Quantitative lunch -- Andrew Lampi.

Item response theory (IRT) is a way of modelling participant’s responses to items that are presumed to tap into a given construct. IRT can model either participants trait levels (on a given trait) and predict their performance on an assessment, or it can model the extent to which a selection of items measures a given trait. Traditionally, IRT is used to understand performance on standardized assessments—such as college admission exams—or to assess the efficacy of psychological measurements—such as mental health inventories. There are other applications for IRT though, beyond these traditional uses. For example, in cognitive and social psychology, research into categorical perception (the study of how we perceive continuous stimuli discreetly) may also yield interesting information when approached from an IRT perspective. In this talk, I outline why approaching categorical perception tasks from an IRT perspective is intuitive and I provide a practical example of how we might accomplish this. In sum, IRT analysis of performance on psychophysical tasks such as categorical perception paradigms is effective, and has several applications for future studies.