Introduction
The ongoing lack of diversity in research participation limits our ability to progress toward health equity, as the very communities experiencing health disparities are those underrepresented in research [Reference Varma, Jones, Oladele and Miller1,Reference Boulware, Corbie and Aguilar-Gaxiola2]. While there have been many explorations of the problem, most have focused on a lack of trust in research on the part of communities experiencing health disparities, focusing on the problem outside of the research community [Reference Knepper and McLeod3]. However, recent attention has been paid to the barriers to diversity in research participation borne of researcher behaviors and practices. For example, recent publications from Wilkins, Manly, et al. and Gilmore-Bykovskyi et al. seek to understand, problematize, and, ultimately, reduce the use of “exclusionary research practices” in research [Reference Wilkins4–Reference Gilmore-Bykovskyi, Jackson and Wilkins6]. Such practices can include but are not limited to lack of attention to the needs of potential participants from a range of backgrounds, withholding opportunities to participate in research from individuals who are not perceived as “good” participants, using needlessly narrow eligibility requirements that affect groups disproportionately (e.g., lack of comorbidities) or setting all recruitment/enrollment efforts in places where there is an existing lack of diversity such as academic health centers [Reference Boulware, Corbie and Aguilar-Gaxiola2,Reference Gilmore-Bykovskyi, Jackson and Wilkins6,Reference Niranjan, Martin and Fouad7]. Our own mixed-methods exploration of attitudes of clinical investigators revealed a high level of recognition of the problem but a low level of actual implementation of inclusive practices [Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8,Reference Passmore, Farrar Edwards, Sorkness, Esmond and Brasier9]. Interviews revealed ideas about who makes a “good” research participant as well as the perception of the problem as unsolvable. These findings, as do those of Niranjan et al. and others, indicate a need for researcher education [Reference Niranjan, Durant and Wenzel10–Reference Boden-Albala, Waddy, Appleton, Kuczynski, Nangle and Parikh12].
With the identification of researcher barriers, several educational programs have been developed to build capacity for inclusive research engagement. An early example of this was the Building Trust between Minorities and Researchers program (2012–2021), which consisted of a two-day workshop focused on raising awareness of research abuse and participant perspectives of mistrust [Reference Hartnett13]. Another recent effort is a massive open online course called Faster Together, Enhancing the Recruitment of Marginalized Communities in Clinical Trials [Reference Kusnoor, Villalta-Gil and Michaels14]. Faster Together is designed for a general audience involved in health-related research recruitment in clinical or community settings. Eight modules featuring videos, text, and quizzes are offered for an audience to move through at their own pace. A study evaluating the Faster Together course demonstrated increased knowledge and intention to change research practices based on pre- and post-test data collected in the first 10 months since the course was released. While 382 individuals enrolled in the course during the first ten months, only 105 participants completed the pre-test, and 14 completed the post-test. Another program is a certification program for clinical research coordinators that includes a series of 1–2 hour sessions, including Just Ask: Equity and Diversity in Clinical Research, as well as more general recruitment educational sessions, such as Using Social Marketing Principles to Design Your Engagement Strategy [Reference Cranfill, Freel and Deeter15]. Evaluation of the Duke certificate program has revealed impressive outcomes, including self-reported competency and manager-reported skill level increases.
In the article, we report on preliminary outcomes from yet another program, the Just Research workshop, established at the end of 2022. Like the programs discussed above, Just Research seeks to promote the capacity for the engagement of participants from groups underrepresented in research. Data collection has included a retrospective pre/post-test and follow-up surveys.
Methods and materials
Just Research workshop
This workshop was developed at the University of Wisconsin–Madison based on our assessment of barriers to inclusive practices on our campus among investigators [Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8] and staff and our experience as members of the Building Trust team [Reference Hartnett13]. Just Research promotes holistic, intentional approaches that prioritize inclusivity by focusing on (1) bias recognition and reduction, (2) the incorporation of community perspectives in research design and procedures, (3) engagement best practices, and (4) critical practice with translating new skills into potential changes in practice relative to each participant’s work [Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8,Reference Passmore, Farrar Edwards, Sorkness, Esmond and Brasier9]. Workshop learning objectives are to understand opportunities within the academic research community to promote diversity in research participation and engagement; to identify principles that guide responsible, respectful, and sustainable engagement; to explore avenues to increase community voice in research participation based on engagement and recruitment science; and to develop practical ways to integrate new skills into one’s work. Through interactive exercises based on real-world experiences and opportunities to explore ways to tailor best practices to their individual needs, Just Research participants can explore researcher “trustworthiness” and build practical strategies to promote inclusivity [Reference Wilkins4,Reference Gilmore-Bykovskyi, Jackson and Wilkins6,Reference Passmore, Gerbitz and Hancock16,Reference Passmore, Fryer, Butler, Garza, Thomas and Quinn17]. The program is available for investigators and research staff (teams are encouraged to participate together). Workshops are full-day, face-to-face, and kept small (∼18–22 participants) to facilitate discussion and activities. To date, we have provided four workshops (four cohorts) in October 2022, February, June, and August of 2023. All have been jointly facilitated by the program’s co-directors (GG-H and SRP), who have considerable experience in academic/community communication and partnership. All participants included in the evaluation reported here received the same curriculum in the same format. To enroll in the workshop, participants self-selected and responded to an open registration link distributed across the University of Wisconsin (UW) Institute for Clinical and Translational Research (ICTR) affiliates and promoted on the ICTR website and social media platforms. Some participants signed up as a team; others registered as individuals.
Just Research evaluation
As our previous work had indicated the lack of self-efficacy regarding inclusive research practices as a barrier [Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8], the Just Research evaluation is specifically designed to address this outcome along with knowledge and skills. We collect both process and outcome-level data, including information on participants’ roles in research, years of research, and experience in clinical research. Surveys are optional and were distributed immediately following the workshop. Participants received three reminders over the following weeks. We tracked outcomes through a retrospective pre-post survey to assess participants’ self-reported attitudes, skills, and self-efficacy on Likert-scale items after the session. Retrospective pre-tests are helpful to avoid a response-shift basis, which can occur when participants’ frames regarding a topic so shift from the experience of the training itself that traditional pre-test/post-test design can essentially measure different phenomena (e.g., participants overestimate knowledge and skills on traditional pre-test and underestimate at post-test after realizing that the topic more complex than previously believed) [Reference Thomas, Wells and Baumann18–Reference Howard, Ralph, Gulanick, Maxwell, Nance and Gerber20]. Main outcomes for pre-post comparison centered around specific inclusive behaviors, including the recognition of bias on the research team, taking action to correct the bias, understanding the importance of community perspective, seeking out community perspectives through partnerships or other relationships, and general ability to integrate inclusive practices (e.g., maximize recruitment through a range of venues), which reflect curriculum objectives. Ongoing change and adoption of inclusive behaviors (e.g., implementation of community-engaged strategies, review of exclusion criteria, and recruitment plans to maximize diverse participation) were assessed at the 9–12-month follow-up with a brief survey also using Likert-scale items and some open-ended questions. All surveys were pilot tested with 8–10 investigators and research staff to ensure that participants clearly and correctly interpreted items. This could only be assessed from cohorts 1 and 2, who participated in the workshop in October 2022 and February 2023 (other cohorts will complete follow-up surveys later in 2024). We created a composite scale of self-assessed knowledge, skills, and self-efficacy from five items (see Figure 1) in the pre- and post-tests in IBM SSPS to allow for the comparison. Reliability was assessed using Cronbach’s alpha coefficient (α 0.85 and 0.98, respectively). Scale scores were compared using a two-tailed t-test. Due to the limited number of follow-up surveys to date, we only present descriptive statistics. All study activities have been reviewed and approved by the University of Wisconsin IRB (ID: 2024-0053)
Figure 1. Participant mean scores on individual items – pre- and post-tests. (n = 70).
Results
In all, 78 investigators and staff participated in 4 workshops. Of these, 70 provided pre/post-data, and 21 participated in follow-up surveys (distributed 9–12 months post-workshop). Table 1 illustrates the characteristics of our 70 workshop participants who have participated in the pre/post-test to date. The majority are affiliated with the School of Medicine and Public Health and identified themselves as research staff. However, we did have participation from investigators as well (23%).
Table 1. Survey participant characteristics (n = 70)
Participants were asked five key questions designed to assess their level of self-assessed knowledge, skills, and self-efficacy. Across all these items, we found improvement in the comparison of means pre- and post-test (see Figure 1).
In comparing retrospective pre/post-test knowledge, skill, and self-efficacy scales using paired sample t-tests, we find a modest but statistically significant difference across all items (t −4.266, SD 0.96504, p < .001). A relatively high standard deviation for each item indicates the wide range of participant skills, knowledge, and self-efficacy both before and after the workshop.
Post-tests included open-ended items requesting workshop feedback, which was largely positive. For example,
It was a wonderful experience, and I will be recommending it to the rest of my research team.
And,
It should be required for people doing human subjects research.
The 9–12-month follow-up survey was completed by 21 of the 38 (55% response rate) participants in cohorts 1 and 2, as they were the only cohorts who completed the program in the follow-up timeframe. There is no reason to believe that these cohorts are unique in comparison to the following cohorts. Data on self-reported inclusive research behaviors tied to their participation in a Just Research workshop is presented in Table 2.
Table 2. Follow-up survey response at 9–12 months post-session (n = 21)
76.2% of participants in the follow-up survey attributed changes in their work as a direct result of their participation in Just Research. In response to an open-ended question about the specifics of such changes, participants reported increased awareness of bias for themselves and, notably, their teams. For some, participation led to more open communication regarding inclusion on the research team. For example,
From my position as a Research Program Assistant, I find myself more comfortable with questioning PI practices and making suggestions for how to open the work to community partners.
Some participants also reported the establishment of new partnerships or community advisory boards. For example,
I have begun making meaningful and bidirectional relationships with members and groups associated with my field of study. Instead of engaging them when I need something, I shifted my focus to curating genuine friendships and relationships.
And,
[We have] more intentional engagement and collaboration to serve historically excluded and underrepresented communities.
Others reported changes in recruitment or other research practices. For example,
I’m now offering recruitment flyers at community locations outside healthcare clinics and have spoken with stakeholders about ways to improve my recruitment practices.
And,
Thoughtfully read through consent to try to ensure that participants are informed. Made modifications to consent and subject facing material if it was not clear or misleading.
Discussion
The path to increasing diversity in research participation is multifaceted and will necessarily include shifts in the attitudes and behaviors of members of our research community. Intentionally or not, there are many ways we, as researchers, “get in our own way” [Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8,Reference Niranjan, Durant and Wenzel10,Reference St. John, Hickey and Kastern21]. For example, driven by traditional approaches that govern eligibility criteria, recruitment sites, and ideas about who is willing and able to participate in research, we may limit diversity in participation [Reference Manly, Gilmore-Bykovskyi and Deters5–Reference Passmore, Kisicki, Gilmore-Bykovskyi, Green-Harris and Edwards8]. Shifting this culture is complex and somewhat uncharted territory, requiring careful assessment and evaluation. Our findings indicate that the program can result in modest but positive short-term outcomes. However, it is essential to note that mean scores on individual items presented in Figure 1 were relatively low, indicating that participants may need additional training. We are most encouraged by mid-term (9–12 months) follow-up findings that indicate behavior changes to promote inclusion. Despite the limitations of our small sample size, our findings that more than 76% of participants reported changes in their research practice due to their experience in the workshop (including increased action responding to bias and racism) are most meaningful and promising. Regarding the 24% of participants who did not report a change, we, unfortunately, know little about their experience as there were no open-ended responses from that group. We are, however, adding six-month follow-up interviews to our evaluation to explore this issue further.
Ultimately, no one intervention or educational program will achieve the goal of meaningful diversity in research participation. Indeed, as others have pointed out, there is a likely need for intervention at various levels and audiences [Reference Boden-Albala, Carman and Southwick11,Reference Niranjan, Durant and Wenzel10]. Just Research is only a small part of this effort, but we are fortunate to be among others working to experiment with various approaches to researcher education [Reference Hartnett13–Reference Cranfill, Freel and Deeter15]. Indeed, all these efforts have data demonstrating that change is possible, and perhaps more importantly, there is a willing audience for these interventions. We hope Just Research and our evidence base will contribute to ensuring that research is representative of patient populations and has the necessary diversity of participation to address health disparities.
Limitations
As is the case with all studies, there are limitations to our work that are important to note. These include our relatively small sample size to date and reliance on self-report. In addition, our sample is dramatically skewed regarding gender (88.6%) identifying as women (88.6%). We will be working on understanding and correcting this imbalance as the program develops if it does not accurately represent the gender distribution of members of clinical research teams more broadly. It is important to note that the distribution in terms of race/ethnicity in our sample is not dissimilar to that of research teams. The majority of our participants identified as White (70%), which accurately reflects the characteristics of faculty and noninstructional academic staff at UW–Madison, which was 69.9% White in 2022 [22]. Interestingly, we saw an overrepresentation of participants identifying as Black or African American compared to the campus-wide faculty and nonacademic staff population (18.6% vs. 2.6%) [22]. The dramatic gender imbalance of the participants is less well understood. As workshop participants either self-select or are encouraged by other team members, it would seem that there is a striking gender imbalance either on research teams or in interest regarding inclusive research practices. This is a topic for future research which we are anxious to pursue. However, it is beyond the scope of this initial attempt to present Just Research outcomes and preliminary data.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/cts.2024.592
Author contributions
Passmore, Green-Harris, Kirschmann, Phelps, Garcia, and Farrar-Edwards contributed to the work’s conception and design. Passmore, Green-Harris, and Filut contributed to the data collection and analysis. All authors made critical intellectual contributions and participated in the manuscript’s drafting.
Funding statement
The 2024 Just Research workshops are made possible by support from the UW Institute of Diversity Science and ICTR, which, in turn, supported by a NIH-NCATS Clinical and Translational Science Award (CTSA 1UL1TR002373) and a grant from the Wisconsin Partnership Program (WPP 5132). Dr Passmore also receives support from the UW School of Nursing and the Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation.
Competing interests
None.
Open access
