Home  >>  Archives  >>  Volume 7 Number 1  >>  st0119

The Stata Journal
Volume 7 Number 1: pp. 22-44

Subscribe to the Stata Journal

Rasch analysis: Estimation and tests with raschtest

Jean-Benoit Hardouin
Department of Biomathematics and Biostatistics
Faculty of Pharmaceutical Sciences
University of Nantes, Nantes, France
Abstract.   Analyzing latent variables is becoming more and more important in several fields, such as clinical research, psychology, educational sciences, ecology, and epidemiology. The item response theory allows analyzing latent variables measured by questionnaires of items with binary or ordinal responses. The Rasch model is the best known model of this theory for binary responses. Although one can estimate the parameters of the Rasch model with the clogit or xtlogit command (or with the unofficial gllamm command), these commands require special data preparation. The proposed raschtest command easily allows estimating the parameters of the Rasch model and fitting the resulting model.
Terms of use     View this article (PDF)

View all articles by this author: Jean-Benoit Hardouin

View all articles with these keywords: raschtest, Rasch model, generalized estimating equations, conditional maximum likelihood method, marginal maximum likelihood method, Andersen Z test, van den Wollenberg Q1 test, R1c, R1m, fit tests, item response theory, U test, splitting test, item characteristics curves

Download citation: BibTeX  RIS

Download citation and abstract: BibTeX  RIS