SciELO - Scientific Electronic Library Online

 
vol.2 issue2Taxonomy of personality descriptorsScales to measure health beliefs: construction and validation author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Avaliação Psicológica

Print version ISSN 1677-0471On-line version ISSN 2175-3431

Abstract

PASQUALI, Luiz  and  PRIMI, Ricardo. Basic theory of Item Response Theory: IRT. Aval. psicol. [online]. 2003, vol.2, n.2, pp. 99-110. ISSN 1677-0471.

The paper offers a historical view and the basic rationale of the modern theory in psychometrics: item response theory (IRT). This theory has its roots in the 1930, but it was fully developed only in the 1950's, and became the standard theory in psychometrics in the 1980's. The IRT is one of the latent trait modeling theories that appeared in the 1930's. Latent trait modeling theories assume that the human behavior, called observable variables, is dependent and caused by latent traits, the hypothetical variables. The IRT assume this modeling and expresses the relationship between these two types of variables through a mathematical equation called the logistic equation. This equation produces a curve called the item characteristic curve (ICC). This curve defines the items parameters (difficulty, discrimination) in terms of the level of the latent trait, symbolized as theta (q). The paper also discusses the advantages that IRT offers over the traditional theory in psychometrics, that is, the classical theory of tests (CTT).

Keywords : Item response theory; IRT; Psychometrics.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )

 

Creative Commons License