Quantitative Methods

Throughout all of my work is an interest in quantitative methods. In particular, I enjoy working with growth modeling techniques, multi-level analyses, epidemiological statistics, probabilistic estimation methods for record linkage, geometric interpretations of multivariate statistics, and curvilinear and nonlinear modeling.

Some of this involves basic research. For example, in the area of developmental epidemiology, I am interested in applying developmental modeling techniques to epidemiological measures of effect, as well as probabilistic strategies for determining whether data records that differ in their key identifiers (e.g., a name or date of birth is different), in fact belong to the same individual. Other activities are more applied in nature and involve using these strategies in research. Below are references and links to articles using some of these tools, including mediational effects in structural equation models, multi-level models, growth modeling, cross-lagged panel analysis, probabilistic linking, and estimating attributable fractions in more complex models. My interest in quantitative methods has led to my role as a statistician or methodologist on a number of projects described in my collaborations.