I primarily study methods for missing data treatment. I also collaborate extensively with social-, behavioral-, and health-scientists.

My research generally falls into one of two broad categories: methodological work developing and evaluating methods to treat missing data or collaborations with substantive researchers from the social-, behavioral-, and health-sciences.

In my methodological work, I mostly explore ways of treating missing data (i.e., data with values in the population but unobserved entries in the sample). Missing values cause myriad problems for statistical modeling, inference, and prediction, so treating missing data is a crucial step in any data analysis. Currently, my work focuses on developing and evaluating multiple imputation techniques. In particular, I am interested in ways of simplifying the process of specifying the underlying imputation models.

In my substantive collaborations, I provide statistical expertise to research teams from social-, behavioral-, and health-sciences. For example, in this project, I helped design a novel planned missing data protocol for non-survey data; in this project, I helped evaluate the efficacy of the Kansas Intensive Permanency Project, and in this project, I helped develop norms for the Supports Intensity Scale. I particularly enjoy collaboration because I like contributing to a group effort, and the diversity of research topics keeps things interesting. As such, I have to agree with John Tukey’s purported quip:

The best thing about being a statistician is that you get to play in everyone’s backyard.

I strive to follow open-science principles in my work. So, open-access flavors of most of my published work are available via the Utrecht University Repository, my Google Scholar profile, or my ResearchGate page.