I have been studying multilevel regression and latent variable models - including latent growth models, parallel process growth models, SEM, factor analysis, latent class models, latent transition models, and growth mixture models - for seven years.  I'm especially interested in models for non-normal data.  Most of my recent contributions to research reports have been in the areas of latent variable models and multilevel regression models, including generalized linear models for longitudinal data.  Current areas of study include survival analysis; the analysis of complex sample data; multiple imputation; and the use of propensity scores for reducing bias.