349. Growth Mindset Messaging

First-generation college students, on average, have lower GPAs and higher dropout rates than continuing-generation students. In this episode, Elizabeth Canning, Makita White, and William B. Davis join us to discuss a growth-mindset intervention that has eliminated this equity gap in a large STEM class.

Elizabeth is an Associate Professor in the Psychology Department at Washington State University. Makita is a graduate student at WSU’s Experimental Psychology Program, and William is a Professor of Molecular Biology and the Interim Vice Provost for Academic Excellence and Student Achievement at WSU.

Show Notes


John: First-generation college students, on average, have lower GPAs and higher dropout rates than continuing-generation students. In this episode, we discuss a growth-mindset intervention that has eliminated this equity gap in a large STEM class.


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: …and features guests doing important research and advocacy work to make higher education more inclusive and supportive of all learners.


Rebecca: Our guests today are Elizabeth Canning, Makita White, and William B. Davis. Elizabeth is an Associate Professor in the Psychology Department at Washington State University. Makita is a graduate student at WSU’s Experimental Psychology Program, and William is a Professor of Molecular Biology and the Interim Vice Provost for Academic Excellence and Student Achievement at WSU. Welcome back, Elizabeth and Makita, and welcome, William.

Bill: Thank you. It’s a pleasure to be here.

John: Today’s teas are:… Elizabeth, are you drinking tea?

Elizabeth: I am drinking tea. I’ve learned from past podcasts with you that I should bring some tea along. So I’m drinking some sweet and spicy black tea today.

Rebecca: Sweet and Spicy… [LAUGHTER] sounds interesting.

Elizabeth: It has some cinnamon and some orange. It’s very good.

John: Very nice. And Makita?

Makita: I’ve got classic Earl Grey today.

Rebecca: Can’t go wrong with a classic. [LAUGHTER]

John: And Bill?

Bill:I have a Raspberry Zinger herbal tea today.

Rebecca: I’ve moved on to Awake tea… this time of day. [LAUGHTER]

John: We recorded another podcast much, much earlier today. So that’s probably needed a bit more. [LAUGHTER] And so that I won’t be awake all night tonight, as I have been for much of the last couple of nights, I have just a pure peppermint tea.

Rebecca: Probably a better choice for this time of day. [LAUGHTER]

John: And for this time of the year.

Rebecca: So we’ve invited you here today to discuss your March 2024 study: “Growth Mindset Messages from Instructors Improve Academic Performance Among First-Generation College Students,” which is quite a mouthful. [LAUGHTER] But before we discuss the study, can you talk a little bit about the difference in outcomes between first-generation students and continuing-generation college students?

Makita: Sure. So when we say first generation, we usually mean someone who’s the first in their family to attend college. For this study, we specifically defined it as a student for whom neither parent has earned a four-year college degree. And these students typically have to face and overcome a lot of social and cultural barriers in order to be successful in college. And although almost a third of the people who go to college are first-generation, they don’t always seem to do as well as continuing-generation college students. On average, they tend to have a little more difficulty adapting to college, they typically earn lower grades and they have higher dropout rates. They also act a little bit differently, they’re less likely to do things like go to office hours, or communicate with their instructors, whether that’s by email or in person to seek clarification on course materials. And just in general, they tend to be less likely to try to access helpful academic resources when we compare them with continuing-generation college students, which would be students for whom at least one parent has earned a four-year college degree. And on top of all of that, the students are also usually… I mean, they have to be… dealing with having less familial guidance than their peers do when it comes to navigating higher education. So if you’re a continuing-generation student, you have a parent who might tell you something like, “Hey, it’s a good idea to go to office hours,” or “You should be asking questions” or “there’s a financial aid department,” important stuff like that. They also may be dealing with a mismatch between the values and culture that they grew up in, compared with the values and expectations of an American university, which is typically very individualistic and may not be as supportive as what they are used to. And of course, they’re also more likely to be working, living off of campus, and they usually get less financial support from their families. But there are other factors that may exacerbate the difficulties that they’re dealing with. And this study is about one of those.

Bill:And one of the things that, just for context at our local institution to keep in mind, is WSU has a lot of, for lack of a better term, brand loyalty. It was not uncommon for me in my class to have fourth generation Cougars. And if you think about that legacy of your great-grandparents had attended this institution, as had other people, that sort of familial knowledge that gets passed along, is a huge benefit for that population of students as compared to the first-generation students as Makita was talking about.

John: One of the things that you mentioned, Makita, was differences in help-seeking behavior. And one of the studies on that is one that we talked about with you and Elizabeth in an earlier podcast, and we’ll include a link to that in the show notes. Could you describe this current study?

Elizabeth: Yeah, so this study we conducted in 2021, I believe. It was the second semester online, so the pandemic year. I know we’re all sort of crossing out that memory from our minds… [LAUGHTER] …that whole year. As instructors, it was a little tumultuous. But Bill and I got together and we designed this study together in his class. He was teaching at the time the introductory biology class at WSU. And what we decided to do is test out some actionable techniques that instructors can do to communicate a growth mindset to their students. So a lot of the research that my lab does looks at how growth mindset messages impact different students from different backgrounds. And at this time, there was a lot of suggestions about how to implement growth mindset ideas in your class, how to change course material, how to integrate it throughout, but there wasn’t really any experimental evidence of specific strategies that instructors can do. So we decided that we would just test it out in Bill’s class. So what we did in this study is we randomly assigned everybody in the course to one of two conditions: a control condition or a growth-mindset message condition. And after the first two major exams in the course, students received an email from Bill, and in the control condition they received a message that is pretty typical of an instructor after an exam has been conducted in their class. It says things like: “exam grades are posted online, here’s how we calculated them,”
“here’s what it means,” and then “feel free to talk to me in office hours, or ask questions about it.“ In the growth mindset message condition, we included all of that information, but we also included a fairly lengthy paragraph… I’ve gotten comments on… but it’s a big chunk of a paragraph that basically says, in Bill’s words, “Here’s what I believe about student performance and exam performance and improving in the class.” And there we integrated growth mindset theory, we talked about how abilities can be improved, we talked about how academic struggles are normal to experience, and that this struggle is part of the process of learning, that is something that is controllable. That you can put in effort that you can use different strategies to make improvements. And I think we used, Bill, if correct me if I’m wrong, we started with a draft of this email that you had already been doing in your class. And we just sort of infused it a little bit more. So Bill really wrote like the first draft of it, and then we kind of went back and forth a little bit to edit it and infuse it a little bit more with growth mindset. But it’s really written in the voice of the instructor. After we implemented this email after the first exam, we did it again after the second exam, changing the language a little bit. We assessed their performance across the semester. So we looked at their exam grades, we looked at their final course grade, and we found that students who were in the growth mindset condition earned higher grades in the course. And this was predominantly driven by first-generation college students. So first-generation college students performed much better in the course, when they received those growth-mindset messages from the instructor.

Rebecca: Can you talk a little bit about the effects and how significant they were?

Elizabeth: Yeah, so when we look at the group differences in the control condition, what we find that’s very consistent across big STEM courses across the university, that continuing-generation students are getting better grades than first-generation students. So in this course, they’re earning about three-fourths of a letter grade higher than first-generation students. So this is a really, really big gap between the continuing-generation students and the first-generation students. It’s not specific to this one course. So we see this in a lot of different datasets. But what was really remarkable in this study is that that difference in their grades was completely eliminated in the growth-mindset condition. So first-generation college students earned about three-fourths of a letter grade higher in the growth-mindset condition compared to the control condition. So this is a pretty large effect on final course grades. And of course, it’s pretty significant in terms of the curriculum that they’re a part of. Bill, if you want to talk a little bit about the students who took this course and how it plays a role in their career trajectories.

Bill:Yeah, thank you for that. I was looking forward to jumping in a little bit. And this is a little bit about my journey too, in this course. So, at this point, just for context, this was the 16th semester that I had taught this course. Largely I had been either the sole instructor or I was the primary instructor of record for the course. It’s taken by about 450 to 550 students each semester, little bit of differential between fall and spring in size, over 50 different majors, or pre-major students take this class. It’s nominally a freshman majors biology course. But there’s a substantial number of students who are neither biology majors or life science majors. And there’s a sizable number of students who are not freshmen. So, in reality, the highest population of my course normally were sophomores, freshmen were the largest population in that spring semester. And that was pretty typical, because it is oftentimes taken as the second semester of an introductory biology sequence. So yeah, and to be honest, over the years before this study, I had been trying to narrow those opportunity gaps between first-generation and continuing-generation students. I knew they existed. I had done many things with the course experimentally over the years, and never had really closed that gap. This was one of the first times, if ever, I saw that gap completely close in anything that I tried, and I tried tinkering in the lecture in the lab, big changes, small changes, a lot had gone on before this.

John: You mentioned that the class is normally about 450 students. Could you tell us a little bit more about the specific course in which you implemented this? You noted that this was during the pandemic. Was this offered remotely? Was it synchronous? Was it asynchronous? And how many students were there in this particular sample?

Bill:So, the course was taught in a synchronous Zoom manner. So both the lecture and the lab were taught using zoom. I believe, if memory serves, I had the largest Zoom Room on campus that semester. I can’t remember the exact number, I want to say my enrollment that semester was somewhere close to 550, or 600

Makita: The final sample is 417, I believe.
Bill:417. That’s right. That’s how many students, so it must have been around probably 450, at the beginning of the semester, or something like that. Yeah, we had synchronous Zoom every day with students. In fact, I learned pretty quickly how to crash Zoom my first semester I taught remotely. [LAUGHTER] So by then, it was a little bit more of a pro, and at least knew how to keep it running for 50 minutes at a time.

Rebecca: And true badge of honor.

Bill:[LAUGHTER] Well, there are many lumps that we took in the pandemic and learning how to teach and Zoom was one of them. [LAUGHTER]

John: I had 350 students in the fall of 2020, and it was not my most enjoyable teaching experience. What are the factors you controlled for? Because you did control for some other variables to separate out the effect of the growth-mindset messaging?

Elizabeth: Yeah, so we controlled for three things in our analytic models. We controlled for students’ prior college GPA… so this was just their self report… GPA that they’ve had in college. Since this was in the spring semester, all the freshmen had had a semester under their belts at that point, to have an idea of what their college GPA was. We included this just because we were looking at performance. So we wanted to filter out or control for their prior performance, that’s a big predictor of their performance in the class. We also controlled for their self-reported personal mindset beliefs. So we asked the students themselves, what do you believe about the nature of intelligence? Do you think it’s something that can improve? Do you think it’s something that you’re just born with? So we measure their own self-reported mindset beliefs, and we control for that and all of the analyses, because we want to look at the effect of the intervention above and beyond student’s own beliefs about ability. And then the last thing we controlled for was their race or ethnic status. So what they self reported to us on a survey. Here, we controlled for that because there is some overlap between race, ethnicity, and first-generation status. We did look at the different groups separately, we don’t have enough power to really look at race ethnicity as a moderator at the intervention. So we control for it in our analyses to look at the effect of first-gen status above and beyond race and ethnicity,

Rebecca: I can imagine, as an instructor who’s fiddled quite a bit in their class to overcome these equity gaps, that it might have been a little surprising [LAUGHTER] how much of a change happened with such a relatively small intervention.

Bill:It was shocking. It’s funny, because one of the things that Elizabeth has really talked about is how we took materials that I actually already possessed in some ways, and modified them slightly in some cases, and more radically in others, but just that transformation was enough to start to really make an impact. And you could see it happening. And that was amazing to me. I remember when I halved my opportunity gaps in my class with an intervention. And I remember dancing and celebrating, [LAUGHTER] and then this one was just on that scale. This was huge and amazing to me.

John: One of the interesting things in this study is you were looking at the effect in terms of grades on subsequent exams, but you didn’t find much of an effect on the second exam from the first intervention. But when you had two of those messages coming out, that’s where you found the relatively large effect. Why do you think the second message was having more of an effect than just a single message?

Elizabeth: Yeah, so we have a couple of theories. We think that one reason could be that they just need the message more than one time. So they need two doses or more than two doses to sort of really get people’s attention and make them change their behavior. That c ould be one reason. Another reason is that it might be enough to just have one message, but it would take time for behavior to change in order to see a difference in their performance. So it could be that after the first exam, they started engaging with the material differently and then that then sort of snowballed to affect their third exam scores, and then again, their final course performance, but we don’t really know exactly what’s going on. Another reason could be that towards the end of the semester, students get a little antsy, and they start waking up a little bit to, “Oh, I should probably be doing something different if I want to have a grade that looks like this.” So it could be timing, it could be dosage, it could be the effects take longer to materialize in terms of performance changes. One of the interesting things that we found here in this dataset is we did try to look at behavior in the way that we had access to, so we pulled all of their activity on the course website. And this semester, our campus was using Blackboard as the course sites. And we pulled all of the Blackboard data. So we looked at whether students in the growth mindset condition were engaging with Blackboard differently. And we found that they were. Students, on average, who were in the growth-mindset condition engaged with the course material 40 instances more, on average, than students in the control group. This is around a 12% increase in webpage engagement. So after they get this message from the instructor saying that intelligence is something that can grow, abilities can improve over time, and here are some strategies that you might implement to realize those changes, they’re going to the course website, they’re clicking on things, they’re re-reading their notes, they’re looking over the course lectures, they’re engaging more with the resources that are posted there. And then not surprising to any instructor ever, [LAUGHTER] that when students are engaging with the course materials that then lead to greater performance in the class. So we found a mediation effect through their behavior with the course material that was posted online. So this is one example of the behavior change that I was talking about earlier, that when they get this message, it can be inspiring, it can inspire a different type of behavior that can lead to better performance in the course.

Rebecca: Those are really strong click rates generally, [LAUGHTER] especially when we’re often complaining that students don’t read our emails or messages. [LAUGHTER]

Elizabeth: Yes. So, we did look at differences in Blackboard in terms of clicking on course material versus going to the page where all their grades were, like the gradebook. And we didn’t find any differences on their engagement with the gradebook. So it wasn’t that they were just tracking it more or anything like that. But it seems to be the case that they’re actually engaging with the substantive course materials more, on average, in the growth- mindset condition.

Rebecca: It’s just incredible. [LAUGHTER] Can you talk a little bit about… and we’re hinting towards this, obviously… what does this study suggest that faculty and institutions can do to improve outcomes of first-gen students, or maybe even, Bill, what you’ve continued or started to do?

Bill:I think, for me, what I took away from it was being very intentional with my students in communicating what I believe about them. And I mean that in the spirit of the old NPR show, or the segment that they used to do, you know, “This I believe,” where people would have those small snippets. I felt like this intervention was something along those lines where I could be authentic in communicating what I truly believed about their abilities. And I did it in a way that, with some small tweaks, obviously had these much deeper and larger payoffs. And I think that’s one of the real nuggets of wisdom here is that when you want to transform, and have better student outcomes, you want to narrow these opportunity gaps for students from certain backgrounds or experiences when they come in. Oftentimes, you don’t have to throw away what you’re already doing. Oftentimes, you already have nuggets of really important things that you’ve built over time. And all you need to do is to work with somebody who sees things just a little differently and can tweak them with you. And I think that partnership with Elizabeth and Makita was really critical for me. They were able to take things that I already had, and show me a different way to just modify what I was already doing. That messaging would go out to every student after that time point. When you see an effect that size, it almost becomes ethical to in fact, it is ethical, in my opinion that you don’t preclude any student from experiencing it. So that should be the way that everyone should be communicating with their students, in my opinion. And you want to make sure that it’s authentic as well, because I started with a set of materials that I had already written and sort of came from my heart, my own experience, that made the authenticity, I think, real whereas if I would have tried to have said something to students that I really truly didn’t believe about their abilities, I think they would have sniffed that out in a heartbeat. And we probably wouldn’t have seen the effect that we did.

Rebecca: It’s amazing what a little bit of intention can do in the design of something, especially because we often don’t realize how big of an impact a simple communication might have. We almost think of it as like a throw away, or a quick thing you might be doing, but you add a little intention to it, and it starts having a really different impact.

Elizabeth: Yeah, I think you mentioned earlier that students might not read the email, you might think that they’re not going to engage with it. And we also thought that, [LAUGHTER] so we think that in designing this intervention, we were really drawing on a lot of the intervention research that had been done previously. So research on wise interventions, mindset theory, all of that was integrated into the design of the message itself. So we started with Bill’s original emails that he was sending out to students and the messages that he had already curated for his class, and timing it in a time in the class in which students are receptive to that information, I think, is also really key. So we don’t know this empirically. But I would guess that if we were to send that email a couple of weeks before the exam, it would not have the same kind of impact. We send it directly when that performance information was posted online. And that’s when students are questioning themselves. They’re saying, “What does this score mean? Does this mean that I’m not good at this? Does this mean that I can continue? Does this mean that I’m smart or not smart? …and that uncertainty, that is the right time to infuse your message as an instructor to counteract those uncertain messages about ability. So I think the timing really matters. I think authenticity, like Bill said, really matters. Makita and I have been working with other institutions to implement a similar type of intervention with instructors. And before we did that Makita did a bunch of focus groups with the students at these institutions and showed them the messages that we used in Bill’s class, and we ended up changing them quite dramatically. The students gave us feedback like “nobody here would ever say something like that,” or “here’s what you need to say to reach students here in my class,” or “here’s what I will be receptive to.” So students in different contexts are going to respond to different messages. And so you really do want it to fit your own class, how you would speak, how you would talk to your students, the needs of the students in that moment might be different from the class they’re in. So it really is not a copy and paste type of situation where you were gonna copy what we did here and put it in your class. I think you do need to kind of go through the process of generating it for yourself and your own students.

Rebecca: I could imagine there would be an element of needing to evolve it over time, too, as your student population changes, and as you change as a teacher.

Elizabeth: Yeah, absolutely.

John: One of the things that I think is particularly nice about this study is that we see higher DFW rates by first-gen students in STEM classes in particular, and so many students end up leaving STEM fields that they intended to major in in their first couple of years. And that’s where the rate of return to education is the highest. And in general, first-gen students come from lower socioeconomic groups and lower family income groups, and if we do want to see more equity in outcomes, getting more first-gen students into the STEM fields could do a lot to help reduce some of those equity gaps in terms of overall outcomes.

Elizabeth: You mentioned the long term nature of these effects. I think it’s a great question. I think it’s something that we’re currently looking at in a project that we’re doing that’s funded by the National Science Foundation, where we’re looking at how these interventions in an introductory course can change career trajectories, that when you get a B instead of a C in a major introductory course, that’s gonna mean something different to that student, that’s going to set them on a path of following that next course sequence, that next step. So that’s the question that we’re asking and the research that we’re doing now, and hopefully, we’ll have some answers in a few years.

John: Peter Arcidiacono at Duke had done a study on the decision to stay in STEM fields. And he did look at some of these questions in terms of the effects on major choice based on the grades that were received. We can share a link to that study in the show notes. In the IRB proposal, did you include the possibility of following up these students in terms of their outcomes a few years later, in terms of retention rates for the students?

Elizabeth: That’s a great question. I believe in terms of IRB, we had students consent to their academic records. So we should be able to go back and look at transcripts from the students who participated in the class. We could look at course taking, we could look at graduation rates, we could look at major. If these students were freshmen/sophomores in 2021, then we’re probably around the time where they would be finishing up so maybe in the next year we could.

Bill:I was just thinking the same. Now’s the perfect time to go back and take a look, because we’ve had enough time since the intervention to see what’s happened. And actually I was gonna go back to one of your earlier points. You were talking about statistics earlier. Basically, this is really the only way that we’re going to increase the number of STEM students going forward. If first-generation students at WSU are 40% of our population today, that means if we don’t improve their outcomes, then we’re going to continue to see the declines in STEM outcomes and STEM graduates that we are trying so desperately to fight against, and also, we’re not going to have the representation in our STEM graduates of the population of the US that we want to see.

Rebecca: I think in addition to the growth-mindset piece, I’m sure that the authentic nature of the messages also just showed deep care. And that can be really important for first-generation students who need that ally or advocate for them in their college journey. So I can imagine those messages coming from Bill in Bill’s voice that make them feel like “Yeah, you can do this, I care about you,” from that perspective is also really important.

John: And there was a study done by a couple of economists, entitled “My Professor Cares,” where they did follow up students four years later, and they did find a significant effect in retention rates four years later by a similar type of intervention, I don’t think it was quite as much growth-mindset focused, but just sending a signal that the professor cared about the students outcomes, a very simple message, made a significant difference in student’s grades as well as longer term outcomes. So it’s nice to see more studies of this, especially focused on issues such as growth mindset.

Elizabeth: Yeah.

Rebecca: So we always wrap up by asking: What’s next?.

Elizabeth: So Makita and I and our lab have been working for the past couple years on following up on these data in different institutions. We used this study as pilot data in a grant application. And now we’re sort of scaling up to other institutions. We have a project with a really large HBCU, a Hispanic-Serving Institution, WSU is continuing to be involved as well, an Asian and Pacific Islander Serving Institution to where we have a number of instructors who vary in different characteristics, different class compositions, and we’re looking at these instructor messages in an experiment in these different contexts to see what matters. Does it help students develop their own growth mindset? Does it change their behavior? Does it work in some contexts, but not others? Do the instructor characteristics matter? So far, in the past two years, we’ve collected data from around 10,000 students who have been involved in this intervention project, and we’re currently in data analysis phase. So stay tuned for future studies and papers that come out of that data set.

John: Makita?

Makita: Well, what’s next for me is Elizabeth and I are actually going to meet right after this to do a debrief and to discuss some of the analyses that I just ran on a study that will be involved in my dissertation.

Rebecca: And how about you, Bill?

Bill:For me, I think about: “How do I proselytize about this work?” Because I think about what little changes make a big difference in an institution. We talk a lot about student retention, we talk a lot about student outcomes, we talk about creating the global citizenry that we want to see emerge from higher education. And the one experience that every student has in common, whether they are a first-generation a continuing-gen, or from some other background, is they have to take classes, and they have to be in classroom environments every day of the week, for the most part, over a course of four years. And I think about what would it look like to have interventions like this, that have evidence that back them, that students get inoculated with repeatedly over multiple time points with different voices, and that faculty engagement in the process of helping our students become better, to find success and to move forward in their careers? To me, that’s the magic of this. I’ve been talking to other people about it and I think about what would a curriculum look like that had small things like this? And how could it transform what we do? And I think that’s a vision that I can buy into, and I hope others are gonna come along for that ride.

Rebecca: I think the scope of the interventions that you’re talking about, that have this kind of impact, are really easy to buy into. Because usually you’re thinking about some drastic, giant investment of time, is what most people are thinking about. And that often is what prevents action from happening. So I’m excited, excited to share this and to hear what comes of your bigger study.

John: And one thing we appreciate is when we see things like this posted on social media so we can hear about it [LAUGHTER] and we can share it out a little bit further. [LAUGHTER] And I hope you’re presenting this at various conferences to reach a larger audience.

Elizabeth: Yeah, we will be presenting some of this work at SABER this summer. So that is the Society for the Advancement of Biology education research, and we’ve presented it in some other psychology conferences throughout the past year as well.

John: Excellent.

Bill:And I gave a plenary talk at the summer educational meeting for the ASBMB, American Society for Biochemistry and Molecular Biology. So I gave a little preview of what we had found. And that generated a lot of discussion and excitement, actually people asking for additional resources and ideas and things. So I think there’s a receptive audience out there.

Rebecca: That’s wonderful.

John: Well, thank you for joining us. It’s always great talking to you and it’s great to hear about this newest study and to meet you, Bill. We’re looking forward to hearing more about how things go in the future.

Elizabeth: Thank you so much. Thanks for having us.

Rebecca: I hope you’ll reach out when your next study is done. [LAUGHTER]


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.