Conjoint Measurement, Error and the Rasch Model

Abstract

The theory of conjoint measurement was a genuine scientific revolution in measurement theory. For many years psychometricians and applied practitioners alike have contended that the Rasch model is a probabilistic version of this revolutionary theory. I disputed this claim from the perspective of representationalism. Michell's realist appraisal of the issue underscored the conclusions in my paper whilst Borsboom and Zand Scholten maintained that the Rasch model conceptually instantiates the theory of conjoint measurement. Nothing in either response refuted my conclusion. Nevertheless, relationships between the Rasch model and the theory of conjoint measurement remain an open research issue.