Item Response Theory (IRT) is a modern psychometric framework that models the probability of item responses as a function of one or more latent traits. This paper provides a comprehensive overview of the theoretical foundations, assumptions, and applications of IRT, focusing on its use in psychological measurement. The aim of this review is to present key IRT models for both dichotomous and polytomous items, compare them to Classical Test Theory (CTT), and discuss their advantages. The review explains item parameters (difficulty, discrimination, guessing), describes item and test characteristic curves, and discusses key assumptions such as unidimensionality and local independence. Special attention is given to the application of IRT across different areas of psychology and its contributions to psychometric practice, such as ensuring parameter invariance, enhancing measurement precision, and detecting differential item functioning. Challenges related to IRT, such as the need for larger samples, complex analyses, and specific issues with non-cognitive measures, are also addressed. In conclusion, despite its practical demands, IRT remains a fundamental tool for developing contemporary psychological instruments, offering greater accuracy and scientific rigor in psychological assessment.
