Explanatory Item Response Modeling (EIRM)

Collaborating Institutions: K. U. Leuven, Belgium; Berkeley Evaluation & Research (BEAR) Center, UC Berkeley

K. U.  Leuven Participants: Paul De Boeck

BEAR Participants: Mark Wilson, Xiaohui Zheng, Leah Walker

Explanatory item response models conceptualize item response theory (IRT) models as a class of generalized linear and nonlinear mixed models and provide a framework for extensions of the standard IRT models. The framework encompasses a broad range of item response models that relate item responses to a large variety of external explanatory variables representing characteristics of items, of persons, and of combinations of persons and items. The framework was initially explored in Explanatory Item Response Models: A Generalized Linear and Nonlinear Approach where standard IRT models and beyond were expressed within the framework and illustrated using real data examples.

The current EIRM project seeks to formalize a representation of a wide range of EIRMs. More complicated IRT models will be explored to emphasize their explanatory uses and psychometric properties. Resulting models will be related to the class of generalized linear and nonlinear mixed models and placed within an expanded framework. The framework expansion will be accompanied by additional illustrative examples. Estimation of the resulting models will be proposed using existing software for generalized linear and nonlinear mixed models.

Website: http://www.springer.com/west/home/generic/search/results?SGWID=4-40109-22-26922428-0&referer=www.springeronline.com

Publications

De Boeck, P. & Wilson, M. (Eds.). (2004). Explanatory item response models: A generalized linear and nonlinear approach. New York: Springer-Verlag.