My interests lie broadly in quantitative methods for the social, behavioral, and medical sciences. I am especially interested in statistical techniques for research synthesis (i.e., meta-analysis), and my methodological research has focused largely on multivariate meta-analysis for (functions of) correlation matrices. Additional interests include practical impediments to multivariate meta-analysis (e.g., missing observations or dependence information, aberrant observations), refinements of univariate techniques for single correlations (e.g., optimal weights, interval estimation), and strategies for other effect-size indices (e.g., subject-level aggregates as study-level covariates, effect estimates from degraded information, sampling variance for approximated effect estimates).
I also dabble in topics that fall under social cognition (e.g., self-discrepancy theory) and trait personality (e.g., Five Factor Model). My interests in these domains mainly concern statistical or other methodological issues, such as psychometrically sound strategies for measuring self-discrepancy and procedures for combining and comparing Big-Five factor results across studies.
Finally, I have had the good fortune to collaborate on projects in a variety of disciplines. Most of these efforts have involved meta-analysis: racial differences in self-esteem (with B. Gray-Little), antecedents of language acquisition style (with J. Jesberg), outcomes of exercise interventions for persons with chronic illnesses (with V. Conn). Others, however, have appealed to me because of novel statistical challenges (e.g., path analysis for 3-mode data on attitudes toward policies about an agricultural problem, with L. McCann).
When developed, this page will include links relevant to the aforementioned research domains.
Research Synthesis Methodology: