Loglinear Models for Observed-Score Distributions

This chapter reviews fundamental statistical concepts and applications of loglinear models for estimating observed-score distributions. It reviews fundamental statistical concepts and applications of loglinear models for estimating observed-score distributions. The application of loglinear models to observed-score distributions encountered in psychometrics is statistical and flexible. For most psychometric test data, a large number of models are usually available that may be considered and evaluated using a large number of fit statistics. Covariance estimates are available for gauging the accuracy of the estimates of a loglinear model. The results of loglinear models can be useful not only for providing accurate and stable estimates of observed-score distributions, but also for interpreting the estimated distributions and conditional item scores from Item Response Theory models.