Timothy R. Johnson

• Bayesian Statistics
• Choice and Ranking Data
• Missing Data
• Simulation-Based Inferential Methods
• Statistical Disclosure Control

Johnson is a professor of statistics in the Department of Mathematics and Statistical Science. He earned a doctorate in quantitative psychology and master's in statistics from the University of Illinois at Urbana-Champaign, and a master's and bachelor's in psychology from Western Washington University.

Response Style. Most statistical models for ratings elicited from human respondents implicitly assume that the use of the response scale is homogenous across respondents, despite evidence that there are considerable individual and cross-cultural differences (sometimes known as “response style”). For some time I have been working on item response models that either account individual or group differences in scale use. More recently I have been developing inferential methods that are robust with respect to such differences.

Coarsened and Aggregated Data. Procedures for data processing, dissemination, and statistical disclosure control sometimes results in particular missing data problems such as coarsening and aggregation. If not accounted for properly, this can compromise inference. But coarsening and aggregation can result in analytically and numerically intractable likelihood functions and posterior distributions. One of my areas of work is in the use of specialized Monte Carlo methods to circumvent these problems.

Collaborative Research. I am involved in a variety of collaborative projects with researchers in the social and natural sciences including psychology, business and economics, and natural resources and ecology.