A Comparison of Multidimensional IRT Linking Methods

Recent work on measuring growth with categorical outcome variables has combined the item response theory (IRT) measurement model with the latent growth curve model and extended the assessment of growth to multidimensional IRT models and higher order IRT models. However, there is a lack of synthetic studies that clearly evaluate the strength and limitations of different multilevel IRT models for measuring growth. This study aims to introduce the various longitudinal IRT models, including the longitudinal unidimensional IRT model, longitudinal multidimensional IRT model, and longitudinal higher order IRT model, which cover a broad range of applications in education and social science. Following a comparison of the parameterizations, identification constraints, strengths, and weaknesses of the different models, a real data example is provided to illustrate the application of different longitudinal IRT models to model students’ growth trajectories on multiple latent abilities.

Download Full-text

Exploring the Correlation Between Multiple Latent Variables and Covariates in Hierarchical Data Based on the Multilevel Multidimensional IRT Model

Frontiers in Psychology ◽

10.3389/fpsyg.2019.02387 ◽

2019 ◽

Vol 10 ◽

Cited By ~ 1

Author(s):

Jiwei Zhang ◽

Jing Lu ◽

Feng Chen ◽

Jian Tao

Keyword(s):

Latent Variables ◽

Hierarchical Data ◽

Multidimensional Irt ◽

Irt Model

Download Full-text

ACT research report series: A comparison of two linking methods for multidimensional IRT scale transformations

PsycEXTRA Dataset ◽

10.1037/e428022008-001 ◽

2003 ◽

Author(s):

Kyung-Seok Min ◽

Jong-Pil Kim

Keyword(s):

Research Report ◽

Multidimensional Irt ◽

Scale Transformations

Download Full-text

A multidimensional IRT model for continuous and graded responses with error in persons and items

10.31124/advance.13227812.v1 ◽

2020 ◽

Author(s):

Pere Ferrando ◽

David Navarro-González

Keyword(s):

Measurement Precision ◽

Multidimensional Irt ◽

Irt Model ◽

Graded Responses ◽

Factor Analytic ◽

Continuous Responses ◽

Categorical Responses ◽

Ordered Categorical

This article proposes two multidimensional extensions of existing DMs: the M-DTCRM, intended for (approximately) continuous responses, and the M-DTGRM, intended for ordered-categorical responses (including binary). A rationale for the extension to the multiple-content-dimensions case, which is based on the concept of the multidimensional location index, is first proposed and discussed. Then, the models are described using both the factor-analytic and the IRT parameterizations. Procedures for (a) calibrating the items, (b) scoring individuals, (c) assessing model appropriateness, and (d) assessing measurement precision are finally discussed. The proposals are submitted to be of particular interest for the case of multidimensional questionnaires in which the number of items per scale would not be enough for arriving at stable estimates.

Download Full-text