A simple model that describes the probability that, when a group of individuals are asked a number of questions, individual j makes a mistake in answering question k. The model, which was suggested by Rasch in 1960, was originally applied to the answers given in an oral reading test and has subsequently been applied as a model of the responses to questions in many different tests (particularly psychological tests). The model is a particular type of logit model and takes the form where Σ_jα_j=0 and Σ_kβ_k=0. See also item response function.