Modeling Polytomous Item Responses Using Simultaneously Estimated Multinomial Logistic Regression Models

Carolyn Jane Anderson, Jay Verkuilen, Buddy L. Peyton

Research output: Contribution to journalArticle

Abstract

Survey items with multiple response categories and multiple-choice test questions are ubiquitous in psychological and educational research. We illustrate the use of log-multiplicative association (LMA) models that are extensions of the well-known multinomial logistic regression model for multiple dependent outcome variables to reanalyze a set of items from a large political psychology survey. In particular, LMA models are shown to provide a useful analysis of items when the proper scoring rule for them is unclear. They also clearly reveal the performance of the items when instructions were altered to suppress "Don't know" responses. LMA models can be fit rapidly using commonly available software.

Original languageEnglish (US)
Pages (from-to)422-452
Number of pages31
JournalJournal of Educational and Behavioral Statistics
Volume35
Issue number4
DOIs
StatePublished - Dec 1 2010

Fingerprint

logistics
regression
political psychology
educational research
instruction
performance

Keywords

  • "don't know" responses
  • Bock's nominal response model
  • item response theory
  • log-multiplicative association models
  • multidimensional latent variable modeling
  • nominal data

ASJC Scopus subject areas

  • Education
  • Social Sciences (miscellaneous)

Cite this

Modeling Polytomous Item Responses Using Simultaneously Estimated Multinomial Logistic Regression Models. / Anderson, Carolyn Jane; Verkuilen, Jay; Peyton, Buddy L.

In: Journal of Educational and Behavioral Statistics, Vol. 35, No. 4, 01.12.2010, p. 422-452.

Research output: Contribution to journalArticle

@article{24a2daef88574f35a07c559625763038,
title = "Modeling Polytomous Item Responses Using Simultaneously Estimated Multinomial Logistic Regression Models",
abstract = "Survey items with multiple response categories and multiple-choice test questions are ubiquitous in psychological and educational research. We illustrate the use of log-multiplicative association (LMA) models that are extensions of the well-known multinomial logistic regression model for multiple dependent outcome variables to reanalyze a set of items from a large political psychology survey. In particular, LMA models are shown to provide a useful analysis of items when the proper scoring rule for them is unclear. They also clearly reveal the performance of the items when instructions were altered to suppress {"}Don't know{"} responses. LMA models can be fit rapidly using commonly available software.",
keywords = "{"}don't know{"} responses, Bock's nominal response model, item response theory, log-multiplicative association models, multidimensional latent variable modeling, nominal data",
author = "Anderson, {Carolyn Jane} and Jay Verkuilen and Peyton, {Buddy L.}",
year = "2010",
month = "12",
day = "1",
doi = "10.3102/1076998609353117",
language = "English (US)",
volume = "35",
pages = "422--452",
journal = "Journal of Educational and Behavioral Statistics",
issn = "1076-9986",
publisher = "SAGE Publications Inc.",
number = "4",

}

TY - JOUR

T1 - Modeling Polytomous Item Responses Using Simultaneously Estimated Multinomial Logistic Regression Models

AU - Anderson, Carolyn Jane

AU - Verkuilen, Jay

AU - Peyton, Buddy L.

PY - 2010/12/1

Y1 - 2010/12/1

N2 - Survey items with multiple response categories and multiple-choice test questions are ubiquitous in psychological and educational research. We illustrate the use of log-multiplicative association (LMA) models that are extensions of the well-known multinomial logistic regression model for multiple dependent outcome variables to reanalyze a set of items from a large political psychology survey. In particular, LMA models are shown to provide a useful analysis of items when the proper scoring rule for them is unclear. They also clearly reveal the performance of the items when instructions were altered to suppress "Don't know" responses. LMA models can be fit rapidly using commonly available software.

AB - Survey items with multiple response categories and multiple-choice test questions are ubiquitous in psychological and educational research. We illustrate the use of log-multiplicative association (LMA) models that are extensions of the well-known multinomial logistic regression model for multiple dependent outcome variables to reanalyze a set of items from a large political psychology survey. In particular, LMA models are shown to provide a useful analysis of items when the proper scoring rule for them is unclear. They also clearly reveal the performance of the items when instructions were altered to suppress "Don't know" responses. LMA models can be fit rapidly using commonly available software.

KW - "don't know" responses

KW - Bock's nominal response model

KW - item response theory

KW - log-multiplicative association models

KW - multidimensional latent variable modeling

KW - nominal data

UR - http://www.scopus.com/inward/record.url?scp=78649809776&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78649809776&partnerID=8YFLogxK

U2 - 10.3102/1076998609353117

DO - 10.3102/1076998609353117

M3 - Article

AN - SCOPUS:78649809776

VL - 35

SP - 422

EP - 452

JO - Journal of Educational and Behavioral Statistics

JF - Journal of Educational and Behavioral Statistics

SN - 1076-9986

IS - 4

ER -