Developing automated coding frameworks using Large Language Models to analyze student self-reflections and learning behaviors at scale, with a focus on maintaining theoretical fidelity and interpretability.
Designing and evaluating digital tools that support community college STEM students in developing self-regulated learning strategies, with particular attention to how these interventions affect diverse student populations.
Developing new statistical approaches that combine inverse propensity score weighting with boosting algorithms to better estimate causal moderation effects with binary treatments and continuous moderators.