Author(s):
Need: Despite decades of effort to diversify science, first generation (FG) students and Persons Excluded due to Ethnicity or Race (PEERs) remain underrepresented in science. One promising solution is interventions targeting students’ beliefs about their abilities (mindset) that shape students’ motivation, engagement, and academic outcomes. Although successful mindset interventions have yielded impressive outcomes, (e.g., increasing grades and persistence in science and closing achievement gaps), there has been considerable variability in the success of interventions. There is recent, growing evidence that the cultural context moderates the effects of interventions. We advocate for shifting away from a deficit model (i.e., “fixing” the student’s beliefs) towards an anti-deficit model, which focuses on the mindset culture that students navigate.Guiding Questions: We experimentally tested interventions targeting two levels: the mindset culture through instructor messages and students’ beliefs directly. The instructor-level intervention consisted of mindset messages in the syllabus and emails following exams (vs. control messages). The student-level intervention consisted of an assignment where students read an article about brain plasticity and writing a letter to a future student endorsing the notion that intelligence can improve (vs. a control focused on brain physiology). We conducted a large-scale (n = 5,197), fully-crossed, randomly-assigned controlled trial in introductory biology courses at 4 diverse institutions (an Historically Black University, an Hispanic-Serving and Asian American & Pacific Islander-Serving Institution; an emerging Hispanic Serving Institution, and a primarily white, First-Gen Forward Institution). Outcomes: We used hierarchical linear modeling to analyze students’ grades as a function of intervention conditions. Our analyses suggested different mechanisms and results at our Historically-Black University, which had a multiple smaller-enrollment sections, than our other 3 institutions, which each had a large section taught by a single instructor. In our large-section courses, we found that the instructor-level intervention improved the grades of PEERs and thus narrowed the equity gap. In the control condition, majority students earned over half a letter grade (7.12%) higher on the midterm exam than PEERs (F(1, 3384) = 82.85, p < .001). However, in the intervention condition, PEERs’ exam performance increased significantly (F(1, 3384) = 5.10, p = .024) while majority students’ performance was unchanged, reducing the gap by 36.9% (reduced to 4.49% difference). In our multiple small-section course, the student-level intervention improved exam grades for FG students (F(8,1543) = 6.45 p = .011). Additionally, both interventions separately increased FG students’ interest in biology (F(8,832) = 1.09 p = .024) and intent to persist in science careers (F(8, 832) = 3.77 p = .052). Broader Impacts: In this study, we make three important advancements. First, we test scalable, low-cost strategies to improve educational equity in introductory biology. Second, we conducted a large experiment in which students were randomized to condition within each section, allowing more power than other studies on this topic, which typically use different sections, instructors, or semesters as control groups and allowed us to control instructor-level characteristics. Third, our large sample included diverse institutions to increase generalizability and to specifically explore effects for PEER and FG students.
Coauthors
Anisha S. Navlekar, Texas Tech University, Lubbock TX; Elizabeth Canning, Washington State University, Pullman WA
