Dr. Krull's research is primarily focused on the development, application, and extension of multilevel random coefficient models in the social sciences. Such models are typically appropriate for the analysis of data with individuals clustered within groups and for longitudinal data with repeated measures of individuals. Dr. Krull's current research projects involve simulation study combined with analyses of existing datasets to explore a number of different topics:
- multilevel analyses of sibling groups,
- multilevel modeling when some "groups" consist of only a single subject or the data are only partially nested,
- possible approaches to growth modeling when subjects change group membership over the course of a longitudinal study,
- optimal design issues in longitudinal studies, and
- tests of mediation in longitudinal data.
Dr. Krull regularly teaches graduate courses and workshops in regression analysis and introductory and advanced multilevel modeling.