Unidimensionality in the Rasch model: How to detect and interpret

Unidimensionality, that is the items in a questionnaire measure only a single construct, is a fundamental requirement for the Rasch model. The paper deals with the detection of unidimensionality making use of principal components analysis of residuals and item fit statistics. Simulated bi-dimensional data sets are analyzed in order to find regularities in the behaviour of these statistical tools. The results are applied to a real database coming from the satisfaction section of the first national survey concerning the social service sector carried out in Italy.