I am a staff learning scientist at Amplify. Previously I was a research scientist in the School of Interactive Computing at Georgia Institute of Technology and a postdoctoral researcher in the Institute for Intelligent Systems at University of Memphis. I received my PhD from Dartmouth College in 2018, where my research focused how retrieval practice (testing) influenced learning of materials that contained causal relations and the complexities of measuring such knowledge.
There is ample evidence that various interventions can improve memory, such as spacing study sessions in time and retrieval practice. However, most research in these areas provide little guidance on exactly how much to space, and which items should be practiced. This is especially true for learning more complex, educationally relevant materials (e.g., STEM content).
I develop and test computational models of learning, and use them to estimate the optimal difficulty at which to practice items to maximize learning of complex, educationally relevant content (e.g., STEM content). Importantly, this method generates specific prescriptions about exactly what to practice next that is sensitive to the practice history of the individual learner. Studying these issues has spawned other lines of research which include how to estimate student-level parameters (e.g., learning rate) in real time (and the consequences of not estimating them), the effects of practice difficulty on mind-wandering, and how to personalize feedback based on student errors.
Click the other tabs for more details on some of my projects and the 'Preprints' tab for manuscripts that explain the models and approach in detail.
Related interests:
AI agents to support learning within large-scale online educational contexts (e.g., MOOCs)
Automated and generalizable learner model development (see my LKT R package)
Optimally scheduling practice for complex learning materials (e.g., STEM topics)
Personalizing feedback based on prior types of errors
Real-time estimation of individual student parameters
Adaptive systematic (nonrandom) error correction due to model misspecification
Category learning generally, and the complex issues that arise when trying to optimize such learning.
The relationship between memory retrieval attempts and subsequent attentional allocation.
The causal structure of learning materials and relation to estimating learner knowledge.