A. Agresti, COMPUTING CONDITIONAL MAXIMUM-LIKELIHOOD-ESTIMATES FOR GENERALIZED RASCH MODELS USING SIMPLE LOGLINEAR MODELS WITH DIAGONALS PARAMETERS, Scandinavian journal of statistics, 20(1), 1993, pp. 63-71
Generalized Rasch models for multiple-response items proposed by Ander
sen (1973) are related to quasi-symmetric loglinear models. The loglin
ear models are obtained by treating subject parameters in the Rasch mo
dels as random effects. Fitting the loglinear models yields estimates
of item parameters in the generalized Rasch models that are also condi
tional maximum likelihood estimates when the subject effects are treat
ed as fixed. For models that apply naturally when there are ordinal re
sponse categories, the related loglinear models are simple quasi-symme
tric models having diagonals parameters. Our results generalize Tjur's
(1982) observation about the connection between binary-response Rasch
models and loglinear models.