Investigating latent constructs with item response models: A MATLAB IRTm toolbox
Open Access
- 1 November 2009
- journal article
- Published by Springer Nature in Behavior Research Methods
- Vol. 41 (4) , 1127-1137
Abstract
Item response theory (IRT) models are the central tools in modern measurement and advanced psychometrics. We offer a MATLAB IRT modeling (IRTm) toolbox that is freely available and that follows an explicit design matrix approach, giving the end user control and flexibility in building a model that goes beyond standard models, such as the Rasch model (Rasch, 1960) and the two-parameter logistic model. As such, IRTm allows for a large variety of unidimensional IRT models for binary responses, the incorporation of additional person and item information, and deviations from common model assumptions. An exclusive key feature of the toolbox is the inclusion of copula IRT models to handle local item dependencies. Two appendixes for this report, containing example code and information on the general copula IRT in IRTm, may be downloaded from brm.psychonomic-journals.org/content/supplemental.Keywords
This publication has 25 references indexed in Scilit:
- Copula Functions for Residual DependencyPsychometrika, 2007
- Examples in which misspecification of a random effects distribution reduces efficiency, and possible remediesComputational Statistics & Data Analysis, 2004
- Explanatory Item Response Models2004
- 2. Random-Effects Modeling of Categorical Response DataSociological Methodology, 2000
- A parametric model for local dependence among test items.Psychological Methods, 1997
- Local Dependence Indexes for Item Pairs Using Item Response TheoryJournal of Educational and Behavioral Statistics, 1997
- Conditional Association and Unidimensionality in Monotone Latent Variable ModelsThe Annals of Statistics, 1986
- Estimating the Parameters of the Latent Population DistributionPsychometrika, 1977
- The linear logistic test model as an instrument in educational researchActa Psychologica, 1973
- Fitting a Response Model for n Dichotomously Scored ItemsPsychometrika, 1970