student’s primary area and/or requiring students to complete a certain number of “breadth” courses. Indeed, such requirements appear to be more and more common in psychology and neuroscience departments. In the specific case of clinical-science training, we propose that clinical-training models that focus less on meeting regulatory demands and focus more on exploring diverse perspectives may improve multidisciplinary fluency. Thus, it is possible that recent efforts to avoid the high regulatory burden of the American Psychological Association accreditation process (Bootzin & Treat, 2015) may have the unintended and welcome consequence of improving multidisciplinary fluency. Indeed, a recent graduate-training reform proposal from leaders in the field made a similar argument: More flexibility would allow training programs to devote more time and resources to developing multidisciplinary fluency (Berenbaum et al., 2021).

At the individual level, researchers may improve their multidisciplinary fluency by engaging with a broader range of conferences and journals. In particular, there are a number of journals (e.g., Philosophy, Psychiatry and Psychology, Interdisciplinary Science Reviews) and academic conferences (e.g., Association for Interdisciplinary Studies) that explicitly aim to promote dialogue across disciplines. Supporting and engaging with such academic spaces is a valuable way to improve multidisciplinary fluency. In addition, to reduce collaborative friction engendered by the curse of knowledge, multidisciplinary team members may benefit from viewing part of their role as “disciplinary translator” for other team members. There is substantial evidence from industrial-organizational psychology that a key driver of team success is organizational structure, such as clearly identifying team roles, duties, and procedures (Kozlowski & Bell, 2019). Thus, efficacy of research teams may be improved by explicitly assigning roles (e.g., team leader, disciplinary translator) in multidisciplinary teams.

Challenge 3: Risk

Multidisciplinary collaborations are often forged with the aim of developing novel approaches to studying historically difficult topics. However, this potential for scientific innovation is counterbalanced by substantial risk. Consider a fairly common multidisciplinary collaboration: A novel task originally developed by neuroscientists is used to study altered cognition in a clinical population of interest. The most obvious risk of such work is that task indices may not meaningfully associate with any clinical measures. In such a case, a great deal of resources (e.g., participants’ time and energy, research hours, grant dollars) were spent enumerating how a novel task is uninformative with respect to the disorder. Even if a significant task association is observed, researchers must contend with measurement invariance: The task may not measure the construct of interest in the same manner between cases and control subjects (or people with more severe symptoms relative to people with less severe symptoms). This risk is substantial in such contexts because tasks derived from other disciplines are rarely designed with clinical populations in mind.

In addition to the scientific risks of multidisciplinary research, there are also career-related risks (alongside a panoply of opportunities to be sure!). Recent work has shown that interdisciplinary early career researchers face significant career impediments (Berkes et al., 2024). Concerningly, Berkes et al. (2024) found that “initially interdisciplinary researchers on average decrease their interdisciplinarity over time.” These findings are generally consistent with our own experiences. As multidisciplinary researchers, it can be difficult to identify clear-cut target audiences for our work. Journals often focus on a single discipline (or a small group of highly related disciplines), and editors are usually able to judge multidisciplinary work by the standards of only a single discipline. Similar issues can arise in securing funding and faculty appointments because grant-funding bodies and academic departments are largely organized according to discipline. In addition, grant-review and academic hiring committees may unfavorably evaluate applicants with interests that span multiple disciplines because this is thought to indicate an “unfocused” research program. Thus, the careers of promising multidisciplinary researchers may be stymied by a lack of institutional understanding and support of their work. Given these career-related risks, it is understandable (and unfortunate) that talented and ambitious young researchers may be reluctant to pursue multidisciplinary research.

Potential Solutions

With respect to scientific risk, clinical psychologists can mitigate some of the riskiness of multidisciplinary collaborations via careful study design and data analysis. In particular, researchers must take steps to ensure the methodologies and concepts from other disciplines can be meaningfully translated into the study of mental-health disorders. Pilot studies are useful for assuring the feasibility, acceptability, and translatability of novel tasks and methods. During data analysis, researchers must carefully assess measurement invariance and inspect the influence of confounding variables. By doing so, researchers may improve the degree to which inferences can be made regarding mechanisms of psychopathology.

In terms of career risk, given the current academic and funding landscape, early career researchers may offset some risk by establishing a firm identity in a single discipline before (or while) branching out into more multidisciplinary work. This approach allows a given researcher to be perceived by academic gatekeepers as fundable and employable while also allowing the researcher to pursue innovative multidisciplinary work. At the institutional level, more efforts should be made to support multidisciplinary research and researchers. In particular, interdisciplinary cluster hires have been suggested as a method of fostering interdisciplinary careers (Sá, 2008); however, the empirical support for this argument is mixed (Bloom et al., 2020). Colleges and universities may further foster multidisciplinary work by modifying tenure and promotion criteria to reward multidisciplinary efforts and allocating resources to multidisciplinary funding mechanisms. Likewise, national and international funding agencies may encourage multidisciplinary careers by incentivizing grant proposals that span disciplines and by evaluating the degree to which multidisciplinary teams have demonstrated the ability to effectively collaborate. By doing so, institutions may “walk the walk” by supporting the type of work that they often claim to value.

Limitations

In the present work, we have discussed challenges we have experienced when collaborating across disciplines that although distinct, may be considered adjacent to one another in a hypothetical “discipline space” (e.g., clinical psychology and neuroscience are closer in such a discipline space than clinical psychology and meteorology). Thus, unique challenges may arise when collaborating across disciplines that share less conceptual and epistemological common ground. Furthermore, although we have, for parsimony, largely described epistemological tenets as being roughly homogeneous in a given discipline, there are well-known examples of subdisciplines that are seemingly in direct opposition to one another. For example, quantitative researchers in a given discipline may encounter unique challenges when collaborating with qualitative researchers of the same discipline (and vice versa) because of opposing viewpoints vis-à-vis how to best accumulate knowledge. Finally, in the present work, we have described only challenges we have faced when collaborating with other academic partners. We suspect that unique challenges arise when academics collaborate with nonacademic partners, such as community members with lived experience, industry partners, and policymakers. Thus, future work would do well to compare the experiences we report here with the experiences of teams operating in other collaborative contexts.

Conclusion

Clinical scientists are well poised to build bridges and forge productive collaborations across disciplines. The challenges and corresponding solutions we have presented are not meant to be comprehensive or incontrovertible. Instead, we hope that this article sparks fruitful dialogue among clinical scientists and their collaborators. By overcoming the challenges presented, we are optimistic that multidisciplinary collaborations will catalyze clinical science and lead to improved understanding and treatment of mental-health disorders.