A Bayesian semiparametric item response model with dirichlet process priors

Kei Miyazaki, Takahiro Hoshino

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

In Item Response Theory (IRT), item characteristic curves (ICCs) are illustrated through logistic models or normal ogive models, and the probability that examinees give the correct answer is usually a monotonically increasing function of their ability parameters. However, since only limited patterns of shapes can be obtained from logistic models or normal ogive models, there is a possibility that the model applied does not fit the data. As a result, the existing method can be rejected because it cannot deal with various item response patterns. To overcome these problems, we propose a new semiparametric IRT model using a Dirichlet process mixture logistic distribution. Our method does not rely on assumptions but only requires that the ICCs be a monotonically nondecreasing function; that is, our method can deal with more types of item response patterns than the existing methods, such as the one-parameter normal ogive models or the two- or three-parameter logistic models. We conducted two simulation studies whose results indicate that the proposed method can express more patterns of shapes for ICCs and can estimate the ability parameters more accurately than the existing parametric and nonparametric methods. The proposed method has also been applied to Facial Expression Recognition data with noteworthy results.

Original languageEnglish
Pages (from-to)375-393
Number of pages19
JournalPsychometrika
Volume74
Issue number3
DOIs
Publication statusPublished - 2009 Sep
Externally publishedYes

Fingerprint

Dirichlet Process Prior
Characteristic Curve
Logistic Model
Logistics
Aptitude
Logistic Models
Model
Facial Expression Recognition
Logistics/distribution
Mixture Distribution
Dirichlet Process
Nonparametric Methods
Increasing Functions
Model Theory
Facial Expression
Statistical Models
Two Parameters
Express
Simulation Study
Estimate

Keywords

  • Dirichlet process mixture
  • Item characteristic curves
  • Item response theory
  • Semiparametric models

ASJC Scopus subject areas

  • Psychology(all)
  • Applied Mathematics

Cite this

A Bayesian semiparametric item response model with dirichlet process priors. / Miyazaki, Kei; Hoshino, Takahiro.

In: Psychometrika, Vol. 74, No. 3, 09.2009, p. 375-393.

Research output: Contribution to journalArticle

@article{9366c18937154d1481247e12893baefa,
title = "A Bayesian semiparametric item response model with dirichlet process priors",
abstract = "In Item Response Theory (IRT), item characteristic curves (ICCs) are illustrated through logistic models or normal ogive models, and the probability that examinees give the correct answer is usually a monotonically increasing function of their ability parameters. However, since only limited patterns of shapes can be obtained from logistic models or normal ogive models, there is a possibility that the model applied does not fit the data. As a result, the existing method can be rejected because it cannot deal with various item response patterns. To overcome these problems, we propose a new semiparametric IRT model using a Dirichlet process mixture logistic distribution. Our method does not rely on assumptions but only requires that the ICCs be a monotonically nondecreasing function; that is, our method can deal with more types of item response patterns than the existing methods, such as the one-parameter normal ogive models or the two- or three-parameter logistic models. We conducted two simulation studies whose results indicate that the proposed method can express more patterns of shapes for ICCs and can estimate the ability parameters more accurately than the existing parametric and nonparametric methods. The proposed method has also been applied to Facial Expression Recognition data with noteworthy results.",
keywords = "Dirichlet process mixture, Item characteristic curves, Item response theory, Semiparametric models",
author = "Kei Miyazaki and Takahiro Hoshino",
year = "2009",
month = "9",
doi = "10.1007/s11336-008-9108-6",
language = "English",
volume = "74",
pages = "375--393",
journal = "Psychometrika",
issn = "0033-3123",
publisher = "Springer New York",
number = "3",

}

TY - JOUR

T1 - A Bayesian semiparametric item response model with dirichlet process priors

AU - Miyazaki, Kei

AU - Hoshino, Takahiro

PY - 2009/9

Y1 - 2009/9

N2 - In Item Response Theory (IRT), item characteristic curves (ICCs) are illustrated through logistic models or normal ogive models, and the probability that examinees give the correct answer is usually a monotonically increasing function of their ability parameters. However, since only limited patterns of shapes can be obtained from logistic models or normal ogive models, there is a possibility that the model applied does not fit the data. As a result, the existing method can be rejected because it cannot deal with various item response patterns. To overcome these problems, we propose a new semiparametric IRT model using a Dirichlet process mixture logistic distribution. Our method does not rely on assumptions but only requires that the ICCs be a monotonically nondecreasing function; that is, our method can deal with more types of item response patterns than the existing methods, such as the one-parameter normal ogive models or the two- or three-parameter logistic models. We conducted two simulation studies whose results indicate that the proposed method can express more patterns of shapes for ICCs and can estimate the ability parameters more accurately than the existing parametric and nonparametric methods. The proposed method has also been applied to Facial Expression Recognition data with noteworthy results.

AB - In Item Response Theory (IRT), item characteristic curves (ICCs) are illustrated through logistic models or normal ogive models, and the probability that examinees give the correct answer is usually a monotonically increasing function of their ability parameters. However, since only limited patterns of shapes can be obtained from logistic models or normal ogive models, there is a possibility that the model applied does not fit the data. As a result, the existing method can be rejected because it cannot deal with various item response patterns. To overcome these problems, we propose a new semiparametric IRT model using a Dirichlet process mixture logistic distribution. Our method does not rely on assumptions but only requires that the ICCs be a monotonically nondecreasing function; that is, our method can deal with more types of item response patterns than the existing methods, such as the one-parameter normal ogive models or the two- or three-parameter logistic models. We conducted two simulation studies whose results indicate that the proposed method can express more patterns of shapes for ICCs and can estimate the ability parameters more accurately than the existing parametric and nonparametric methods. The proposed method has also been applied to Facial Expression Recognition data with noteworthy results.

KW - Dirichlet process mixture

KW - Item characteristic curves

KW - Item response theory

KW - Semiparametric models

UR - http://www.scopus.com/inward/record.url?scp=84895901434&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84895901434&partnerID=8YFLogxK

U2 - 10.1007/s11336-008-9108-6

DO - 10.1007/s11336-008-9108-6

M3 - Article

AN - SCOPUS:84895901434

VL - 74

SP - 375

EP - 393

JO - Psychometrika

JF - Psychometrika

SN - 0033-3123

IS - 3

ER -