Welcome

228 Lazenby Hall
Department of Psychology
1827 Neil Avenue
Columbus, OH 43210
Phone: 614 292 4940
Email: pek.5@osu.edu

I am an Assistant Professor in the Quantitative program within the Department of Psychology at Ohio State University.

Research Interests

My quantitative work is centered on the topic of quantifying the uncertainty inherent in results obtained from fitting latent variable models to data. Latent variable models include factor analysis models, structural equation models (SEM), structural equation mixture models (SEMM), multilevel models (MLM) and latent growth curve models.  One aspect of this research focuses on the issue of quantifying uncertainty due to cases, involving the identification of influential cases which unduly impact model results. Parameter estimates are another aspect of model results which carry uncertainty in the form of sampling variability, typically communicated by confidence intervals and confidence regions. Work in this area has focused on the development, estimation and evaluation of confidence regions for multiple parameters in latent variable models. A third aspect of this research concerns fungible parameter values, or a set of suboptimal parameter estimates obtained when model fit is slightly worse than optimal. Here, the suboptimal model fit associated with fungible values is of no practical difference compared to model fit associated with optimal parameter estimates. Fungible parameter values may exhibit large deviations from optimal parameter estimates, and the difference between fungible values and optimal estimates quantify the stability of parameter estimates across small changes to model fit. My research interests also extend to statistical graphics, where development has focused on visualizing and graphically communicating these various kinds of uncertainty.

Recently, my quantitative developments have been applied toward bridging the gap between methodological developments and their use by substantive (i.e., non-methods) researchers. This line of work has focused on the estimation and reporting of effect sizes and their confidence intervals, the application of data transformations, as well as aspects of research design (e.g., measurement validity and power analysis).