Research
Methodological research grounded in applied problems in education and social science
My research has two connected strands: methodological research in quantitative modeling and measurement, and applied research in education and social science. The relationship is reciprocal. Applied studies reveal where existing methods are fragile or incomplete; methodological studies then develop, test, and translate better quantitative guidance for practice.
Research Program
Methodological Research
Bayesian multilevel and longitudinal methods
My primary methodological line studies prior specification, calibration, random effects, and inference when higher-level information is limited. My dissertation, Informative Variance Priors for Bayesian Multilevel Models, developed this program through simulation and large-scale assessment applications. It now extends to prior translation across software, Bayesian multilevel item-response models, random-slope covariance structures, and longitudinal dyadic models.
Measurement, psychometrics, and SEM
A second methodological line examines what happens when measurement data are sparse, imbalanced, or heterogeneous across groups and contexts. I study fit-index behavior, category collapsing, condition-specific decision rules, item response theory, measurement invariance, differential item functioning, and assessment fairness.
Applied Research
My applied work uses quantitative methods to answer substantive questions while identifying design and measurement problems that motivate new methodological research.
Learning, measurement, and evaluation
Large-scale assessment, mathematics achievement, educational measurement, causal evaluation, and explainable machine learning in educational settings.
Measurement, prevalence, and program evaluation
Survey design, prevalence estimation, latent-variable measurement, complex sampling, and program evaluation in research on child trafficking and related social outcomes.
Measurement and design across settings
Cross-context measurement, respondent-driven and link-tracing sampling, measurement invariance, and reproducible workflows for interdisciplinary research.
From Applied Problems to Methods
This applied-to-methods pathway is central to how I define my research program.
Current & Emerging Work
Computational Experiences in Introductory Mathematics: A Longitudinal Study of Mathematical Agency and STEM Progression
I lead the University of Georgia component and provide methodological leadership in measurement, psychometrics, and longitudinal modeling. The submitted UGA subaward request is $250,000.
Long-Term Agenda
My long-term program develops quantitative methods for settings where information is limited by small higher-level samples, sparse response patterns, or complex study designs. I aim to connect methodological development with applied research, reproducible tools, and graduate training.