An approach to the theory of psychological testing based on the assumption that the probability of a particular response to a test item is a joint function of one or more characteristics of the individual respondent and one or more characteristics of the test item. The relationship of these parameters to the probability of a particular response is defined by an item response function, and their values are estimated from the observable responses of respondents to the test item. It is the principal alternative to classical test theory. Also called latent trait theory. See also Rasch scale. IRT abbrev.