Introductory college labs provide students with a valuable hands-on introduction to experimental science (Gormally et al., 2011; Kloser et al., 2013). In labs, students typically work together in small groups to engage in often-unfamiliar scientific work in a culturally rich environment. The experience of navigating instructional spaces can strongly influence students’ science interests (Hazari et al., 2013), self-efficacy (Sawtelle et al., 2012), and identities as science people (Carlone & Johnson, 2007). Moreover, this experience can differ significantly depending on students’ identities (Hazari et al., 2010; Rosa & Mensah, 2016). Unless explicit efforts are made to create an equitable and inclusive learning environment, interactions with peers may serve to perpetuate or exacerbate inequities with regard to who is recognized as a scientist, who experiences stereotype threat (Maries et al., 2020), and who benefits from this first taste of experimental work (Danielsson & Linder, 2009; Van Dusen & Nissen, 2020). Thus, how students collaborate in introductory labs can have an impact on inclusion and diversity in our fields (Mulnix et al., 2016).

Over the past few years, we have used ethnographic observations and interviews to study a variety of ways in which female students in our introductory physics labs experienced inequitable learning opportunities, including gendered task division (Doucette et al., 2020); students “splitting” rather than “sharing” lab work (Doucette & Singh, 2022); and women experiencing isolation, stereotype threat, and lack of recognition for their work (Doucette & Singh, 2021). Related patterns have been reported elsewhere (Quinn et al., 2020). Physics has strong stereotypes about being a field for brilliant and aloof men, and women often report a “chilly climate” in physics (Hall & Sandler, 1982; Simon et al., 2017). The prominence of gender-based inequities in our prior investigations is consistent with these stereotypes. Thus, when the COVID-19 pandemic forced us to design online versions of our introductory physics labs for the fall and spring semesters of the 2020–21 academic year, we were already concerned about the issue of gender-based equity and resolved to investigate and attempt to minimize these issues in the online lab. Despite research-based best efforts (e.g., Oleson, 2020), our first attempt at equitable collaboration in our online lab course was not realized. However, notable improvement in the second iteration helped us identify some effective strategies, which we describe in this article.

Methods

The introductory physics lab course at our large, primarily white, research-intensive university enrolls students in health science and physical science tracks, but not engineering students, and meets weekly for 13 weeks over the course of a semester. Typically, 55% of the students in this lab course identify as women. About half of the students are simultaneously enrolled in the second half of an introductory physics sequence, and the other half have already finished it. Students may take the lab at any time in their studies, with most taking it during their third year.

In the lab, students completed an individual lab investigation using an iOLab device (Bodegom et al., 2019; Leblond & Hicks, 2021), then met in groups of four in Zoom breakout rooms to work through a collaborative activity. New groups were randomly assigned each week. Students were each asked to adopt one of four different roles while doing this collaborative work: manager, experimentalist, archivist, or skeptic/theorist. The collaborative activities, written by Danny Doucette, required students to work as a group to combine data from their individual investigations, design and conduct a new experiment, and write a reflection on the experimental process. The groups submitted one single report, which was generated collaboratively using the iOLab Lesson Player and exported as a PDF for submission to our learning management system.

To understand students’ experiences in this lab course, the authors conducted interviews with 11 students (nine women, two men) near the end of the fall 2020 semester. The students were selected randomly from a pool of 24 who responded to an emailed invitation to participate, and the interviews were part of an Institutional Review Board–approved study with students’ informed consent. These interviews suggested the existence of pervasive and problematic interactions in which women were being stereotyped and ignored. Next, we extracted the groups’ responses on the reflection portion of their collaborative activity report and compared the length of responses from groups with different gender compositions. This analysis suggested that groups with an isolated female member may have had lower levels of engagement, but it also showed that providing clear grading expectations and feedback may have alleviated this issue. Finally, we revisited our use of roles in this group work and evaluated the effectiveness of roles as a tool to help students collaborate equitably while doing group work in the introductory physics lab. We found evidence that roles may be effective at helping students interact in productive and equitable ways if they are integrated meaningfully into the work students do.

Interviews

During the hourlong, semistructured interviews, we asked students about their impressions on the effectiveness of the lab course, their experiences with remote learning in this and other lab courses, and their interactions with other students in the lab groups. We coded the responses using an emergent coding scheme, which produced two key themes relevant to students’ gendered experiences in the introductory physics lab. We conducted interviews near the end of the fall 2020 semester but did not repeat the interviews in the spring 2021 semester because we were informed that instruction would return to an in-person format in fall 2021; thus, we did not need to continue refining the online lab course.

The first theme describes women in the lab being talked over, ignored, or not being taken seriously when they worked in groups with men. One woman described feeling as if she were “being overshadowed by the men” when she worked in a group with only one woman. She suggested that instructors try “to make sure that not just one woman is in a group.” This advice aligns with research results suggesting that small groups with only one woman are less effective (Heller & Hollabaugh, 1992; Sullivan et al., 2018).

Another woman explained feeling as if the men in her group ignored her and did not take her seriously. She explained that “sometimes the guys don’t take what I say seriously. … A guy would say something and then I would say something different. They would just ignore what I say. We would ask the instructor or like the TA [teaching assistant], and then I would end up being right.” This experience was common among the women we interviewed. Another woman told a similar story in which the men she worked with supported one another when there was a disagreement:

I was the only girl and there were three or four other guys. … I would say something, which would be right ... but other people would disagree with what I said. Which is fine, like, they can disagree. But I also felt like if it was one guy against me, all the other guys would just side with that guy even though they didn’t even have a proper reason for doing it. … Then the TA would come in and corroborate what I said. And then people would be like, OK, and then change it ... I felt that a lot.

In our interviews, students also described witnessing other students experience gender-based discrimination. One student described witnessing the experiences of a woman who was struggling with some physics concepts and ignored by the rest of her group when she asked for help. She “was just talked over and wasn’t given as much consideration. Nobody would take the time and explain it … and they would just push through it. [She] would have to deal with it.”

Whether they were ignored, talked over, or not taken seriously, it was clear that the women in our introductory physics lab were experiencing gender-based discrimination. It is possible that the new and uncertain learning environment (Zoom breakout rooms), lack of clarity about grading expectations or how groups were meant to work together, or the culture of physics could have contributed to these unproductive interactions. Regardless of the specific cause, women in our online introductory physics labs were experiencing an inequitable learning environment.

The second theme involves the stereotypes associated with women’s contributions to, and expectations for, graded work in the lab. Many of the students we interviewed, both men and women, described a stereotype that women in our labs were willing to do more work for the same task and that they would insist on submitting better and more thorough work for grading. Whether or not it was true (Meece & Jones, 1996), the existence of this stereotype may have affected the students’ group work.

One woman outlined the stereotype: “I think everyone that I’ve worked with wants to get a good grade. But I think sometimes when I do work with guys-only groups, if we don’t understand something, they’re like, ‘Oh well, what we’ve put is good enough.’” On the other hand, she continues, “When I worked with just two girls, they were like, ‘No, we have to make sure this is right.’” This student described a stereotype that women are generally hardworking perfectionists and men are slackers or social loafers (Oleson, 2020) willing to do only enough to get a C.

The existence of this stereotype may have had an impact on the work students did in the lab. One woman described needing to do most of the work during a collaborative activity in her Zoom breakout room in which she was the only woman, as none of the men were willing to contribute. “It was just silent, nobody did anything. I typed up all the answers. They didn’t even check my work.” In this case, we hypothesize that the men slacked off because they believed the work would be done by a woman who, as the stereotype may have led them to believe, would work hard to complete the activity regardless of their contributions (or lack thereof). As a result, not only was the work unfairly distributed, but none of the students benefited from the discussions that would follow from engaged collaboration.

We included roles (manager, experimentalist, archivist, and skeptic/theorist) in the labs in an attempt to spread out the work. Each week, groups were asked to self-assign roles in such a way that everyone got to do each of the different roles about the same number of times over the course of the semester. However, some of the students co-opted this system. One woman explained that when she worked with men, “they are always first to say, ‘I’ll be the skeptic.’ ‘I’ll be the manager.’ So then I’m stuck with the role of the experimentalist or the archivist, which is traditionally a little bit more work than the other two [roles].” In this case, the student perceived that some of the men she worked with chose roles that would allow them to minimize the amount of work they would contribute to the group collaborative activity. Again, the men choosing to be the skeptic or manager were trusting in a stereotype that the women in their group would take on the more involved roles.

Another woman described feeling “a little nervous” when she was the only woman in a group with three men and assigned to be the experimentalist. She described feeling uncertain when speaking up about some issues in their work, telling herself, “Oh, we can talk about it at the end. … We can revisit it, and I can like look through things as well.” After they finished the Zoom call, she looked back through their work and “realized that [she] would not put a lot of the answers that they put, because a lot of the answers that they put were very skimpy and not very detailed and didn’t include a lot of the stuff that [she] would have told them to include.” She continued, “So I had to spend an extra 10 to 15 minutes trying to, like, fix the answers a little bit.” She felt torn between her desire to ensure their work was complete and correct and her worry that she would validate the stereotype by speaking up. In the end, she described feeling “stressed” and ultimately decided to “just let it go.”

Regardless of whether or not it was true that women were more willing to work hard to get good grades, the women we interviewed described needing to navigate those stereotypes. It was as if the women were walking on eggshells, aware of their difference within the culture of physics and feeling the need to carefully weigh their engagement in group work. Uncertainty about their role and whether their ideas and work would be accepted may have led women in our lab to contribute less to the group work than they otherwise might have.

Engagement

The interviews provided anecdotal evidence that the gender composition of groups may have affected the groups’ cohesion and effectiveness. In particular, several women mentioned that they felt ignored or overshadowed when they were the only woman in their group. Other women described how they felt the need to navigate stereotypes about the types of expectations men and women held toward their work in the physics lab.

To evaluate the impact of different group compositions on the effectiveness of their collaboration, we measured the length (in characters) of the reflection portion of the collaborative activity from 1,342 different groups of students during the fall 2020 and spring 2021 semesters. These reflections called for students to collaboratively respond to prompts about the nature of experimental science, the application of a physics concept in a novel situation, or even the nature of their collaborative work. We created a Z-score (i.e., scaled so the average = 0 and the standard deviation = 1) each week and combined all of the weeks for each semester. The average Z-score for each group composition (e.g., all men, one woman in a group with two or three men) from fall 2020 is plotted in Figure 1.

The reflections spanned a spectrum from cursory and perfunctory to thoughtful and in-depth analyses. We chose to measure the length of the reflection portion of the collaborative activity because we hypothesize that groups that are more engaged will have lengthier discussions, will surface and share more creative or divergent ideas, and will be more likely to integrate different perspectives as they compose their reflections. We also analyzed reflections using a rubric to assess quality via a few metrics (e.g., originality, thoroughness, relevance to the prompt), but this assessment largely reproduced the patterns we saw from the length of the reflections.

The data in Figure 1 suggest two takeaway messages, in spite of the large error bars. First, groups that are all men or all women tend to have the highest level of engagement (i.e., they tend to write the most). Second, groups that include exactly one woman have the lowest level of engagement. One immediate takeaway, based on Figure 1 as well as the interviews described earlier, is that instructors should avoid forming groups of three or more students that include just one woman.

In spring 2021, we made two changes to the collaborative activities. One of those changes was the introduction of detailed grading rubrics. We expected that a change in grading policies may have had a substantial impact on the work students submitted, so we present the engagement data (i.e., length of the reflections response) in Figure 2. We note two changes from Figure 1. First, the level of engagement from all groups that included women was higher in spring 2021 than it was in fall 2020. This suggests that all of the groups that included women wrote somewhat more in spring 2021 than their counterparts in fall 2020, as compared to the “All Men” groups. Second, we note that groups with exactly one woman no longer had the lowest level of engagement. Instead, all of the groups that included women had comparable levels of engagement.

By comparing Figure 1 and Figure 2, we suggest that the introduction of grading rubrics had an impact on the level of engagement for groups of students and that the impact was particularly positive for groups that included both men and women. The grading rubrics that we introduced in spring 2021 provided detailed performance indications, including specifying that a creative, thorough, and/or insightful response would be required for full marks. The grading rubrics also allowed graduate student teaching assistants, who both instruct and grade for the labs, to provide more useful feedback to students, more quickly, about their collaborative activity work. Overall, these results suggest that detailed grading rubrics may have had a positive impact on equity in our introductory physics labs.

Roles

In addition to the introduction of grading rubrics, we also made modifications to how we included roles in the collaborative activity for the spring 2021 semester. In fall 2020, students would type their name, their role, and the number of times they had previously done that role on the first page of the collaborative activity report. The first page also included a one-sentence description of each role (manager, experimentalist, archivist, and skeptic). Other than this requirement, there was no particular direction given to students about how to do the roles or why they mattered.

Our interviews with students at the end of the fall 2020 semester revealed that the students found the roles useful as a way to alternate which student was the experimentalist and archivist (roles that were perceived to require more work) but that they did not see much value in the roles other than this reason. However, one student suggested that it was possible to make it so that “different roles involved different forms of thinking.” For example, while serving as the archivist, this student “tried to listen to everyone to make sure [they weren’t] missing someone’s portion of their discussions.”

Inspired by this student’s idea, for the spring 2021 semester, we shared with students a video of four graduate students working on one of the collaborative activities in such a way as to emphasize different types of thinking associated with the different roles. We changed the skeptic to theorist, which called for that student to be the reference person for issues related to physics concepts, theory, or calculations, in addition to taking the lead on interpreting data. We also rewrote the role descriptions on the first page of the collaborative activity report to be more detailed, and we added a “role reflection” in which each student would write one or two sentences describing how they fulfilled their role while working on the collaborative activity that week.

Our goal in changing the use of roles to become more integrated with the lab was to help students become better able to collaborate effectively with their peers. We expected that more engaging, meaningful interactions with other members of their group would help students learn more about physics concepts and experimental physics. To evaluate the impact of the change to our use of roles, we compared student responses to the survey item “Interactions with my peers helped me learn in this lab course,” which was added to an end-of-semester survey in both fall 2020 and spring 2021. Responses to this survey item are tallied in Figure 3.

Using a Mann-Whitney U test, there is a significant difference (p = 0.023) between the distribution of responses to the survey item in the fall and spring semesters, with the spring semester having a higher fraction of students who agreed that interactions with their peers helped them learn in the lab course. These results suggest that a better integrated use of roles may have been associated with more productive group interactions by improving how students collaborated with one another in the collaborative online lab activities.

Conclusion

During the 2020–21 academic year, we sought to provide students with opportunities for synchronous, productive group work in our online introductory physics labs through the introduction of collaborative activities completed in groups of four students via Zoom breakout rooms. In interviews, students reported that women in the class were ignored, talked over, and not taken seriously when working in mixed-gender groups. In the second semester, we introduced explicit grading rubrics and improved the way that we were using roles for group work. These changes may have been associated with an improvement in the engagement of mixed-gender groups and an improvement in students’ perception that interactions with peers helped them learn in this class.

The changes we made to our online introductory course focused on improving group interactions in order to improve equity in student experiences and outcomes. In addition to modifying learning activities and procedures, it may also be possible to introduce new activities or interventions to improve group interactions. One area of future study is to investigate how psychological interventions (Binning et al., 2020) or discussions about equity (Daane et al., 2017; Rifkin, 2020) in science could improve the nature of classroom collaboration in the lab context.

In summary, these results point to two strategies instructors can use to make introductory science labs more equitable and inclusive. Although our study focused on online lab courses, in the future, we will investigate the effectiveness of the following successful strategies to improve equity for in-person instruction: (i) Provide students with clear and explicit performance expectations (e.g., via rubrics), and (ii) incorporate roles fundamentally into assigned group work.

Danny Doucette (djdoucet@ncsu.edu) is a teaching assistant professor in the Department of Physics at North Carolina State University in Raleigh, North Carolina. Chandralekha Singh is a distinguished professor in the Department of Physics & Astronomy and the director of the Discipline-Based Science Education Research Center at the University of Pittsburgh in Pittsburgh, Pennsylvania.