Preparation to address ethical challenges is an essential component of graduate training, and no less so for the psychological sciences. However, in the absence of uniform guidelines, approaches to training vary in form and quality. Classroom lectures and online training seem to be the mechanisms of choice, but these fall short. First, such approaches conflict with the scholarship on teaching and learning that makes it clear that having a meaningful impact depends on having students actively engaged in constructing their own learning. Second, research is consistent with intuition that the impact of courses is likely to be far less than what happens in a graduate student’s research environment. The conclusion is that promoting an ethical culture, and for the training of graduate students in particular, will be well served by enhancing the role of mentors. Examples of options to consider are: (1) recognizing that a primary advisor can be a mentor, but should certainly not be considered the only mentor; (2) emphasizing the importance of mentoring for individuals from underrepresented groups (e.g., because of gender or ethnicity); (3) strengthening the APA code of ethics to more fully articulate the full range and importance of mentoring; (4) developing and implementing mechanisms to evaluate and reward effective mentoring; and (5) providing targeted training for faculty advisors to empower them with tools and resources to be effective mentors for ethics generally and the responsible conduct of research specifically.
Keywords: Responsible conduct of research, Research ethics, Graduate students, Mentors, Teaching and learning
“Like any relationship, creation of a foundation of trust, nurturance, and support is the best context for a mentee to experience maximal growth.”
Science faces a crisis in integrity (Alberts et al., 2014; National Academies of Sciences, Engineering, and Medicine, 2017). While intentional lapses in honest reporting are a significant threat (National Academies of Sciences, Engineering, and Medicine, 2017), it is arguably even more important that the integrity of science can be compromised unintentionally through insufficient understanding or awareness of good practices of science (Ioannidis, 2005). The latter challenge is probably best illustrated by worries about the reproducibility of science (Baker, 2016). The field of psychological sciences is not immune to the reproducibility crisis (Open Science Collaboration, 2015). Some authors have even proposed that with the right methodological choices you can prove anything in psychology (Simmons et al., 2011). The risks here are that deficient research practices can result in false positive findings. To the extent this is true, the implication for translation of psychological findings into practice is at least disturbing and possibly harmful. The subject of this manuscript is how to minimize these challenges by better preparing the next generation of researchers. What, if anything, can be done to strengthen the integrity of science, and in psychological sciences in particular?
Guidelines and training to protect scientific integrity
Following WWII, the prominence of psychological sciences and growing ethical concerns led to the establishment of the American Psychological Association code of ethics (1953). Over 2000 psychologists created a document outlining standards for public responsibility, client relationships, teaching, research, writing and publishing, and professional relationships. The code of ethics was established to acknowledge the societal responsibility of psychologists to not only better understand human behavior through ethical research, but also, in treating mental illnesses through ethical clinical engagements. Since then, the APA code of ethics has been revised to reflect current ethical challenges (e.g., cultural competence, open science) through more specific and updated standards closely mirrored guidelines for responsible conduct of research (American Psychological Association, 2017).
Notwithstanding the APA code of ethics, responsible conduct of research (RCR) was not a prominent concern before the 1980s. However, in the face of several high-profile cases of research misconduct (e.g., summarized in Kalichman, 2013a), the National Institutes of Health (NIH) established a new requirement that trainees supported by NIH training grants should receive RCR instruction (NIH, 1989). Although at least in part a response to cases of research misconduct, the requirement could be characterized as calling for teaching about rules, regulations, and to some extent good practices of research, largely consistent with the APA code of ethics guidelines (e.g., need for institutional review board approval, informed consent, protection of human participants, mentoring). This approach to mitigating misconduct is questionable. While it is plausible that someone might be dissuaded from committing research misconduct if they were first told it is against guidelines or regulations to lie, cheat, or steal, it is more likely that most adults already know this. On the other hand, for cases of research misconduct that might be “accidental,” this approach may have some value. In any case, there is an implicit assumption that trainees would benefit from a better understanding of how research should be done.
The scientific community has largely focused on education or training as the remedy for deficiencies in research conduct, resulting in a proliferation of courses, workshops, seminars, and resources to promote RCR (Kalichman, 2013a). Many APA accredited psychology graduate programs have made research ethics courses a core requirement. Although such requirements as well as training required by the U.S. federal government (e.g., through the National Institutes of Health (NIH) and the National Science Foundation (NSF)) ensure access to RCR training, each institution has the liberty to decide what is to be taught within these courses ensuring a lack of consistency in goals and content covered (Phillips et al., 2018). To date, the overarching goals of RCR programs remain unclear.
While it might seem enough to say the goal is “responsible conduct of research,” that begs the question of what should be covered and how it should be covered. One answer is current NIH guidelines for RCR training in which nine topics are recommended (NIH, 2009). These include: conflicts of interest, human and animal subjects, mentoring, collaborations, peer review, data management, research misconduct, authorship and publication, and social responsibilities. However, as these are only guidelines, it might seem more useful to simply ask teachers of RCR about their goals for teaching (Kalichman and Plemmons, 2007). The answers were wide-ranging, including: (1) conveying new knowledge (e.g., ethical theories, rules and regulations); (2) nurturing skills (e.g., ethical decision-making, stress or time management); (3) encouraging positive attitudes (e.g., recognizing the importance of promoting RCR, continued learning, and valuing openness and transparency); (4) stimulating positive behaviors (e.g., conducting responsible rather than fraudulent research, being a good mentor, taking responsibility to learn and follow relevant rules); and (5) building community (e.g., providing opportunities for trainees to speak to one another in class, but to also do so outside of class with peers, mentors, faculty, and others). All of these are desirable outcomes, but not only is it unrealistic to expect to meet so many different goals at the same time, classroom education and training are unfortunately unlikely to have significant impact.
The hurdles limiting the impact of courses are numerous. First, it is certainly unrealistic to think that any one course, much less a seminar or workshop, could address all of the possible issues (Kalichman, 2013b). It would of course be even more dubious to expect much more than conveying limited knowledge if the approach were simply an online tutorial (Kalichman, 2005). Second, it should not be surprising that a single course is likely to be less important in shaping outcomes than the pervasive, daily exposure a researcher experiences over the course of a career (e.g., Anderson et al., 2007). Third, researchers may work in an institutional environment in which leadership, policies, and resources could unfortunately send a clear message contrary to the efforts made in the classroom. Finally, the culture of science is one in which the incentives are often perverse—publication in high impact factor journals, high volume of publications, and first authorship is valued and rewarded over publishing work more likely to be reproducible (Oransky & Marcus, 2016). The nature of the incentives widely perceived to define science would likely undermine the impact of attempts at training or education. In light of these challenges, it is important to ask how best to provide training.
For most educational goals, the scholarship on teaching and learning is clear: It is better for students to be actively engaged rather than being a passive audience for a lecture or simply reading material online (National Research Council, 2000; Brown et al., 2014; Freeman et al., 2014). While lectures or online training may have some value, meaningful impact should be expected only when learners are in effect generating their own learning by being actively engaged. This can be done in many ways ranging from small group case discussions to debates to role playing to conducting surveys to games (Plemmons and Kalichman, 2008). However, while such activities definitely strengthen courses, it remains true that the greatest impact is likely going to come from a trainee’s experiences in the laboratory, a culture and context largely dictated by their mentors.
It is increasingly recognized that graduate students and trainees receive the majority of their ethics education and information regarding the practical implementation of ethical standards, not from formal classroom training, but rather from the ‘hidden curriculum’— socialization into RCR that includes formal instruction from principal investigators (PIs) (who often serve as primary mentors), observed behaviors of other researchers, day-to-day interactions with members of the research team, and experiencing the culture of the research environment (Fryer-Edwards, 2002). In this setting, it is the PI who serves not only to set the cultural norms with regard to ethical scientific practice, but along with other senior laboratory staff also serves as a professional role model for trainees. PIs, thus, lead by example, modeling both “good” and “bad” behaviors that a trainee may internalize and integrate as normative behaviors of a seasoned scientist (Fryer-Edwards, 2002; Institute of Medicine and National Research Council, 2002; Peiffer et al., 2008; Satalkar and Shaw, 2019; Plemmons et al., 2020).
Role of mentorshipAs for RCR training and goals, mentorship in academic research has varied widely among and even within institutions. Who is considered a mentor versus a PI or an advisor, the degree of interactions, and responsibilities of mentors to their protégés are among many factors to be considered (Cobb et al., 2018). While the expected roles of advisors and mentors overlap and are variably defined, a primary “advisor” often has a supervisory role (i.e., their guidance might be seen more as expectations rather than advice). On the other hand, a mentor does not necessarily have a formal relationship with the mentee so their role is much more advisory than supervisory (i.e., their guidance would be considered as suggestions rather than expectations). In clinical psychology, formal mentorship expands beyond a primary advisor, and also includes clinical and dissertation advisors (Cobb et al., 2018). This further confounds the proper dissemination of ethical practices to students given not only the variance in ethical standards across institutions, but also variance in RCR teaching provided by mentors or advisors.
Despite the variation, mentoring in psychological training can improve academic productivity, professional competence, and the success of psychology graduate students (Johnson et al., 2000; Williams-Nickelson, 2009, Mangione et al., 2018). This is an opportunity to ensure trainees are well-trained and prepared to engage in best practices for how research is conducted, reported, and applied. Unfortunately, in practice, training in clinical psychology (e.g., scientist-practitioner, practitioner-scholar) does not necessarily promote the direct use of mentors (Clark et al., 2000).
Mentoring challenges in the psychological sciences
Few studies have addressed mentoring for psychology graduate students, but such research has increased in recent years. One area of concern is that graduate students in clinical psychology programs are less likely than those in experimental psychology programs to receive mentoring (Johnson et al., 2000), although clinical psychology Ph.D. students were more likely than Psy.D. students to receive mentorship (Clark et al., 2000). While it is marginally “good news” that 50–70% of graduate students surveyed received mentoring, it is concerning that only about 8% of mentors initiate the relationship (Clark et al., 2000). Even more worrisome, 11% of survey respondents report ethical concerns regarding their mentor or mentor relationship (e.g., research related concerns, sexualized relationships), underscoring the difficulty of addressing ethical training and concerns within the context of the mentor-mentee relationship.
While traditional mentorship meetings are considered the cornerstone of graduate student training, there is a general lack of consensus as to how (i.e., what content, in what format, and with what frequency) mentorship meetings should address ethical practice in science (Institute of Medicine and National Research Council, 2002). Compared to traditional graduate programs, clinical psychology graduate students have the challenge of dividing time between research and clinical training, with students spending less time with their research advisors, and having less exposure to the modeling and teaching of ethical research behaviors (Johnson et al., 2000).
About half of clinical psychology trainees’ time is spent in external clinical settings that often last a year or less; the structure of clinical psychology training models, thus, pose an additional challenge for students to establish meaningful mentorship relationships. This discrepancy may be more pronounced in programs where clinical training is prioritized over research (e.g., professional versus Ph.D. programs).
At least one other aspect of mentoring in the psychological sciences that warrants attention is mental health. Recent research has highlighted the mental health crisis in graduate students (Kemsley, 2017) and clinical psychology students are not immune to these statistics (Rummell, 2015). It is particularly concerning that psychology students often have unaddressed mental health concerns— something that could impact, not only trainees’ overall well-being and productivity (Liu et al., 2019), but also their ability to be effective clinicians. Institutions have a responsibility to promote a culture that communicates to trainees that success in graduate school, including academic productivity (e.g., publications, grants) and clinical proficiency, depends first on their physical and mental health. Providing resources for students to receive mental health treatment when necessary, for example, is a critical factor in maintaining graduate student health and productivity.
Other mentoring challenges in academia
A common challenge to addressing ethical dilemmas in the context of academic mentorship lies in the very nature of the mentor-mentee dynamic. The relationship between mentor and trainee is a power-discrepant one in which mentors have implicit and explicit dominance. Additionally, mentors are often expected (if not required) to financially support their trainees (Gruber et al., 2020). Institutions and mentors alike, promote a culture in which graduate students are expected and rewarded for scholarly contributions above other career-oriented pursuits and, at times, well-being (e.g., taking on additional clinical hours, working long hours/weekends). Paradoxically, the power imbalance between any teacher/mentor and student/trainee, combined with an overarching culture of research productivity, may inadvertently promote unethical behavior. For this reason, mechanisms are needed to critically monitor the mentor-mentee relationship and mitigate risks.
As part of RCR training and mentorship, it is also important to consider multicultural competency and ethical concerns regarding research within diverse populations. This plays out in two ways: (1) the training of underrepresented minority (URM) students and (2) the teaching of multicultural ethics. Certain barriers are currently in place limiting appropriate RCR training through mentorship in these groups. For example, traditional one-on-one mentoring is not equitably available to all graduate students, particularly women and members of URM groups (Holt et al., 2016). Further, a major contributor to the on-going difficulties in recruitment and retention of URM researchers, are barriers posed by institutions themselves (de Dios et al., 2013). Such barriers limit accessibility to diverse mentors in research settings, who are important not only to trainees’ development, but who could also provide services to clients who might benefit from a clinician from an underrepresented population. Additionally, little is known overall about the role of gender and ethnicity in graduate students’ ethics experiences or their impressions of RCR training at an individual or institutional level and further research is warranted (Fisher et al., 2009).
Even though efforts to ameliorate some of the aforementioned barriers are still works in progress, academic advising has been shown to contribute to URM student success (Museus & Ravello, 2010). Specifically, three themes were identified in what made an advising relationship most effective: (1) advisors that showed concern about all aspects of the student’s life extending beyond research; (2) advisors that connected students to resources once they expressed concerns; and (3) advisors that proactively connected students to resources (Museus & Ravello, 2010). These three themes may seem intuitive but are especially critical when advising students of color given that many of these groups are exposed to additional stressors (e.g., microaggressions, discrimination) that increase stress and affect psychological well-being (Miller & Orsillo, 2020). In the same vein, given the changing demographics in the United States, diverse research/clinical teams are imperative in addressing mental health disparities response to this challenge, funding institutions, such as the National Institute of Health and the National Science Foundation, have made concerted efforts in recent years to shift diversity-related issues from the periphery to a central focus and value. Not only is recruitment of participants from URM populations now mandated, but also new efforts are designed to recruit and train underrepresented researchers (Valantine et al., 2016).
Guidelines for Mentoring
Guidelines calling for effective mentoring (e.g., American Psychological Association, 1953) are important, even though such guidelines alone are insufficient to meet this goal, much less ensure effective RCR mentoring. Despite a clear distinction between mentoring and advising in clinical psychology (Cobb et al., 2018), quantifying the strength of mentoring lacks standardization, especially when trying to gauge RCR training. To meet this challenge, Fisher and colleagues proposed scales to measure mentoring, institutional attitudes, and perceived preparedness in RCR training (Fisher et al., 2009). Development and use of such objective, reliable, and valid measures is important in helping to capture differences among programs/institutions and may help guide improved mentoring and accountability.
Another challenge for mentoring guidelines is that the APA code of ethics speaks exclusively to the teaching and training of students via mentors, but the role of students themselves in teaching and training other psychology graduate students is often missed in the discussion. Considering the importance of peers and fellow researchers in providing informal mentorship (Lewinski et al., 2017) and setting the research environment culture, this narrow focus is a clear short-coming. There is room to improve the APA code of ethics by expanding expectations for the training of graduate students to be effective, ethical mentors (Brown, 2016). In doing so, students can perhaps view themselves as members of a community that values ethics, rather than being passive recipients of ethical training from their mentors and institutions.
In addition to having students more actively engaged in peer training, having access to multiple mentors may also be a remedy for the potentially damaging effects of the power-discrepant nature of the mentor-mentee relationship and trainee perceptions of the competitive nature of the laboratory setting (Martinson et al., 2005; Anderson et al., 2007). As with other dimensions of mentoring, multiple mentors can help increase the chance a mentee will hear and experience important messages about good practices and responsible conduct of research. A multiple-mentor model (Gruber et al., 2020), that includes non-evaluative mentors, and even assigned peer mentors, would allow students to discuss topics they might otherwise hesitate to discuss with formal mentors (e.g., what to do when others in the lab are suspected of unethical behaviors, how to navigate situations arising from the power differential between a trainee and mentor such as pressures to publish, or inappropriate demands on time). Support from mentors with differing perspectives and skillsets may also serve to buffer dysfunctional mentoring relationships and increase the likelihood of appropriate whistleblowing (Terry and Ghosh, 2015).
Interestingly, multiple mentorship models are already part of clinical psychology programs in the context of clinical training. Multiple tiers of supervision (which can include primary and secondary clinical supervisors, post-doctoral trainees, and interns) share some similarities with the proposed structure of RCR mentoring and training. However, the goals of clinical and research training do not always align (e.g., in research, standardization of measures is highly valued, but in the clinical context, more precise diagnostic evaluations may depend on personalization). Furthermore, depending on the institution, academic research roles may be perceived as more valuable than clinical roles impeding exploration and acquisition of additional ethical training earlier in student’s careers. Likewise, professional schools in clinical psychology may emphasize clinical careers over academic roles, thereby limiting access to appropriate RCR training. Having access to several mentors who can speak to the ethics and challenges inherent to research and clinical work, particularly when their objectives clash, is particularly important in providing graduate students with a wide range of training to handle the various dilemmas to be faced as an independent clinical psychologist. In short, while mentors, particularly multiple mentors, have a central role in trainee success, there is room to improve mentoring as well as the culture of graduate training programs in the psychological sciences.
Recommendations for Institutions to Strengthen Mentoring and Culture
In light of the limitations of relying solely on traditional classroom-based RCR training, it is imperative that mentors explicitly address ethical issues in a practical and laboratory-relevant fashion. Trainings such as those proposed by Peiffer and colleagues (2011) and Kalichman and Plemmons (2017) are geared towards providing mentors and educators with tools and strategies for engaging trainees in regular RCR discussions. These “train the trainer” curricula have been well received by trainees and mentors. Furthermore, Peiffer and colleagues argue that integrating such routine discussions promotes RCR as a core value of the laboratory and inculcates moral and ethical questioning as an automatic response to ambiguous situations among trainees. In surveying their participants, Peiffer found that trainees and staff both agreed that they could freely report protocol violations and mistakes to their superiors. Overall, such changes in laboratory culture could ultimately translate to reductions in questionable scientific practices.
A positive mentoring relationship is a prerequisite for effective RCR mentoring, but trainees often lack the power to find and maintain such a relationship. Institutions have a responsibility to openly acknowledge the power differential inherent in the mentor-mentee relationship and provide trainees with clear guidance for what to do when ethical dilemmas, particularly in the mentor-mentee relationship, emerge (e.g., seek consultation from other mentors). Each graduate program could provide and maintain guidelines somewhere easily accessible to students, and these resources (e.g., an ombuds office) could be presented at initial onboarding. Further, program-specific committees comprised of students and faculty, could meet annually to discuss these evolving guidelines, as well as, present bi-annual surveys to students to gauge institutional climate.
If mentoring of trainees is important, then it is reasonable to expect that mentors should be evaluated at least in part on their mentoring. The focus of evaluation could nominally include frequency of mentorship meetings, frequency with which the mentor promotes discussion among trainees, and adequacy with which relevant ethical issues are addressed. Such evaluation of the mentor-mentee relationship in graduate programs would serve to hold mentors accountable for covering ethical issues and may even stimulate discussion of ethical issues relevant to a trainee’s particular research environment. It is perhaps worth considering that such evaluations would best be accomplished through a committee outside the department in an effort to promote change, minimize potential backlash, and protect the students. In addition to assessing these mentorship duties, it might also be useful to assess trainee perceptions of their preparedness to address ethical dilemmas in the context of their research environments and their comfort in bringing ethical questions to their supervisors’ and mentors’ attention. Assessments of trainees’ general preparedness, even if results are summarized across multiple students/cohorts, might serve as a barometer for the quality and effectiveness of mentors’ promotion of ethical practices.
Institutions could provide aggregated results of student feedback surveys to faculty members and elicit from mentors proposed means for remediation when applicable.
Finally, in addition to the significant role of mentorship in shaping trainee perceptions of appropriate scientific practices, the role of the institution to establish and promote the overall scientific culture must also be acknowledged. Unlike mentorship which can be variable and unpredictable, the institutional climate can be thought of us a constant and necessary backdrop for consistent and effective training (Fisher et al., 2009). To effectively enforce RCR discussions in the context of the mentor-mentee relationship, institutions could: 1) promote a multi-mentor structure (e.g., resources and strategies for obtaining non-evaluative mentors or identifying others that might serve as professional role models); 2) formally describe expectations for mentors including descriptions of how to engage mentees in RCR discussions; 3) offer or perhaps require mentor discussion-based training, including problem solving, laboratory-relevant scenarios and dilemmas; and 4) develop and conduct formal evaluations of mentorship effectiveness in promoting ethics-relevant discussion among trainees. These steps to strengthen mentoring would arguably make a significant impact on the culture of ethics in clinical psychological sciences.
Research in the psychological sciences has a profound impact on every aspect of our lives. The quality of that research depends on robust training and mentoring for the next generation of scientists. This article recommends several strategies institutions can implement to strengthen mentoring in the responsible conduct of research.
Preparation of this manuscript was supported in part by grants from the National Institutes of Health (NCATS/NIH UL1TR001442) and the National Science Foundation (CCE STEM #1835029). The opinions expressed in this article are the authors’ own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, the National Science Foundation, or the United States government.