Abstract:
|
Effective analysis of "big data" in education research has the potential to transform both research and practice, and has important applications for the already established field of Discipline-Based Education Research (DBER; NRC, 2012), a promising avenue for improving student performance and retention in STEM (García-Peñalvo et al., 2017). Additionally, to evaluate higher education data we must draw on the expertise of data scientists and statisticians who are trained to study the complexities inherent in education research (Talanquer, 2014). Developing robust methods to analyze educational data as well as disseminating those results are imperative to improving educational research. Linear quantile mixed effects models using M-estimation methods to model trajectories in STEM is used to assess equity in a new way for students and implement personalized interventions to help students succeed during their time at the university and identify possible issues in terms of equity. The impact of supplemental instruction is examined using quantile regression methods and the student equity index for students at a large scale university is used to evaluate diversity and equity on the campus.
|