Rasch Model as Additive Conjoint Measurement (From Robust Estimation in Latent Trait Analysis - Final Report, 1980 - See NCJ-72747)

• NCJ Number: 72748
• Author(s): B Perline; B D Wright; H Wainer
• Date Published: 1980
• Length: 19 pages
• Grant Number(s): 78-NI-AX-0047
• Sponsoring Agency: US Dept of Justice
• Publication Format: Article
• Language: English
• Country: United States of America
• Note: Reprinted from Applied Psychological Measurement, V 3, N 2 (Spring 1979), P 237-255

Annotation

Rasch's psychometric model is presented as a special case of additive conjoint measurement and applied to the prediction of parole outcome.

Abstract

Because the theory of conjoint measurement has been formulated deterministically, there have been some difficulties in its application. This paper indicates that the Rasch model, which is a stochastic model, does not suffer from this fault. The exposition centers on the analyses of two data sets, each of which was analyzed using Rasch scaling methods, as well as some of the methods of conjoint measurement. Results are compared in using the different procedures in predicting parole outcomes with a sample of parolees. The statistical tests of fit and the tests of additivity imply that the data were not described well by an additive representation. However, despite the poor data fit, the Rasch model performed close to optimally; i.e., the minimum stress monotonic function which best produces additivity in these data was very close to a linear transformation of the logistic function. Tabular and graphic data and 29 references are provided. (Author abstract modified)