Identifiability of Spurious Factors Using Linear Factor Analysis with Binary Items

Samuel B. Green

Research output: Contribution to journalArticle

29 Citations (Scopus)

Abstract

The purpose of this study was to evaluate the ro bustness of some linear factor analytic techniques to violations of the linearity assumption by factoring product-moment correlations computed from data con forming to an extended, three-parameter logistic model of item responding. Three factors were crossed to yield 81 subcases: the number of underlying dimen sions (0, 1, or 2), the number of items (10, 15, 20, 25, 30, 35, 40, 45, or 50), and the number of subjects (100, 250, or 500). The mean eigenvalues for the sub cases were evaluated using parallel analysis and the scree technique. The mean eigenvectors were visually inspected. For almost all subcases with one or two un derlying dimensions, a single spurious factor was able to be identified using parallel analysis. However, in comparison with the nonspurious factors, it was small in magnitude and, in practice, factors of this relative size might be interpreted as trivial. It was concluded that researchers may have some confidence in inter preting linear factor analysis with binary items if they are using a test instrument that has been carefully de veloped.

Original languageEnglish (US)
Pages (from-to)139-147
Number of pages9
JournalApplied Psychological Measurement
Volume7
Issue number2
DOIs
StatePublished - 1983
Externally publishedYes

Fingerprint

Statistical Factor Analysis
factor analysis
Logistic Models
Research Personnel
factoring
confidence
logistics

ASJC Scopus subject areas

  • Psychology (miscellaneous)
  • Social Sciences (miscellaneous)

Cite this

Identifiability of Spurious Factors Using Linear Factor Analysis with Binary Items. / Green, Samuel B.

In: Applied Psychological Measurement, Vol. 7, No. 2, 1983, p. 139-147.

Research output: Contribution to journalArticle

@article{e1a01697786e440084bcbc5fa98d05a9,
title = "Identifiability of Spurious Factors Using Linear Factor Analysis with Binary Items",
abstract = "The purpose of this study was to evaluate the ro bustness of some linear factor analytic techniques to violations of the linearity assumption by factoring product-moment correlations computed from data con forming to an extended, three-parameter logistic model of item responding. Three factors were crossed to yield 81 subcases: the number of underlying dimen sions (0, 1, or 2), the number of items (10, 15, 20, 25, 30, 35, 40, 45, or 50), and the number of subjects (100, 250, or 500). The mean eigenvalues for the sub cases were evaluated using parallel analysis and the scree technique. The mean eigenvectors were visually inspected. For almost all subcases with one or two un derlying dimensions, a single spurious factor was able to be identified using parallel analysis. However, in comparison with the nonspurious factors, it was small in magnitude and, in practice, factors of this relative size might be interpreted as trivial. It was concluded that researchers may have some confidence in inter preting linear factor analysis with binary items if they are using a test instrument that has been carefully de veloped.",
author = "Green, {Samuel B.}",
year = "1983",
doi = "10.1177/014662168300700202",
language = "English (US)",
volume = "7",
pages = "139--147",
journal = "Applied Psychological Measurement",
issn = "0146-6216",
publisher = "SAGE Publications Inc.",
number = "2",

}

TY - JOUR

T1 - Identifiability of Spurious Factors Using Linear Factor Analysis with Binary Items

AU - Green, Samuel B.

PY - 1983

Y1 - 1983

N2 - The purpose of this study was to evaluate the ro bustness of some linear factor analytic techniques to violations of the linearity assumption by factoring product-moment correlations computed from data con forming to an extended, three-parameter logistic model of item responding. Three factors were crossed to yield 81 subcases: the number of underlying dimen sions (0, 1, or 2), the number of items (10, 15, 20, 25, 30, 35, 40, 45, or 50), and the number of subjects (100, 250, or 500). The mean eigenvalues for the sub cases were evaluated using parallel analysis and the scree technique. The mean eigenvectors were visually inspected. For almost all subcases with one or two un derlying dimensions, a single spurious factor was able to be identified using parallel analysis. However, in comparison with the nonspurious factors, it was small in magnitude and, in practice, factors of this relative size might be interpreted as trivial. It was concluded that researchers may have some confidence in inter preting linear factor analysis with binary items if they are using a test instrument that has been carefully de veloped.

AB - The purpose of this study was to evaluate the ro bustness of some linear factor analytic techniques to violations of the linearity assumption by factoring product-moment correlations computed from data con forming to an extended, three-parameter logistic model of item responding. Three factors were crossed to yield 81 subcases: the number of underlying dimen sions (0, 1, or 2), the number of items (10, 15, 20, 25, 30, 35, 40, 45, or 50), and the number of subjects (100, 250, or 500). The mean eigenvalues for the sub cases were evaluated using parallel analysis and the scree technique. The mean eigenvectors were visually inspected. For almost all subcases with one or two un derlying dimensions, a single spurious factor was able to be identified using parallel analysis. However, in comparison with the nonspurious factors, it was small in magnitude and, in practice, factors of this relative size might be interpreted as trivial. It was concluded that researchers may have some confidence in inter preting linear factor analysis with binary items if they are using a test instrument that has been carefully de veloped.

UR - http://www.scopus.com/inward/record.url?scp=84965404438&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84965404438&partnerID=8YFLogxK

U2 - 10.1177/014662168300700202

DO - 10.1177/014662168300700202

M3 - Article

AN - SCOPUS:84965404438

VL - 7

SP - 139

EP - 147

JO - Applied Psychological Measurement

JF - Applied Psychological Measurement

SN - 0146-6216

IS - 2

ER -