Publications & Resources

Measurement Models as Narrative Structures

May 2006

Robert J. Mislevy and Chun-Wei Huang

Advances in cognitive research increase the need for assessment that can address the processes and the strategies by which persons solve problems. Several psychometric models have been introduced to handle claims cast in information-processing terms, explicitly modeling performance in terms of theory-based predictions of performance. Cognitively based item response theory IRT models incorporate features of items that influence persons’ responses, and relate these features to response probabilities. A number of these models are discussed in terms of the interplay between the formal characteristics of the probability models and the substantive narratives they support. Examples include the Linear Logistic Test Model, mixtures of IRT models, and multivariate structured IRT models.

Mislevy, R. J., & Huang, C.-W. (2006). Measurement models as narrative structures (CSE Report 680). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).