Three Research Questions Guide Our Work

Engineering and computing education remains a critical ingredient for US competitiveness, workforce development, and technological supremacy now and into the future.  Understanding the ways in which students succeed and fail in STEM majors, and developing powerful ways to support them, will pay dividends for our students, our institutions, and our nation. Predictive models for student academic performance largely rely upon cognitive measures of achievement such as high school GPA, SAT scores, and similar measures of past performance. These models have consistently demonstrated that only a reasonably small part (R2 ~ 0.25) of the total variance in predicted academic performance is explainable using cognitive factors. More recent work includes non-cognitive and affective (NCA) variables (such as mindset) to improve the predictive power of academic performance models. Nonetheless, there remain substantial gaps in our understanding of how NCA profiles of STEM students can be used to support their academic success. The central intellectual contribution of this research-to-practice project is the development, implementation, and evaluation of NCA-based interventions for diverse STEM students in multiple settings.

This multi-institution research team engages student affairs practitioners in the development and delivery of NCA-based interventions. The unique coalescence of expertise and experience among the team strongly promotes the multiple perspectives that are required to deploy interventions with diverse students in different settings. Importantly, the student affairs collaborations allow access to information about life events (‘obstacles’) faced by students, and this research explicitly connects student academic performance to such obstacles, as mediated by the NCA profile. This obstacle data, and its connection to both NCA profile and academic performance, represents a truly unique and deeply valuable contribution of this research program.

This research is guided by three crucial research questions:

  • RQ1. What are the NCA profiles of engineering and computing students, and to what extent do profiles vary by institution, academic program, demographics, or over time?
  • RQ2. In what ways are NCA factors predictors of academic performance, and how do they mediate a student’s response to academic or personal obstacles they may face?
  • RQ3. To what extent can NCA-based interventions improve academic performance and the perceived quality of the undergraduate experience, and how do students at different institutions experience those interventions?

The project has important intellectual merit because it is the first project to systematically examine student academic performance in the face of specific obstacles as mediated by their NCA profile and cognitive makeup. This project leverages the important role of academic researchers and student affairs practitioners as co-educators who promote engaged student learning. This research will fill important gaps in the literature through the explicit inclusion of student-level obstacle data in academic performance models. The mixed-methods approach described here integrates diverse perspectives and the local culture/environment at each partner institution into the intervention design and delivery. This is the first research program of its kind to connect obstacles, NCA profile, and academic performance in a rigorous way.

The project has broader impacts because operationalization of the ‘same’ intervention in multiple settings, and recognizing the role of local context in the implementation and outcomes, has broad relevance to all who support STEM students. The role of both traditionally-defined and ‘latent’ diversity in answering the research questions holds important implications for the research and practitioner communities alike. This research demonstrates that powerful alliances between academic and student affairs experts are required to shape effective interventions supporting student success.

This material is based upon work supported by the National Science Foundation under Grant No. DUE-1626287 (Purdue), DUE-1626185 (Cal Poly), and DUE-1626148 (UTEP). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.