When we sort students into cooperative learning groups, we often attempt to create balanced groups that reflect the diversity of the students in our classes. In this episode Olga Stoddard joins us to discuss her recent research that suggests that this approach can be harmful for female students in classes in which a majority of the students are male.
Olga is an Assistant Professor of Economics at Brigham Young University, a Research Fellow at IZA (the Institute of Labor Economics), and the Research Director at the Science of Diversity and Inclusion Initiative, and the Co-Director of the Gender and Civic Engagement lab at BYU.
- Stoddard, Olga B.; Karpowitz, Christopher F.; Preece, Jessica (2020) Strength in Numbers: Field Experiment in Gender, Influence, and Group Dynamics, IZA Discussion Papers, No. 13741, Institute of Labor Economics (IZA), Bonn
- Zölitz, Ulf and Jan Feld (2018), “The effect of peer gender on major choice.” University of Zurich, Department of Economics, Working Paper.
- Sandberg, S. (2013). Lean in: Women, work, and the will to lead. Random House.
John: When we sort students into cooperative learning groups, we often attempt to create balanced groups that reflect the diversity of the students in our classes. In this episode we discuss recent research that suggests that this approach can be harmful for female students in classes in which a majority of the students are male.
John: Thanks for joining us for Tea for Teaching, an informal discussion of innovative and effective practices in teaching and learning.
Rebecca: This podcast series is hosted by John Kane, an economist…
John: …and Rebecca Mushtare, a graphic designer.
Rebecca: Together, we run the Center for Excellence in Learning and Teaching at the State University of New York at Oswego.
John: Our guest today is Olga Stoddard. Olga is an Assistant Professor of Economics at Brigham Young University, a Research Fellow at IZA (the Institute of Labor Economics), and the Research Director at the Science of Diversity and Inclusion Initiative, and the Co-Director of the Gender and Civic Engagement lab at BYU. Welcome, Olga.
Olga: Thanks so much for having me. It’s great to be here.
John: …really pleased to have you.
Rebecca: Today’s teas are:
Olga: I was going to be prepared and I have my mug, but unfortunately, it’s only filled with water because I ran out of time to heat it. [LAUGHTER] So, water for me today.
John: Tea is mostly water. We’re recording this in mid February when there’s a bit of a nationwide snow covering. And I’m drinking spring cherry green tea to set a better mood for the future.
Rebecca: I think that seems like a good plan. And for a change, I’m drinking Chai.
John: Wow. Okay, I don’t think I’ve seen you drink that on here before.
Rebecca: It’s not a common one for me. But it’s nice to mix it up occasionally. Of course my Chai doesn’t have dairy in it. So it’s just the tea part of the Chai
Olga: Is it flavored Rebecca?
Rebecca: Yeah, it’s nicely spiced.
John: We do normally in our office have a variety of flavored Chai teas, but they’re safely locked up in our building. We haven’t visited in a long while. We’ve invited you here today to discuss your research with Chris Karpowitz and Jessica Preece concerning how the gender composition of teams affects women’s participation and role in team activities. Could you tell us a little bit about the study?
Olga: Yeah, absolutely. So this study was a collaboration with a top 10 accounting program in the US. We partnered with them to randomly assign different gender compositions of teams in this program. So, like many programs in the US, especially business programs, like MBA programs, this particular program relies on a pedagogical group-based approach in which students are assigned into teams, in this case, teams of five. And they work together quite intensively throughout the semester. So throughout the four months that’s their first semester in the program, they work on assignments together, they meet socially outside of the classroom, they even do some of the exams as a group. And so there’s a lot of interaction between those students within those seats. for that period of four months. Normally, because this program has a really small percentage of women, so about 25% of the students in the program are women, the way that these groups had been formed in the past is to assign one woman per group, so as to sort of dilute the women, to have men have experience in an academic and professional setting interacting with women. There is some prior research in the laboratory that has shown that this really is detrimental for women’s ability to be influential, for their willingness to participate, to be engaged. And so what we wanted to do is we wanted to test whether that laboratory evidence plays out in a similar way in the, so to speak, real world setting, more naturally occurring kind of environment. So we partnered with the program and randomly assigned some women to be in the condition where they were the only woman in the group. So the status quo, this is how things have been done. One woman and four men in a five person group, and then other women were randomly assigned to be in a condition where they were in the majority. So there were three women and two men in a five person group. We then tracked these students for the following two years. We had them complete monthly surveys and peer evaluations of their group members. We had them come into the laboratory twice a semester, where we had them work on a team-building exercise, and we watched who’s participating. These exercises were recorded, so we could see who’s speaking, who is interrupting, with speaking for how long. so that we could precisely measure women’s participation, but also measure their level of influence. Because on these tasks, the way that we designed them, women could exert more or less influence depending on certain decisions they make. So we had different ways to measure their level of influence, their participation, and whether others perceive them to be influential, and sort of more like leaders in the group. And so we did that for the following two years. We had two cohorts of students participate in the study. And what we found is that women who were randomly assigned to be the lone woman in the group were perceived to be significantly less influential, and were actually exerting a lot less influence in the group than the equally qualified women who had been assigned to be in the majority in their group. And so we saw really striking differences across those two conditions. Again, these are equally qualified, very well prepared academically women. This program is very competitive. They have prior leadership experience. And yet we find these huge differences across the two conditions in our case, depending on whether women were in the minority or the majority, they were seen significantly less influential by their peers.
Rebecca: Was the perception of the women in the groups different from the start of the study or the beginning of the group formation versus the end of the group formation? Or was it kind of consistent?
Olga: Yeah, that’s a good question. So one advantage of our study is that we can track these students over a relatively long period of time. Most laboratory studies up to date have relied on sort of these one-shot types of interactions, where strangers meet for a period of an hour or so and never interact again. One thing we wanted to know is do these patterns that had been observed in the lab to date, exacerbate over time, or do things get better as team members get to know each other, they get to experience women’s authority or their expertise. And what we found is that it’s mixed evidence on this. So in these surveys, these are monthly surveys that students have to fill out about each other… we call them peer evaluations… and in these peer evaluations, we ask them “Who is the most influential member of your group?” And they state who is the most but also who’s the least influential. What we found that over time, over the course of those four months that these students work together as a group, there is an improvement for the lone women, that their peers perceive them to be more influential over time. For the women in the majority, there seems to be no change. And so we do see the gap closing by the end of the semester, relative to the large gap in the beginning of the semester, but only in the survey data. Once we actually look at the data from the lab, where we observe students interacting in teams, where we can measure who is exerting influence on a task, we see no difference over time. So it seems that there is some improvement for the lone women in these sort of general assessments of influence in these monthly “Who was the most influential member of your group over the course of the month?” But when you actually get down to the specific tasks, we don’t see any improvement for women over time.
John: I know you were looking at this in a very broad context, in terms of teams and organizations and firms and so forth. But in terms of classroom groups of the sort that you were actually experimenting with, a growing number of classes in pretty much all disciplines now rely on group activities. What does the study suggest about how we form these groups in terms of the gender composition of groups, so that everyone can have an active role in the group?
Olga: Like you said, both in the workplace and in many academic settings, group work is crucial. And many faculty members rely on group- based activities. Understandably, they prefer collaborative thinking and develop the skills that students will need as they go on in workplaces where increasingly there’s reliance on group work. And so certainly the implications from our study are that assigning groups in which women are the lone woman or in the minority is going to have costs for women, costs in terms of participation, in terms of influence, in terms of whether they’re seen as authoritative, as leaders in the group. Those are the types of questions we ask and things that we can measure. And so certainly, if at all possible, groups that are gender balanced, or groups in which women are in the majority, are going to be significantly better for women in terms of these types of outcomes. Now, I would add a couple of caveats here. One is that in our study, we can track the grades. We can see what students actually get at the end of the semester. And we find no penalty for women, as far as grades can tell, when they’re in the minority. The women who are in the minority receive about the same grades as women who are in the majority. However, the grades in this program are largely group based. So it may not be surprising, because so much of the grade is based on the group work that we’re not finding those differences. Moreover, we don’t know how women get to those grades. It’s possible that because of these influence gaps, they’re having to work extra hard to get the same grade, or to be seen as sufficiently expert in that particular class. And so those are the two caveats that, even though we don’t observe differences in grades in our study, that doesn’t mean that there aren’t underlying differences in how hard students are having to work or how much effort they’re having to exert. I would also note that, regardless of the gender composition, there were no differences in man’s perception. So the man, whether they were in the minority, or in the majority, saw no deficit in influence, they were equally likely to be seen as a leader, they were seen equally influential. And so, if one thinks well, putting men In the minority is going to all of a sudden hurt the men in the group. That’s not what we’re finding. And there is in fact quite a bit of literature now confirming that. There are laboratory studies and studies in different settings, like nursing school, where men are in the minority, and in fact are not incurring any kind of deficit as far as influence or participation or authority that the women are incurring in these kinds of settings in which they’re a minority. I would also mention one study, it’s a working paper by a PhD student at University of Zurich, and it’s a really great working paper. She’s looking at a setting in which women are a minority… economics… a setting we’re familiar with. And in that setting, she’s using some data from, I believe it’s University of Zurich, it might be another university in Europe, but at that university, they also created different study groups, just like in our study, except these are larger study groups. These are sections of about 50 to 60 students, and they also randomly assigned gender compositions of these study groups. And what she shows is that, over time, the women that are assigned to be in a group in which they’re a minority, are much more likely to drop out of the study group altogether; that they not only incur these potential influence deficits, which we document in our study, but there are, in fact, very serious consequences to their ability to thrive in that class, or to thrive in that environment in which there are a minority. So that’s closely related, of course, to our study, and confirms really similar patterns.
John: We’ll share a link to both studies in the show notes. You mentioned that in disciplines like economics, and more generally, in STEM fields, women are often underrepresented as students, but they’re also underrepresented in faculty. It’s likely that these types of issues will carry over into group meetings and team meetings and department meetings and so forth on campuses. What can women and departments do to address this problem?
Olga: Yeah, that’s a very good question. Certainly, the setting in which we study these topics is student groups. But we are more than confident that these kinds of patterns replicate in a variety of settings, including professional settings, whether you’re a faculty or a student, being in the minority as a woman entails these costs to your level of influence, to your ability to exert influence, to your ability to be heard and taken seriously. And certainly there are other studies that have found very similar patterns in other kinds of settings. So I would not be surprised that if we ran this study in a professional setting or a workplace, we would find very similar patterns among women at all levels, including leadership. Certainly, some studies have confirmed similar patterns among the board of directors, female directors. The question of what can women do to sort of fix this is a really complicated one. And I say that because, what we find in our particular study, for example, is that women can’t just overcome that deficit by working extra hard. One thing that we observe is their levels of participation, how much time they put into coursework, and things like that. And we find that to be the same, regardless of the condition in which they’re in. They’re working extra hard already. Another thing that we observe is their talk time. In this laboratory setting, we can measure how long each person talks. And so you might say, well, maybe women, they’re just not leaning in, maybe they’re not participating enough in these group discussions, and so of course, they’re not seem as influential. Well, we find that’s not the case. These women are in fact leaning in. They’re speaking just as much regardless of the condition in which they’re in. The women and the minorities are going out of their way to try to get their opinions heard. They’re speaking just as much, as far as we can tell, based on the speaking turns and speaking time that we can observe. And so the failure to lean in can’t explain this gap in influence. So the common sort of Sheryl Sandberg “lean in” approach is that women just need to participate more and become equal participants in the process. That doesn’t seem to be supported by our research. Even when they try to do that, that doesn’t help them overcome this gap in influence. And so that’s kind of a depressing thing to discuss, that there isn’t much women actually can do to change those kinds of outcomes when they find themselves in these settings where they’re underrepresented. That it’s really men’s attitudes and men’s behavior that seems to be changing when women are in the minority versus women in the majority. So in our study, it’s men that are evaluating women as more influential when they are randomly assigned to be in a group with more women relative to when they’re in the group with just one woman. But of course, these underlying causes are really structural. So if you were to ask me, you know, what can organizations do to avoid those kinds of consequences for them, and I would say, “Well, number one is they need to hire more women.” Creating an environment in which women are no longer in the minority is certainly the direct implication of our research. However, that might be the more longer term goal. If organizations, say a tech firm, says over time, “We’re trying to hire more women, but we just don’t even have enough qualified women in the pipeline. What can we do now? How can we fix this given that women are still going to be in the minority for a while…” Then thinking about the structures of the teams and how they’re assigned, but also the norms within those teams? So for example, my co-author Chris Karpowitz has done some research in the past about the norms of deliberation and whether teams make decisions by majority rule, or whether teams make decision unanimously. That seems to be really important to women’s ability to contribute in environments in which they’re underrepresented. So maybe restructuring some of the team norms so that decisions have to be made unanimously, such that women’s voices are heard and they’re able to contribute even when they’re in the minority.
John: One thing I’ve been thinking when I read your paper and during our discussion is that there’s a similar cultural issue that affects teaching evaluations. And there’s at least some research that suggests that the negative bias that students may have in evaluating female professors can be overcome somewhat when students are made aware of the existence of this. And one nice thing about studies like yours is that it is making people more aware of this. But it would be interesting to see if students were given information about this at the start of their group formation, if that may affect the way in which group behavior is formed.
Olga: I am aware of those studies and I like them very much, because they show us one way, an easy nudge, which can change behavior, in this case, in the context of student evaluations of teaching. So in our study, of course, we try to keep the framing about students’ participation in this research, very neutral. We didn’t want them to be primed that this was a study about gender dynamics in groups and things like that. But I can envision future work thinking about the next step, which is what can be done to reduce this gender gap, what can be done to improve outcomes for women when they do find themselves in the minority, and one of those could be making students aware then making these patterns a lot more salient. Because honestly, if you probably ask a lot of the students whether they think that women in these groups are incurring any kind of penalty, they would probably say, “No.” The majority of the male students would probably not think that these things are happening. They’re happening in a subconscious basis, not through explicit discriminatory practices. It’s certainly possible that some male students are explicitly discriminating. But one measure that we have of that is how satisfied students are with their groups. And what we find is actually, regardless of the condition in which women are in, they report very high levels of satisfaction with their group. So even when they’re in the minority, and we can see that they’re incurring this really strong cost or deficit of influence, they still report being equally satisfied with their groups, and as happy with the group interactions as the women in the majority. So it seems that even the women themselves are not often recognizing that these deficits are occurring, let alone the male students in the group.
Rebecca: They’ve experienced it forever, it doesn’t seem different, right? [LAUGHTER]
Olga: That’s right, and this is not the first setting in which they’re experienced in this. There’s research showing that these kinds of patterns exist as early as school levels, where difference in competition is found as early as kindergarten, basically. And so the socialization that takes place even prior to college is probably conditioning women to feel that that is a normal kind of environment.
Rebecca: Your study reminds me a lot of conversations around all girls schools in K-12 and some of the benefits of that for women and also thinking about compositions of committees and things that might exist in professional environments where they’re trying to diversify, and they diversify by having token representation. And we often see that that can be problematic, but this is demonstrating other ways in which it can be problematic, which I think is a lot of interesting food for thought.
Olga: Yeah, absolutely. That was one of the biggest motivation… thinking about this is when you look at these policies, both private and public initiatives that are aiming to diversify these settings, like school boards or corporate boards, political assemblies, often, like you said, the solution is let’s just add one or two token women or minorities to the setting to help us be more diverse, and certainly we wanted to know what impact is that having on the women that are added, the women that become those token or lone members of the group and it’s not looking great. [LAUGHTER]
John: It’s a cultural issue and cultural changes tend to be slow. And as you said before, the only real solution is to have more balanced representation in all groups.
Olga: Absolutely. Yeah. And often, of course, what you hear, especially in the private sector is, “Well, it’s a meritocracy. Everybody can apply for these jobs, and we’re just not finding enough qualified women. And you know that certainly could be a valid concern in some stages of application process, but it is an important hurdle to overcome and think about how do we get more women into the funnel? How do we make sure that our women persist through the application process and actually make it into these jobs, because there are barriers at different levels, at different stages of that process that lead to these gender disparities in the share of women that go into these occupations, it’s not all choice. Choices are made, not in a vacuum, they’re made based on the constraints and information that people have. And so making these environments more appealing, more welcoming to women, should be an important objective of any organization that is struggling to increase diversity, gender diversity, in their rank and file.
Rebecca: As someone who teaches when an area of design that is also not balanced, [LAUGHTER] I teach in a more tech heavy side, it’s much more male dominated, because there’s more code and stuff involved and so historically, there’s less women, I’m thinking about all the group work that I do in my own classes, in the context of your research, and thinking about how productive and exciting it’s been to see some groups of all women, and what that looks like and what that feels like. But also having that little voice in the back of your mind saying maybe we need diverse teams that represent different kinds of people, because we’re designing for different kinds of people. And that, for the benefit of males in the class of interacting with women, maybe it benefits them, but they already have a benefit. And so that’s a really interesting consideration that I don’t think we often think about… not in a systematic way… or thinking about groups. I thought about majors and all kinds of things when I was formulating my groups, but I didn’t necessarily think about this.
Olga: Yeah, and I think that’s very common, especially in environments where there are serious binding constraints, you only have a few women. So I’m at BYU, and we have our share of women in the majors only about 20%. So any faculty trying to form group is going to be faced with these really serious constraints. One thing I would say is, in addition to this quantitative evidence that has been generated over the years showing how harmful it may be for women to be in the minority, there’s also, in our study, some qualitative evidence that we find. And since we’ve presented this study in different places, it’s been such an interesting experience, because you get these women just nodding their heads and saying, “I know exactly how this feels having been in the minority, and having compared my experience as a woman in the majority, just how much more heard and influential I feel in those kinds of settings.” So I think compiling qualitative evidence, pointing to the fact that it is significantly more difficult for women in the minority in these group settings to exert their influence and to get their voices heard.
John: Are you thinking of extending this research to other areas in terms of say race or other categories in which there may be similar effects?
Olga: Yeah, absolutely. So the original study certainly can only speak to gender, we have very few non-white students in the sample and can’t say very much since they weren’t randomly assigned across the group composition. But our goal long term is to look at whether these patterns extend beyond the gender domain. My guess is that we’re going to find very similar patterns for racial minorities, for example, who find themselves being underrepresented in many kinds of similar settings. They may even be exacerbated relative to the gaps that we find for women. And so we’re very interested, we’re in conversations with one firm and another institution trying to design a study that might work but this is a work in progress. And I hope it happens, because certainly we want to know whether other kinds of minorities find themselves in similar predicaments when they are underrepresented.
Rebecca: It also seems like it would be interesting to know whether or not, if you have multiple people from different underrepresented groups, if that somehow starts treating that more as a majority of underrepresented people, or if it’s just specific to a particular group at any given time.
Olga: Yeah, that’s a good point. One thing that we are doing is we do have a study in the field that is sort of following up from their original study, which includes the groups in which women are still in the minority, but they’re not the only woman. In our original setting there’s either one woman and four men, or three women and two men. So it’s not a symmetric kind of setting. And that’s by design, because there’s so few women in that program that if we created two women groups we wouldn’t have enough sample size to confidently say whether these results are statistically significant. But in the follow up study that we have been doing in the field, actually, for the last year and a half, we do have groups with two women and groups with two men. So we can compare sort of more symmetric, does it help to have another woman in a team? Or does it not make a difference, because you’re still in the minority. Some preliminary findings that we have, are that, unfortunately, it’s not tipping the scale… that unless women are in the majority, they’re still going to incur those deficits in terms of influence. And that’s supported by some of the prior laboratory research. But this is still ongoing… so, unfortunately, not the full findings yet. Another interesting extension of this work that we have started implementing, sort of by accident, or by necessity, rather, when COVID head and a lot of the group interactions have moved online… our entire lives have moved on to virtual settings… we wondered whether these same patterns would be exacerbated in virtual settings. There’s some anecdotal evidence that it’s even harder for him to get their voices heard in these kinds of settings. And so the study that we had been running in person has been turned into a study using Zoom as a platform. So we can now, at the end of this semester and next semester, say something about whether these patterns are different in online settings versus actual face-to-face settings, and what kinds of additional burdens may fall on the women when they’re having to influence outcomes or participate in the deliberative process in an online setting.
Rebecca: …sounds fascinating.
John: It’s a great natural experiment. …let me rephrase that… [LAUGHTER] we should probably not refer to the pandemic as a great experiment…
Olga: I know.
John: …but it does provide an interesting source of data on that issue, and virtual work is likely to become much more common in the future anyway.
Olga: Absolutely. It doesn’t seem like it’s going to go away. Even if the pandemic ended today, people are getting used to these kinds of interactions. There are advantages to them in terms of flexibility and the kinds of geographical constraints that no longer seem to apply. But they may also have these unintended hidden costs that I think are important to be able to quantify, particularly as it relates to these gender and racial disparities that already exist in a lot of these settings and workplaces.
Rebecca: So we always wrap up by asking: “What’s next?”
Olga: So this study has really led us to think carefully about these gender disparities, and to try to understand what kinds of interventions can help improve the outcomes for women. So the next step is certainly for us to try to test and evaluate the effectiveness of some of these interventions. So for example, I mentioned we’re doing a study in the field using Zoom as a platform for team meetings, we’re playing around and designing different kinds of changes in group norms, which operates through Zoom on, for example, who gets to start the conversation, or timing each participant in the group, so they know how long they’ve been speaking for… things that have been possible through technology, and trying to see whether those kinds of interventions will help improve the outcomes for women when they’re in the minority. So that’s one direction in which we are continuing this research agenda. And then another one, of course, is looking at other kinds of minority status. So particularly looking at race, we’re very interested in collaborating either with firms or other institutions that have ethnic or racial minorities, and are interested to know what implications do these settings have on their minority employees or students?
Rebecca: Looks like a lot of great work coming down the pike. I’m excited to hear what you find.
Olga: Thanks, Rebecca, thank you so much.
John: You’re doing some wonderful work, and I’m looking forward to seeing more of it in the future.
Olga: I really appreciate it. Thank you guys. It’s been a pleasure. Thanks so much.
Rebecca: Thank you.
John: If you’ve enjoyed this podcast, please subscribe and leave a review on iTunes or your favorite podcast service. To continue the conversation, join us on our Tea for Teaching Facebook page.
Rebecca: You can find show notes, transcripts and other materials on teaforteaching.com. Music by Michael Gary Brewer.