Loglinear multidimensional IRT models for polytomously scored items

Henk Kelderman; Carl P.M. Rijkes; Carl Rijkes

doi:10.1007/BF02295181

Loglinear multidimensional IRT models for polytomously scored items

Henk Kelderman, Carl P.M. Rijkes, Carl Rijkes

Faculty of Behavioural, Management and Social Sciences

Research output: Contribution to journal › Article › Academic › peer-review

80 Citations (Scopus)

454 Downloads (Pure)

Abstract

A loglinear IRT model is proposed that relates polytomously scored item responses to a multidimensional latent space. The analyst may specify a response function for each response, indicating which latent abilities are necessary to arrive at that response. Each item may have a different number of response categories, so that free response items are more easily analyzed. Conditional maximum likelihood estimates are derived and the models may be tested generally or against alternative loglinear IRT models.

Original language	Undefined
Pages (from-to)	149-176
Journal	Psychometrika
Volume	59
Issue number	2
DOIs	https://doi.org/10.1007/BF02295181
Publication status	Published - 1994

Keywords

Multidimensional item response theory
Rasch model
multidimensional Rasch model
polytomous responses
METIS-135298
goodness-of-fit testing
loglinear model
IR-85754
partial credit model

Access to Document

10.1007/BF02295181

Kelderman_10.1007_BF02295181

Cite this

@article{8adf96cc2df64754ab646c8b5c495d48,

title = "Loglinear multidimensional IRT models for polytomously scored items",

abstract = "A loglinear IRT model is proposed that relates polytomously scored item responses to a multidimensional latent space. The analyst may specify a response function for each response, indicating which latent abilities are necessary to arrive at that response. Each item may have a different number of response categories, so that free response items are more easily analyzed. Conditional maximum likelihood estimates are derived and the models may be tested generally or against alternative loglinear IRT models.",

keywords = "Multidimensional item response theory, Rasch model, multidimensional Rasch model, polytomous responses, METIS-135298, goodness-of-fit testing, loglinear model, IR-85754, partial credit model",

author = "Henk Kelderman and Rijkes, {Carl P.M.} and Carl Rijkes",

year = "1994",

doi = "10.1007/BF02295181",

language = "Undefined",

volume = "59",

pages = "149--176",

journal = "Psychometrika",

issn = "0033-3123",

publisher = "Springer",

number = "2",

}

TY - JOUR

T1 - Loglinear multidimensional IRT models for polytomously scored items

AU - Kelderman, Henk

AU - Rijkes, Carl P.M.

AU - Rijkes, Carl

PY - 1994

Y1 - 1994

N2 - A loglinear IRT model is proposed that relates polytomously scored item responses to a multidimensional latent space. The analyst may specify a response function for each response, indicating which latent abilities are necessary to arrive at that response. Each item may have a different number of response categories, so that free response items are more easily analyzed. Conditional maximum likelihood estimates are derived and the models may be tested generally or against alternative loglinear IRT models.

AB - A loglinear IRT model is proposed that relates polytomously scored item responses to a multidimensional latent space. The analyst may specify a response function for each response, indicating which latent abilities are necessary to arrive at that response. Each item may have a different number of response categories, so that free response items are more easily analyzed. Conditional maximum likelihood estimates are derived and the models may be tested generally or against alternative loglinear IRT models.

KW - Multidimensional item response theory

KW - Rasch model

KW - multidimensional Rasch model

KW - polytomous responses

KW - METIS-135298

KW - goodness-of-fit testing

KW - loglinear model

KW - IR-85754

KW - partial credit model

U2 - 10.1007/BF02295181

DO - 10.1007/BF02295181

M3 - Article

SN - 0033-3123

VL - 59

SP - 149

EP - 176

JO - Psychometrika

JF - Psychometrika

IS - 2

ER -