Improving STEM Motivation and Retention with Postsecondary Classroom Interventions: A Meta-Analysis

Eben Witherspoon
American Institutes for Research

Need: Fewer students graduating with STEM degrees can hinder efforts to meet the growing demand for a workforce with STEM skills and experience (U.S. Bureau of Labor Statistics, 2020). Identifying interventions that influence postsecondary STEM retention decisions is a key effort in addressing this issue. While academic achievement is one important predictor of retention, various motivational factors also predict undergraduate persistence, even after controlling for prior measures of performance and academic ability (see Robbins et al., 2004). Prior meta-analyses of undergraduate STEM interventions have either examined classroom-based instructional innovations while focusing on achievement outcomes only (Freeman et al., 2014) or examining motivational interventions while excluding pedagogical and curricular innovations (Lazowski & Hulleman, 2016). This project builds on this earlier work by studying whether both motivational interventions (e.g., values affirmations) and instructional interventions (e.g., flipped classrooms) improve both retention and retention-related motivational outcomes.

Guiding Questions: Through a systematic meta-analysis of the effects of classroom-based motivational and instructional interventions in STEM undergraduate courses, this study seeks to understand the average effect these interventions have on retention in STEM and on motivational outcomes related to retention in STEM. Perhaps most importantly for practitioners, we will also explore factors contributing to the heterogeneity of these effects, identifying certain characteristics of interventions that have the greatest impact and the contexts and populations in which specific interventions are most effective.

Outcomes: At this early stage of the project, we have used litsearchr, a state-of-the-art automated search process in R (Grames, 2019), to systematically search the literature of several online databases. This process uses network analysis to adapt keywords to specific databases and capture the broadest number of relevant articles, resulting in identifying more than 18,000 abstracts. We have also developed an abstract screening protocol for determining the eligibility of articles for further review and potential inclusion in the meta-analysis. Initial screening results show an inclusion rate of 26%, indicating a large body of relevant literature meeting our abstract screening eligibility criteria.

Development of full-text screening and coding protocols are in progress. Some interesting challenges encountered so far include defining the boundary of a classroom when identifying “classroom interventions” (e.g., summer bridge programs), the treatment of samples of pre-professional students and STEM courses for non-majors (e.g., Chemistry for nursing majors), and determining the eligibility of attitudinal outcomes (e.g., enjoyment of the course).

Broader Impacts: This study will provide insight into the relative sizes of the effects of a range of classroom-based motivational and instructional interventions on undergraduate students’ retention and motivation to persist in STEM. Importantly, project findings will be disseminated using an open-access, interactive website that allows STEM researchers and practitioners to identify interventions that are particularly effective when selecting, designing, and administering interventions targeting a specific retention issue, student population, and STEM domain. Understanding which interventions support students’ decisions to persist in certain STEM courses, majors, and careers will provide another piece of the puzzle for universities aiming to increase and broaden STEM participation.


Crystal Aguilera, American Institutes for Research, CA; David I. Miller, American Institutes for Research, IL; Rebecca Steingut, American Institutes for Research, Washington, DC; Elli J. Theobald, University of Washington, WA