Brian Junker's research spans a range of topics in statistics, including mixture and hierarchical models for multivariate discrete measures; nonparametric, semiparametric, and parametric inference for latent variable models; and interdisciplinary applications of statistics, particularly in education, psychology, and the social sciences. He has studied latent variable models employed in the design and analysis of standardized tests such as the SAT and GRE in order to assess whether they are the right tool for the problem.