Mental FLOSS for faculty: effects of an inter-institutional community of practice on faculty mindset and inclusive teaching in higher education

ABSTRACT

Upon entering college, students from marginalized identities are equally as likely as their peers to declare an interest in science, technology, engineering, and math (STEM). However, many of these students leave STEM before graduation, leading to underrepresentation in academia and the workforce. In the Faculty Learning Optimizes Student Success (FLOSS) Program, we sought to reduce this disparity by introducing faculty to the diversity of needs and perspectives of their students, educating them on best practices for inclusive teaching, and fostering a community of practice to support implementation of these practices. We assessed the program by surveying faculty before and after participation on their (i) role and responsibilities in building classroom equity and inclusion, (ii) knowledge and confidence to build inclusive learning environments, and (iii) sense of belonging within a teaching community. As a result of participating in FLOSS, faculty were less likely to assign deficits to students’ individual characteristics (e.g., academic preparation), and more likely to recognize deficits in their own teaching and/or the academic environment. Faculty also reported increases in their knowledge of inclusive teaching strategies and their confidence in implementing them, as well as a strong sense of belonging to a teaching community.

INTRODUCTION

STEM (Science, Technology, Engineering, and Math) education is of paramount importance in the United States, serving as a cornerstone for innovation, economic growth, and global competitiveness. STEM disciplines provide the skills and knowledge necessary to address complex challenges, ranging from healthcare and environmental sustainability to cybersecurity and space exploration. With technological advancements driving rapid changes across industries, a strong foundation in the STEM disciplines is essential for preparing the workforce of tomorrow. This can only be achieved through the establishment of a strong STEM education-based school-to-career pipeline, where individuals are equipped with the tools to succeed in a dynamic and interconnected world and the continued fostering of a dynamic STEM workforce engaged in research, development, and practice (1). These efforts are ongoing, with the ultimate goal to drive societal progress and prosperity.

While we continue to work on growing and strengthening the STEM academic-to-career pipeline, it is imperative that we also further diversify it. An analysis of 1.2 million US PhD recipients over the past 30 years found that students from demographically minoritized groups demonstrated higher rates of novelty and innovation in their dissertation work compared to peers with majority identities (2). Similarly, modern science is increasingly dominated by teams over solo researchers (3), with diverse teams shown to produce more novel results than homogeneous teams (4). Moreover, the lack of diversity in some fields can lead to potentially serious consequences. For example, pulse oximeters, skin temperature sensors, and other critical medical diagnostic tools were developed and calibrated on White skin, ultimately rendering them less effective on patients with darker skin tones (5–7). Furthermore, artificial intelligence (AI) detectors are more likely to falsely implicate the work of non-English speakers (8), while facial recognition programs struggle with non-White faces—either failing to accurately match them (e.g., reference 9) or even failing to detect and recognize them as faces at all (10). These and numerous other examples stress the importance of fostering broader representation and inclusivity in STEM education and professions.

Despite the need to further diversify STEM being well understood, in the United States, the number of underrepresented minorities (URM) occupying professional STEM positions does not align with their representation in the overall workforce (11). While this issue varies within subcategories of STEM jobs, data show that there are areas where these differences are significant. For example, women represent half of the U.S. workforce but only one-third of the STEM workforce (12). Furthermore, URM collectively holds 25% of all STEM jobs compared to one-third of jobs overall (13), with approximately 10% working professionally within the sciences or engineering (1).

We can link these diversity-related trends to the representation, or lack thereof, that we see in STEM disciplines at U.S. colleges and universities. The number of URM who graduate with a bachelor’s degree in a STEM discipline is significantly lower than their representation within the population. The underrepresentation is even greater when considering the number of URM earning a doctorate (1). The disparity between the representation of URM graduating with STEM degrees and moving into related fields does not seem to be due to a lack of interest, however. Instead, the issue can be more broadly attributed to a leaky STEM academic (i.e., education to career) pipeline (14) and systemic barriers that continue to limit diversity within the STEM labor force (15).

Upon entering college, students from marginalized identities are equally as likely as their non-marginalized student peers to declare an interest in STEM (16). Different rates of persistence in these majors begin to develop over time (17–19); however, this results in significantly fewer members of underrepresented groups completing STEM degrees and entering relevant fields. The disproportionately low persistence rates of underrepresented students majoring in STEM disciplines have been attributed to a number of factors, such as under-preparation (20), lack of support from family or other social groups (21), or a lack of a STEM identity or sense of belonging (21).

While URM students can and do encounter challenges during their K-12 experiences that could ultimately guide them out of the STEM academic pipeline, Hrabowski (1) suggests that more good may be done toward retaining and growing the numbers of college-educated students filling STEM-related jobs, at least in the short term, if we focus our attention on making transformative change at the secondary level. In addition, he suggests that interventions that target students early on (i.e., when they are taking introductory science courses) and that are designed to bolster the academic performance of all students while aiming to be inclusive in their design are key. The initiative that we introduce in the current study meets these criteria.

In our study, we targeted the connection between faculty and students within the classroom environment, a setting in which numerous interventions designed to promote academic success in STEM students have been situated (e.g., reference 22. While it may not be the only variable at work, it is well established that student performance in the classroom can be a function of instructor approach and mindset (23, 24). Numerous studies have presented data highlighting a disparity in grades and test scores between minoritized students and their peers in STEM courses (e.g., references 25, 26). These disparities are common in the large-enrollment introductory classes typical of STEM majors (e.g.,references 27–29), which are also critical points in a student’s decision to continue in STEM (30, 31). Students report that introductory STEM courses “obscure the subject, diminish students’ confidence, and discourage them from becoming scientists” (17). In addition, students from underrepresented backgrounds are disproportionately less likely to persist in STEM if they receive low or failing grades in introductory courses (32). For example, Harris et al. (31) found that women, first-generation students, underrepresented minorities, and students from low socioeconomic backgrounds received lower grades in introductory chemistry compared to their major-group peers, and if they received below a C- in the course, they were less likely to retake the class in an attempt to earn a higher score. This typically meant that these students were unable to continue in those majors.

In response to these well-known opportunity gaps (see references 33, 34 for why we avoid using the term “achievement gap”), educators have devoted significant time and effort toward identifying and implementing remedies designed to reduce these score differentials. However, these efforts have often taken place through a lens of student deficit, where the focus is on establishing mechanisms by which individual students or groups of students can raise their performance. Students are encouraged to adopt a growth mindset (35) and to cultivate “grit” (36) to persist through the challenges they encounter. Programs are developed to provide additional support for students with lower levels of academic preparation and to teach study skills. While these are important initiatives and can lead to successful outcomes, they are focused on treating symptoms rather than addressing the causes. Categorizing a student as “underprepared” suggests that any problems that need to be addressed lie with the student, who needs to be “fixed” rather than recognizing that our students arrive in our classrooms from a wide variety of academic backgrounds, thus requiring that we adjust our teaching accordingly.

To effect greater and more lasting change, it is critical to work on identifying and remedying systemic barriers and reimagining the culture and environment in which our students are learning. One factor in this change is to shift our mindset away from viewing the student as the problem and toward a focus on identifying where the “deficit” lies with respect to the broader academic culture and in how we teach. Students from marginalized backgrounds can be subject to stereotype threat, in which they feel added pressure and anxiety from the risk of conforming to negative stereotypes about their race, gender, or other aspects of their identity (21, 37). They may also face unconscious or implicit bias, which can affect how they are perceived by their peers (38), the grades they are assigned (39, 40), and eventually contribute to cycles where implicit stereotypes lead to differential participation/results, which then reinforce the stereotype (41, 42). Common grading and assessment practices, such as curving scores and basing grades on only a few high-stakes assessments, are rooted in historical practice rather than accurately representing students’ knowledge (43). As such, the ability to achieve high scores can be linked to succeeding in the system, rather than mastering the content, which disadvantages both first-generation students who may lack knowledge of the “hidden curriculum” necessary for university success, as well as students who are unaccounted for in the existing system (e.g., due to neurodiversity, disabilities, or other reasons).

These factors, among others, can contribute to creating a classroom culture where minoritized students do not feel a sense of belonging, which can raise doubt in their own STEM identities and affect persistence in STEM (44, 45). In addition to improving the learning environment, redesigning instructional methods to be more inclusive and to promote engagement has been shown to significantly affect outcomes for minoritized students. For example, implementing active learning reduces failure rates and improves academic outcomes for all students (46), but disproportionately benefits those from underrepresented groups (26, 27, 47). Alternative grading schemes such as specifications grading (48) or ungrading (49) hold promise for more clearly linking grades to mastering content, rather than mastering the hidden curriculum or being subject to instructor bias (unconscious or otherwise).

Another commonly cited example of an inclusive teaching practice is the concept of a growth mindset. Proponents of a growth mindset frequently focus on the role of student mindset and cultivating student’s belief in their ability to succeed (e.g., reference 50). However, the mindset of the instructor can be equally important. STEM courses taught by professors who held fixed beliefs about student ability had racial “achievement gaps” that were twice as large as those taught by faculty who held growth mindsets (51). If instructors believe (consciously or not) that some of their students are simply less likely to succeed, that attitude will affect how they teach and interact with their students.

The Faculty Learning Opportunities for Student Success (FLOSS) program described in this study was developed to address the effect of these broader issues of classroom climate, course structure, and instructor mindset/approach on faculty instructional approach, course design, and sense of self. As a professional development program, FLOSS was distinct in bringing together biology faculty from three different institutions of higher education—a community college; a primarily undergraduate, masters-granting university; and a large R1 institution. These three institutions serve unique roles in the diverse and comprehensive higher education landscape found within the San Francisco Bay Area, but intersect in the diversity of their student bodies and in seeking to identify and address obstacles to their students’ success. This semester-long program was designed to strengthen faculty understanding of the diversity of needs and perspectives of their students, introduce strategies and best practices for inclusive teaching, and foster an intersegmental community of practice to support implementation of these approaches. As a community of practice, faculty were joined by a common interest in teaching undergraduate biology, and they regularly engaged and worked together as a cohort in FLOSS workshops. Program workshops were structured to first identify the systemic barriers that their students are facing and then to explore evidence-based pedagogical interventions.

In this initial iteration of the FLOSS program, we sought to assess the impact of the program on faculty instruction, first with regard to shifting faculty mindset about their role and responsibility in building equity and inclusion in their classrooms. Since it is critical to transform mindset into practice to effect change, we also assessed whether the program provided essential knowledge and confidence in inclusive teaching methods, as well as whether the community of practice model provided faculty the necessary support for their efforts toward implementing pedagogical innovation.

METHODS

FLOSS program

The FLOSS program was developed out of a shared recognition among the three participating institutions that opportunity gaps persist among their biology students and that there was mutual interest in creating resources specific to biology instructors for addressing those gaps through inclusive teaching practices. Once the program team was established, we released a needs assessment survey to all biology faculty at our institutions to establish (i) the major challenges they faced in teaching undergraduate biology, (ii) their desired learning priorities for a teaching community of practice, and (iii) the concerns they had about participating in this faculty development program. The results from this needs assessment were used to shape the structure and content of the FLOSS program.

Workshops and community of practice

The FLOSS program included two key elements: a series of eight workshops that cover topics in inclusive teaching (Table 1), and a community of practice structure that involves faculty participants actively engaging with each other during these workshops to share their stories and experiences, seek and receive feedback, and exchange resources. Workshops were held synchronously over Zoom for 2 hours approximately every other week during the course of a 14 week semester. The first cohort of participants took place from January to June 2021 (“Spring 2021”), and the second cohort took place from September 2021 to January 2022 (“Fall 2021”).

TABLE 1.

FLOSS workshop overview for the second cohort (Fall 2021)^a

Workshop	Topic
1. Getting to Know Your Students	Explore campus data resources to become familiar with student demographics and need
2. Teaching to a Diverse Student Body	Exploration of diversity’s dimensionality
3. Building Inclusive Classrooms	Identifying equal access barriers to classroom opportunities and resources
4. Metacognition	An examination of how students learn and methods to support them
5. Introduction to Active Learning	Active learning techniques and their contribution to student success
6. Assessment and Grading	An examination of the goals of assessment and how to meet them
7. Backwards Design	Techniques for effective curricula and assessment planning
8. Peer Feedback and Discussion	Incorporating previous discussions into existing or new lesson plans

^a For the first cohort (Spring 2021), workshop 6 featured “adaptive learning,” which was replaced due to participant feedback.

In addition to the workshops, participants were enrolled in a companion Canvas course that provided additional resources, materials, and discussion boards for reflection between workshops. A public version of this Canvas course is available at https://canvas.instructure.com/courses/7434274 and includes the lesson plans to facilitate each workshop.

Participants

Participants in the FLOSS program were recruited from biology faculty who held professor or lecturer status at the three partner institutions: a community college, a primarily undergraduate masters-granting university institution with a small number of masters students, and a large R1 institution, all within the San Francisco East Bay region. Participants were recruited through broadly circulated email advertisements, announcements made during community/department functions, and directly through word-of-mouth and personal recruitment by members of the FLOSS team.

Participation in the program was voluntary, and participants who attended at least six of the eight workshops were awarded a $1,000 stipend at the end of the program. Both the Spring 2021 and Fall 2021 cohorts included faculty from all three institutions, as well as both new faculty and tenured faculty with over 20 years of experience. The Spring cohort had 10 participants, and the Fall cohort had nine. As the faculty in both cohorts had similar demographics and participated in a nearly identical program, both cohorts were combined for analysis.

Survey

Assessment of our 2020–2022 FLOSS program was driven by analysis of pre/post surveys completed by all faculty participants. Surveys consisted of questions within the following main themes regarding the participants and their teaching experience: (i) skill or knowledge in teaching/pedagogical approaches, (ii) sense of support in implementing pedagogical change, (iii) sense of being a part of a teaching community, (iv) knowledge about diversity, equity, and inclusion (DEI) in the classroom, (v) their perspective regarding student growth mindset, (vi) their teaching style and approach for each class they currently teach, and (vii) personal demographics. All questions within these themes were present in the pre- and post-survey, allowing for direct comparison between the participants’ answers before and after completing the FLOSS program. Most questions were taken directly from published and/or validated survey tools, then compiled into the main themes. Survey tools consisted of both Likert-style and open-ended question types. These survey tools will be cited as we introduce each survey question.

Pre-surveys were distributed via Qualtrics and completed during the week immediately prior to the first workshop. Post-surveys were distributed via Qualtrics and completed during the week immediately following the last workshop.

Analysis

Qualitative analysis

To understand whether participation in FLOSS had impacted participants’ recognition and understanding of their role regarding diversity, equity, and inclusion within their classrooms, we coded responses to two open-ended survey questions. Responses were pooled between each program cohort (19 participants total); however, each survey question was optional, such that the sample size may be slightly smaller for any one question. Results in the pre-/post-domain are normalized by the number of responses in that category.

Because of the breadth of each question, it was not immediately clear what the appropriate granularity of coding the responses would be. We used bottom-up coding to identify the themes and coding for each question. Researcher 1 labeled the data according to specific themes (e.g., an explicit or implicit response), then organized these themes in terms of broader themes (e.g., the “Importance” of the topic, indicated by either an implicit or explicit response). Researcher 1 then drafted a codebook detailing these specialized and broad themes. Researcher 1 and Researcher 2 reviewed the codebook together, fine-tuning the descriptions of the themes. Researchers 1 and 2 subsequently coded five responses to each question independently, then reviewed and confirmed the codebook further to a finalized state based on this subset. Next, Researcher 1 coded all responses to the two qualitative survey questions, while Researcher 2 coded a subset of five responses. The average Cohen’s kappa value for the five interviews coded by both Researchers was 0.70, which indicates a substantial level of agreement (52). Lastly, any uncertainties within all coded responses were reviewed among Researchers 1 and 2. The data were discussed until consensus was met on whether the response in question should be coded present or absent.

Survey Question 1: Importance and action taken by faculty regarding equity in the classroom

To identify participants’ awareness of the impact equity could play in their teaching decisions, we asked, “How does equity (addressing the unique traits of individuals towards creating a fair and impartial learning environment) come to play in your teaching?” This survey item was generated internally by our team while developing the survey instrument. The question was created to be a direct indicator of an individual’s perspective(s) of the role of equity in their teaching. Each response was coded into one of two primary themes: “Importance,” suggesting a recognition of the importance of equity in teaching, or “Action,” which included a statement of actions that one could, has, or will take regarding equity in the classroom. In this case, we were identifying whether participants volunteered descriptions of what impact equity had on teaching decisions.

Codes for “Importance” included “Explicit,” “Implicit,” and “None.” Each code was exclusive. Coding a response as “Explicit” indicated the response stated clearly that equity is important and/or a priority. An example of such a response was “equity is very important in my teaching.” “Implicit” coded responses included responses where equity was considered, and how to improve it in the classroom, without an explicit “equity is important” statement. An example of such a response was “I try to make sure everyone feels they belong in the group despite differences in background. I also try to include as much foundation as possible (however quickly) or provide additional resources for it.” Responses were coded as “None” when there was no mention of equity, or statements dismissing a connection between equity and their teaching. An example of “None” was “Generally, it doesn’t. But I do try to keep communication channels open and ask students to come to me with any challenges that are in the way of their learning.” There was no hierarchy between “Explicit” and “Implicit” responses, but it is important to note that with the goal of improving equity in the classroom, implicit statements of importance were not necessarily accompanied by concrete impact and/or action.

Codes for “Action” included “Specific,” “General,” and “No Statement.” Each code was exclusive and there was no case of potential overlap with the Importance theme in the data. “Specific” coded responses stated a specific action that has happened or that they intend to try regarding equity in their teaching. An example of such a response was “I spend quite a lot of time emailing students, especially those who are not turning in Exit Tickets, asking if they are OK, what’s up. Listening is fine, but you have to be proactive if you want to hear things worth listening to, for example, what is going on in their lives that is making it hard to concentrate and participate. That sort of thing…” Coding a response as “General” indicated that it included a statement of non-personal actions. They might reference a broad societal importance of an action, or even suggest what should happen, but in a general capacity. An example of “General” was “Very important. Reaching out to students, extra office hours, being responsive to emails are all critical.” Responses were coded as “No Statement” when there was no statement of actions. An example of “No Statement” was “The material for the… class I teach is complex and extensive. The study body is widely variable; some students have had unlimited resources to succeed in learning these topics; others have not.” It was not necessarily expected that respondents would include a statement of action, given that it was not an explicit request in the question, so there was no hierarchy between these codes.

Survey Question 2: Faculty perspective on a potential deficit in equity in a classroom—Is it due to students or their environment?

To identify participants’ awareness of the DEI issues their students may face in a classroom, we asked, “When considering diversity, equity, and inclusion, what are the primary issues that your students face in any classroom?” This survey item was generated internally by our team while developing the survey instrument. The goal of the question was to identify an individual’s knowledge and viewpoints of their students and the barriers they face. More specifically, the primary research question guiding this item’s coding process was whether the deficit in equity was assigned to students or nonstudents. Thus, the primary theme was broken into two codes: “Student” where equity issues were assigned to some characteristic the student is bringing to the class, typically background or under-preparation, or “Other” where issues were assigned to factors outside of the student’s control. These two binary categories were coded non-exclusively, so it is possible for one response to be coded as true for each category. An example of a response that was coded as both “Student” and “Other” was “I think some students are less prepared and other[s] have financial challenges/family responsibilities.” Here, the response mentions students as less prepared, which is coded as “Student,” and also refers to financial challenges and/or family responsibilities which is coded as “Other.” A response that was coded only as “Other” was “belonging, assumptions made by other students, teaching too much material, too many terms…” There were no pre- or post-survey responses that included only “Student” in our coding.

It became clear that there were distinct subcategories in the data for Survey Question 2, so the “Other” category was broken down into subthemes. “General” was prescribed to general statements of equity issues without mentioning a specific source. “Institution” coded responses included mentions of inequities in access to resources at the institutional/department level. Responses coded as “Responsibilities” included external factors that reduce the student’s time/energy resources that can otherwise be dedicated to learning. “Instructor” coded responses included factors relating to the classroom and instructor as the source of a (the) deficit. Each code was nonexclusive, such that more than one of the sub-themes could apply to each response. We primarily found overlap between the primary “Student” and “Other” themes, but we also found a few occasions of overlap among the four subthemes.

An example of a response that was coded for multiple subthemes was “...Students under severe financial or caregiver constraints may not have enough time to devote to their studies. Students with disabilities may struggle with the way the information is presented or assessments administered,” which was both “Responsibilities” and “Instructor.” An example of a response that was coded as “Institution” was “Not feeling like they belong, that the ’system' is designed around them.” And finally, an example that was coded as “General” was “Not all students' voices, backgrounds, and personal histories are equally valued.”

Quantitative analysis

Three Likert questions on the pre-post survey were analyzed to determine if the FLOSS program was successful in (i) cultivating knowledge and confidence in faculty to build inclusive learning classrooms and (ii) fostering a community of practice among faculty participants.

Survey Questions 3 and 4: Knowledge and confidence in inclusive teaching practices

To assess knowledge and confidence, we compiled the 19 participants’ pre- and post-responses to two survey questions. Survey Question 3 asked: “Please indicate your skill or knowledge in the following areas:” with 8 pedagogical areas listed, each with a 6-ranked Likert scale (None = 1, Very Low = 2, Low = 3, Average = 4, High = 5, and Very High = 6). This survey item was used from a previously published instrument that assessed skill and knowledge of pedagogical areas among biology graduate students and post-docs after a teaching fellows program (53). The pedagogical areas listed were aligned with our FLOSS workshop curriculum’s learning goals such that each faculty should have gained knowledge of each, assuming we were successful. Survey Question 4 was “I do not feel that I have enough knowledge to implement inclusive teaching” with a 7-ranked Likert scale (Strongly disagree = 1, Disagree = 2, Somewhat disagree = 3, Neither agree nor disagree = 4, Somewhat agree = 5, Agree = 6, and Strongly agree = 7). This survey item was pulled from a previously published instrument that assessed skill and knowledge of inclusive teaching among college science instructors after participating in a summer institute (54). Both Survey Question 3 and Survey Question 4 were viable and representative metrics for our FLOSS participant population, as they were designed for higher education science instructors participating in a teaching-focused professional development program. For each survey question’s analysis, we created individual data frames for pre- and post-responses and then tested the exact permutation from the mean among each group using a paired Fisher’s exact test.

Survey Question 5: Formation of a community of practice

To assess whether FLOSS was successful in fostering a community of practice among the participants, we compared pre- and post-responses to Survey Question 5: “I feel I am part of a community of teachers who are interested in active learning” with a 7-ranked Likert scale (Strongly disagree = 1, Disagree = 2, Somewhat disagree = 3, Neither agree nor disagree = 4, Somewhat agree = 5, Agree = 6, and Strongly agree = 7). This survey item was adapted from Miller et al.’s (53) instrument assessing whether their participants felt they were part of a community of scientific teachers, which again was a viable representation of our FLOSS participant population. We tested the assumption that pre and post were not significantly different with a paired Fisher’s exact test.

RESULTS

Qualitative results

Finding #1: Increased faculty perception of the importance of equity in their teaching

Faculty were more likely to mention the importance of equity in their teaching as a result of having participated in FLOSS (Fig. 1A). Prior to participating, most faculty (44%) made an implicit statement about equity as being important in their teaching, with slightly fewer (31%) stating explicitly how equity plays a role. When asked how equity was involved in their teaching, 25% of faculty responded with no mention of the importance of equity. However, after participating in FLOSS, every faculty member stated either explicitly (25%) or implicitly (75%) that equity was important to their teaching.

Fig 1.

Distribution of faculty responses to the question “How does equity come to play in your teaching?” before (pre) and after (post) participating in the FLOSS program. Responses were coded as either explicit, implicit, or no mention of equity’s importance (A). Responses were also coded as either specific, general, or no mention of individual action toward considering equity in their teaching (B). N = 19.

For example, one faculty member, before participating in FLOSS, said, “Generally, it (equity) doesn’t [come into play in my teaching]. But I do try to keep communication channels open and ask students to come to me with any challenges that are in the way of their learning.” This same individual, after participating in FLOSS, shifted from no specific mention of equity’s importance to responding with an implicit recognition of equity: “Coming up with a number of different ways to both teach material and allow engagement and assessment in the classroom [is how equity comes into play in my teaching].” Within the context of building inclusive learning environments, implicit versus explicit statements are not ranked, as it is valuable to both recognize the importance of equity and also to actively work on developing it in the classroom.

This finding implies that after completing FLOSS, all participants recognize that equity is important for their teaching. By recognizing the significance of creating equitable classrooms, whether implicitly or explicitly, it is possible that faculty who participated in FLOSS are more likely to look beyond purely content-related concerns and consider equitable best practices when designing or revising their courses, syllabi, assessments, and lectures.

Finding #2: Increased general and specific mentions of implementing actions for equity

Faculty were more likely to mention actions that one could take toward building equity in the classroom as a result of having participated in FLOSS (Fig. 1B). Before joining FLOSS, most faculty (44%) did not mention actions toward equity in their teaching. Of those who did mention an equitable teaching action, 19% made a general statement, while 38% had a specific mention of an action they or any instructor could make toward integrating equity into their teaching. After participating in FLOSS, there was a shift such that the majority of faculty (58%) had specific mentions of action, 25% generally mentioned action, and only 17% had no mention of action in their response. Thus, in addition to recognizing the importance of equity in teaching (Finding #1), faculty were also more aware of actions for equitable teaching practice.

For an example of a shifting mindset, one faculty member originally said, “Minimally, I'm not sure if I know what those (equitable) issues are”—a response coded as no mention of actions one could take toward equitable teaching. This same individual, after participating in FLOSS, said, “I try to bridge the gap to students who do not have the same access or who did not have access in the past,” which was coded as a general mention of action. In the case of action, being specific is a stronger indicator of understanding for the faculty, compared to a general response. However, even general mentions of action can lead to future specific faculty actions. In the case of this particular faculty member, they might be inspired to reconsider their curricula or course resources through an access lens, taking an important step toward a more inclusive classroom.

Finding #3: Shifting instructional deficit from students to instructors

Prior to participating in FLOSS, faculty most often assigned deficit to students or—within the “other” category—a general barrier when asked about the primary issues their students faced in a classroom (Fig. 2). In our non-exclusive coding, these majority assignments represented 47% assigned deficit to students, and an equal 47% assigned deficit to general barriers. The remaining pre-survey response category representation included 27% of the respondents mentioning instructor deficit, 13% mentioning student responsibilities, and 13% the institution as a barrier. After participating in FLOSS, faculty responses to the same question shifted dramatically such that the instructor was cited most frequently as the deficit, with all other categories being mentioned less than half the time. Instructors were assigned deficits across 46% of responses, with students cited 31% of the time, general at 15%, student responsibilities at 15%, and institution at 15%. The decrease in comments about student deficit and the subsequent increase in the instructor deficit narrative after participating in FLOSS is noteworthy. Simultaneously, it appears faculty became more specific and/or recategorized their perspectives of general sources of deficit, as this category dropped dramatically in representation among responses after FLOSS participation. Attribution of deficits to student responsibilities and institutions was very similarly represented among responses between the pre- and post-surveys.

Fig 2.

Distribution of faculty responses to the question “Considering DEI, what are the primary issues your students face in any classroom” before (pre) and after (post) participating in the FLOSS program. Responses were non-exclusively coded with the deficit, or fault, being assigned to the student, instructor, a general barrier, student responsibilities, or the institution. N = 19.

These results suggest that faculty who have participated in FLOSS assigned less responsibility to a student’s individual characteristics (e.g., academic preparation), and more to the instructional environment that faculty create in their classrooms. In one example, before FLOSS, one faculty member said: “... [student] motivation to participate/interact/seek help [is a primary issue students face in any classroom],” while after FLOSS they said: “Exclusion from participation; by access, socially, [and] by representation through material [are primary issues students face in any classroom].” In highlighting the student’s lack of motivation in the pre-program response, there is a suggestion that the student’s personal “grit” and predetermined inclination are to blame for the issues they might face in a classroom. By contrast, their post-program response highlights how students may be extrinsically excluded, both by the instructor (“access” and “representation through material”) and the peer community and/or culture (“socially”) within that classroom. This change in mindset shifts faculty toward recognizing their own role and responsibilities in addressing classroom inequity.

Quantitative results

Finding #4: Non-significant increases in faculty skill/knowledge of teaching

All participants responding to the pre- and post-surveys reported improved or equivalent pedagogical knowledge in eight targeted areas after participating in FLOSS. This benefit was seen in both the raw, individual participant data (not shown), as well as the average pre-/post-scores (Fig. 3). The largest increases in reported knowledge were for “Creating a classroom that is inclusive to all students,” and “Including regular, formative assessments in my courses.” However, there was no significant difference between average pre- and post-skill/knowledge levels for any area (A–H, P = 0.19), likely due to low sample size (n = 19). While not statistically significant, the increases at the personal level are promising after this initial stage of program implementation.

Fig 3.

Average self-reported faculty knowledge/skill before and after participation in FLOSS. Topics A–H highlight key components of FLOSS’s teaching curricula. N = 15, 79% response rate. We tested the assumption that pre and post were not significantly different with a paired Fisher’s exact test. P = 0.19.

Finding #5: Non-significant increases in faculty confidence for implementing inclusive teaching

To incorporate inclusive and equitable teaching practices, faculty must have both knowledge of those practices as well as enough confidence in their knowledge to actually institute those changes. Before completing the FLOSS program, 80% of participants reported they did not have enough knowledge to implement inclusive teaching (i.e., they lacked confidence in their knowledge, Fig. 4). After FLOSS, we saw a complete switch, with 80% reporting at least some level of confidence, as indicated by responding Strongly disagree, Disagree, or Somewhat disagree to the statement “I do not feel that I have enough knowledge to implement inclusive teaching.” However, there was no significant difference between average pre- and post-responses (Fisher’s exact test, P = 0.78), potentially due to our low sample size. We are nonetheless encouraged by the potential for FLOSS to build faculty confidence for incorporating inclusive and equitable teaching practices, and we propose that including data from any future FLOSS cohorts to increase our sample size could support this promising shift in instructor confidence.

Fig 4.

Faculty agreement/disagreement levels to feeling of confidence with implementing inclusive teaching practices before and after completing FLOSS workshops. N = 15, 79% response rate. Paired Fisher’s exact test, P = 0.78.

Finding #6: High rates of faculty sense of belonging to a teaching community

After participating in FLOSS, the majority of faculty (86.7%) agreed that they felt they were part of a community of teachers who are interested in active learning (Fig. 5). This was a slight increase considering 70% agreed on some level prior to FLOSS. While not significant (P = 0.67), this shift toward an increased sense of community is anecdotally supported by participant comments in internal program evaluations outside the scope of this paper. Based on the pre-survey, it is possible that participants had some community among their peers prior to FLOSS, yet the formal community of practice we created stood out and/or enhanced any existing peer groups. To our knowledge, none of the faculty from different institutions knew one another prior to FLOSS. So, if the sense of community was enhanced, it was among departmental peers and new relationships were bridged between institutions.

Fig 5.

Faculty agreement/disagreement levels to being part of a community of practice before and after completing FLOSS workshops. N = 15, 79% response rate. Paired Fisher’s exact test, P = 0.67.

DISCUSSION

When faculty focus on improving classroom equity and student performance through identifying and addressing perceived deficits in their students, they are regarding the students as the factor that needs to be fixed. This perspective allows instructors to distance themselves from the challenges they have identified (55), rather than consider how their interactions with their students, their educational approach, and their classroom climate may be causal or contributing factors. The FLOSS program workshops were structured to first identify the systemic barriers that their students are facing and then to explore evidence-based pedagogical interventions. After completing the FLOSS program, faculty participants demonstrated a shift in their understanding of their role in matters of diversity, equity, and inclusion within their classroom setting. By recognizing that the deficits identified may be attributed, at least in part, to instructor contribution, a factor that they can directly manage, faculty are empowered to take a more proactive role in addressing inequity in their classrooms. This shift in mindset allows instructors to critically reflect and consider how they can change their teaching practices to make their courses more inclusive and conducive to student success.

Adopting this more inclusive mindset is a critical component. Simply implementing inclusive teaching practices, without the corresponding change in instructor mindset, can limit the effectiveness of the intervention or lead to the development of new issues. If instructors continue to view their students through a deficit lens, they may unintentionally convey this perspective through subtle behaviors or actions in their teaching practice or when interacting with students. Learning environments that explicitly champion equity and inclusivity while perpetuating disruption through microaggressions or implicit bias threaten students’ sense of belonging due to conflicting messaging (56).

During the FLOSS program, in addition to providing faculty participants with tools for assessing their mindset and educational approach, we introduced faculty participants to active learning strategies and offered them opportunities for practice. We emphasized that, while incorporating active learning into a course has been shown to increase equity (26), building an inclusive classroom is an ongoing process that requires continual cycles of learning, implementation, assessment, and reflection. Furthermore, these pedagogical tools are most effective when utilized in a manner that is appropriate for the specific students in your course (57), and with consideration for the dynamic nature of variables that are both internal and external to the classroom. Understanding this will assist instructors in appropriately assessing the effectiveness of implementing new pedagogical strategies. An instructor who locates a deficit in their students may fault the students if the implementation of new pedagogical strategies does not go well, rather than reflecting and potentially adjusting their instructional practice.

Interestingly, the FLOSS curriculum did not aim to directly educate participants about locating deficits in their instructional methods rather than in their students. Instead, this shift in mindset emerges organically from beginning to understand both the systemic nature of the issues faced by their students (Workshops 1–3) and the demonstrated effects of implementing inclusive best practices in teaching (Workshops 3–7). It is not uncommon for educators to demonstrate resistance when told to change their mindset and/or update their instructional practices. This can be mitigated, however, by making this experience more collaborative, where faculty view themselves as working with those offering the professional development. Buy-in is fostered by giving faculty the opportunity to explore the rationale behind the suggested changes, offering them the time and space to consider the potential outcomes of proposed interventions, and providing them with the resources to effect the desired changes (58). We believe that faculty participants exhibited the gains they did because, through FLOSS, we offered them a collegial and interactive environment in which to explore the various topics presented. In addition, we created a framework for the individual relationships built during the program to extend over a longer term. Ideally, our communities of practice would persist beyond the completion of the FLOSS program. However, sustaining a long-term community of practice for STEM education reform requires significant resources that were beyond the scope of our program (59). Nevertheless, we hope to obtain future data on whether/how the shift in deficit modeling leads to practical changes in instruction for our faculty participants and/or resulting effects on student success at our institutions.

One example of a program that has been successful in putting this mindset into action is the Biology Scholars Program (BSP), which has been active at the University of California, Berkeley, since 1992. Founder John Matsui’s primary recommendation for diversifying STEM is to “Fix our institutions, not our students” (60). BSP enrolls students from a range of backgrounds, with a focus on those that have been historically marginalized and/or underrepresented. These students have statistically lower high school GPAs and SATs than non-BSP students, yet graduate with biology degrees at significantly higher rates than non-BSP students regardless of race/ethnicity (61). Lower high school GPA/SATs are often used as a proxy for students being “underprepared” (e.g.,reference 32), which is a common student deficit explanation for the origin of disparities in college course grades/academic success. Crucially, despite enrolling students who might be classified as “underprepared,” BSP programs provide support without any “remedial” framing (60). Instead, BSP recognizes that incoming GPAs and test scores are not true measures of a student’s ability or a prediction for success or failure, and it is the learning environment that can make the difference.

Within the scope of an individual faculty member’s classroom, relocating the deficit focus from the students to the instructor is an important first step, but it is critical that this shift in mindset translates into actual change in the classroom. In order for pedagogical change to occur, faculty must have both knowledge of inclusive teaching practices and the confidence and support to implement them. In their seminal paper on how active learning can help to narrow opportunity gaps in STEM, Theobald et al. (26) proposed a “heads-and-hearts” approach incorporating deliberate practice (heads) and a culture of inclusion (hearts). The FLOSS program followed a similar approach for working with our faculty participants.

Knowledge of inclusive teaching practices

Within the FLOSS program, the eight FLOSS workshops provided the deliberate practice component. The workshops were designed to be highly interactive and model inclusive teaching strategies. Participating in the workshops gave faculty the opportunity to experience and then practice a variety of different techniques for structuring instruction and interaction. The accompanying Canvas course outlined and referenced detailed information about the pedagogical approaches utilized in each workshop, and faculty were frequently asked to reflect on how they could incorporate what they learned into their courses. In the final workshop, participants brought in examples of their own teaching and provided feedback to each other on how they could use what they had learned through FLOSS to make their teaching more inclusive. At the end of the FLOSS program, participants indicated an increase in their knowledge about inclusive teaching practices and were also more likely to mention actions that they could take in response to equity concerns. In addition, after FLOSS, fewer participants ascribed the fault for inequity to “General” barriers, indicating that they were now equipped with more specific terms and ideas for their discussions on equity in the classroom.

Support and confidence for implementing inclusive teaching

One of the major barriers for STEM faculty in implementing pedagogical change is a perceived conflict between their professional identity as scientists vs. their identity as instructors, with faculty being afraid to “come out” to their peers as a teacher in addition to being a researcher (62). Working together in a community of practice can help to ameliorate this conflict, as participants are all joined by common experiences and shared interest in seeking to improve their teaching. This supportive community thus provides a safe environment for exploring inclusive teaching ideas and experimenting with new pedagogical strategies (the “hearts,” or culture of inclusion, component).

Communities of practice are able to succeed in effecting transformational change where one-off workshops or individual studies (e.g., through readings or asynchronous modules) struggle because of the emphasis on social support and continuous engagement, and practice with the subject. Perhaps unsurprisingly, the same elements that lead to increased student learning also apply to effective learning for faculty: (i) engaging in deliberate practice and (ii) in an environment that is inclusive and supportive. Communities of practice have been successful in increasing adoption of evidence-based teaching strategies (63, 64), and these changes have been sustained for years after the communities of practice concluded (65). In addition, members of communities of practice can themselves become “linkage agents” who disperse what they learned to non-participants (66). Thus, the effects that we see with FLOSS may continue to spread throughout the individual departments and institutions of the participants.

The FLOSS program intentionally gathered faculty from three different categories of institutions of higher education that are linked both geographically and through students who transfer or cross-enroll. Each of these institutions, nevertheless, generally serves different student populations, and the intention was that having these faculty work together would help to expand the perspective of the faculty at other institutions. Just as diverse working teams are more effective than homogeneous teams for solving problems (67), a diverse community of practice has a wider range of experiences and ideas to bring to the central question of how we can build a more inclusive classroom. Within the workshops, participants discussed the similarities and differences in their student populations, while also recognizing the commonalities in many of the barriers that students face. While our pre-/post-surveys did not find significant differences in faculty sense of community as a result of participating in FLOSS, this may have been the result of both low sample size and the imprecise wording of the question. In informal program feedback, participants unanimously cited the community of practice when asked in an open prompt to describe the main strength of the program.

Finally, the role of the community of practice may have been particularly important for FLOSS participants as the program took place in the Spring and Fall of 2021, when the COVID-19 pandemic was still causing major disruptions to higher education. During this time of crisis, faculty communities of practice were instrumental in providing support as instructors were forced to adopt new modalities of teaching and learning (68, 69).

Conclusion and future directions

There have been a number of recent calls to recognize the importance of shifting the deficit framework away from students and onto faculty and institutions (e.g., references 60, 70). The FLOSS program is an encouraging model for how a relatively low-commitment faculty development program can lead to this shift in faculty mindset while also providing the knowledge and support needed for implementing inclusive and transformational pedagogical practices. As FLOSS completes its pilot stages, we hope to scale up this work in the future to pursue additional research questions.

Does this change in faculty mindset lead to a change in their teaching practices?

The FLOSS program was designed to provide faculty with the tools for building equity in their classrooms, and further research is needed to assess whether and to what degree inclusive teaching practices were actually implemented. Surveying faculty about their practices is insufficient for evaluating effectiveness, as faculty perceptions of their own practices are not always accurate when compared against an objective observer (71) or against the perceptions of their students (72). If changes in practices are observed, then it would be worthwhile to examine the relative roles of mindset shift, knowledge of best practices, and the support of a community of practice in facilitating those changes. Finally, we hope to follow up with program participants to assess whether the effects of FLOSS on mindset, knowledge, and sense of community are maintained or shift over time. Longer-term follow-ups are also desirable as there may be lags in implementation (even among faculty who rate themselves as prepared), or faculty may need time to experiment to determine the most effective strategies for their contexts (72). The results of implementation may likewise shift over time as both faculty and students become more comfortable with these new pedagogical strategies (e.g., reference 73).

How does a shift in faculty mindset lead to changes in student outcomes?

The main goal of FLOSS was to improve equity in our classrooms and thus, consequently, improve the success and persistence of minoritized students in STEM, so further work will be needed to determine whether and how student outcomes are affected. Does faculty participation in FLOSS lead to reduced opportunity gaps in the courses that they teach? For the faculty teaching introductory-level courses, are there long-term effects on student success and persistence (e.g., references 29, 31, 74)? To what extent are the changes in student outcomes mediated by improved affective traits for their students (e.g., sense of belonging, self-efficacy, self-as-scientist)? As faculty move away from a student deficit model in thinking about their teaching, they ideally shift toward a “dynamic model” where they consider those affective aspects of learning alongside cognitive aspects (75).

Ultimately, in our efforts to achieve equity in our classrooms, we must seek to follow the call of Handelsman et al. (70): “[S]top trying to ‘fix’ the students and fix our classrooms instead.” STEM faculty have a central role in cultivating a learning environment in which disparities are reduced and all students have the opportunity to succeed. It is critical that we ensure equitable access to opportunities for all students, regardless of background or socioeconomic status, so that we can cultivate a pipeline of diverse talent and empower individuals to pursue fulfilling careers in STEM. These students will become the future generations of scientists, researchers, and healthcare practitioners that drive innovation, fuel societal progress, and strengthen the nation’s economy. This work demonstrates that communities of practice may be effective agents of this change.

ETHICS APPROVAL

The study was approved by the University of California, Berkeley’s Institutional Review Board (CPHS # 2021-03-14114).