Workshop on Longitudinal Item Response Models

Monday, October 7, 2013 - 1:00pm

Ronli Diakow (New York University) will conduct a workshop on Longitudinal Item Response Models. Abstract: Item response theory provides a sophisticated machinery for analyzing assessment data, but the emphasis is usually on individual assessments. Modern methods of longitudinal data analysis provide flexible and powerful ways to model change over time, but these models often treat the measured variables of interest as known. In this workshop, we will discuss the analysis of data from assessments given at multiple occasions using models that lie at the intersection of these traditions. We will address issues from both traditions such as the flexible treatment of time and measurement invariance. The primary statistical framework will be item response modeling. Connections will be made to ideas and equivalent models from the larger statistical framework of generalized latent variable modeling, in particular developments from hierarchical linear modeling and structural equation modeling. The first two hours will consist of lecture and discussion while the final hour will focus on an empirical application of the models in ConQuest, Stata, and MPlus.

Room: 2515 Tolman