More than a token photo: humanising scientists enhances student engagement

Despite broad consensus that highlighting counter-stereotypical scientist role models in educational materials promotes equity and success, the specific elements that make these materials effective remain untested. Are pictures of counter-stereotypical scientists enough to communicate to students that scientists come from a variety of backgrounds, or is additional information required? To parse the effects of including visual depictions and humanising information about scientists featured in biology course materials, we distributed three randomised versions of assignments over several academic terms across 36 undergraduate institutions (N > 3,700 students). We found that the inclusion of humanising information about scientists was key to increasing student engagement with the biology course materials. Structural equation modelling revealed that the extent to which students related to scientists mediated the positive effect of humanising descriptions on student engagement. Further, our results were strongest among students who shared one or more excluded identity(s) with the featured scientists. Our findings underscore the importance of providing students with examples of humanised and relatable scientists in classrooms. Rather than simply adding a photo to increase representation, showcasing scientists as actual people (enjoying hobbies, experiencing setbacks, etc.) is a promising intervention to engage students with identities systemically excluded from biology. 

Methods
These data describe survey responses from students immediately after working through a quantitative biology activity that featured a counter-stereoyptical scientist in one of three ways: 1) provided no information about the scientist (the control treatment), 2) included photos of the scientist (the visual treatment), or 3) included both photos and interview questions answered by the scientist about their experiences as a counter-stereotypical scientist (the humanizing treatment). Students in biology courses taught by 43 different instructors at 36 different US undergraduate universities and colleges participated in the study. Over the duration of the course, students engaged with three different quantitative biology activities from a single treatment and were surveyed after each activity. Our study included a total of 12 different quantitative biology activities, and the specific activities implemented in each course were chosen by the instructor to best match the course content. We collected demographic information from students and created student-scientist demographic matching categories based on their self-reported gender and race/ethnicity identities. The survey data are organized into two different datasets: the quantitative and qualitative datasets. The quantitative dataset includes student responses to survey items on a 7-point scale. The qualitative dataset includes codes for how students responded to the open-ended prompt, "Describe how you related to the featured scientist in the activity, if at all." The datasets were error-checked to remove all surveys that could not be assigned to a student. All students that did not consent to participate in the study were removed from the datasets.