ABSTRACT
Regression models have a familiar suitability as explanatory models: a criterion or dependent variable is explained in terms of predictors or independent variables. In the explanatory item response modeling (EIRM) approach, item responses are seen as dependent variables to be explained by independent variables, such as properties of persons, items, and pairs of persons and items. Examples of person properties are gender, age, socioeconomic status, or perhaps treatment status, or location on another test variable. Examples of item properties depend on the domain of testing. For example, in a spatial ability test, the angle of the required rotation is such a property, while in a mathematical test, the number and the different kinds of numerical operations are interesting explanatory properties. Because of its explanatory nature, the EIRM approach has the intrinsic qualities to function as a vehicle for research based on substantive theories. It can help to meet a basic concern expressed bySijtsma (2012) “Ask what psychometrics can do for psychology.”
