Type I and Type II Error Rates and Overall Accuracy of the Revised Parallel Analysis Method for Determining the Number of Factors

Samuel B. Green, Marilyn Thompson, Roy Levy, Wen Juo Lo

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Traditional parallel analysis (T-PA) estimates the number of factors by sequentially comparing sample eigenvalues with eigenvalues for randomly generated data. Revised parallel analysis (R-PA) sequentially compares the kth eigenvalue for sample data to the kth eigenvalue for generated data sets, conditioned on k− 1 underlying factors. T-PA and R-PA are conceptualized as stepwise hypothesis-testing procedures and, thus, are alternatives to sequential likelihood ratio test (LRT) methods. We assessed the accuracy of T-PA, R-PA, and LRT methods using a Monte Carlo approach. Although no method was uniformly more accurate across all 180 conditions, the PA approaches outperformed LRT methods overall. Relative to T-PA, R-PA tended to perform better within the framework of hypothesis testing and to evidence greater accuracy in conditions with higher factor loadings.

Original languageEnglish (US)
Pages (from-to)428-457
Number of pages30
JournalEducational and Psychological Measurement
Volume75
Issue number3
DOIs
StatePublished - Jun 6 2015

Fingerprint

Type II error
Error Rate
Testing
Likelihood Ratio Test
Eigenvalue
hypothesis testing
Hypothesis Testing
testing procedure
Alternatives

Keywords

  • factor analysis
  • parallel analysis
  • revised parallel analysis

ASJC Scopus subject areas

  • Developmental and Educational Psychology
  • Education
  • Applied Psychology
  • Applied Mathematics

Cite this

@article{ee6966c4b7bb4359a0c69e9a37cb0e86,
title = "Type I and Type II Error Rates and Overall Accuracy of the Revised Parallel Analysis Method for Determining the Number of Factors",
abstract = "Traditional parallel analysis (T-PA) estimates the number of factors by sequentially comparing sample eigenvalues with eigenvalues for randomly generated data. Revised parallel analysis (R-PA) sequentially compares the kth eigenvalue for sample data to the kth eigenvalue for generated data sets, conditioned on k− 1 underlying factors. T-PA and R-PA are conceptualized as stepwise hypothesis-testing procedures and, thus, are alternatives to sequential likelihood ratio test (LRT) methods. We assessed the accuracy of T-PA, R-PA, and LRT methods using a Monte Carlo approach. Although no method was uniformly more accurate across all 180 conditions, the PA approaches outperformed LRT methods overall. Relative to T-PA, R-PA tended to perform better within the framework of hypothesis testing and to evidence greater accuracy in conditions with higher factor loadings.",
keywords = "factor analysis, parallel analysis, revised parallel analysis",
author = "Green, {Samuel B.} and Marilyn Thompson and Roy Levy and Lo, {Wen Juo}",
year = "2015",
month = "6",
day = "6",
doi = "10.1177/0013164414546566",
language = "English (US)",
volume = "75",
pages = "428--457",
journal = "Educational and Psychological Measurement",
issn = "0013-1644",
publisher = "SAGE Publications Inc.",
number = "3",

}

TY - JOUR

T1 - Type I and Type II Error Rates and Overall Accuracy of the Revised Parallel Analysis Method for Determining the Number of Factors

AU - Green, Samuel B.

AU - Thompson, Marilyn

AU - Levy, Roy

AU - Lo, Wen Juo

PY - 2015/6/6

Y1 - 2015/6/6

N2 - Traditional parallel analysis (T-PA) estimates the number of factors by sequentially comparing sample eigenvalues with eigenvalues for randomly generated data. Revised parallel analysis (R-PA) sequentially compares the kth eigenvalue for sample data to the kth eigenvalue for generated data sets, conditioned on k− 1 underlying factors. T-PA and R-PA are conceptualized as stepwise hypothesis-testing procedures and, thus, are alternatives to sequential likelihood ratio test (LRT) methods. We assessed the accuracy of T-PA, R-PA, and LRT methods using a Monte Carlo approach. Although no method was uniformly more accurate across all 180 conditions, the PA approaches outperformed LRT methods overall. Relative to T-PA, R-PA tended to perform better within the framework of hypothesis testing and to evidence greater accuracy in conditions with higher factor loadings.

AB - Traditional parallel analysis (T-PA) estimates the number of factors by sequentially comparing sample eigenvalues with eigenvalues for randomly generated data. Revised parallel analysis (R-PA) sequentially compares the kth eigenvalue for sample data to the kth eigenvalue for generated data sets, conditioned on k− 1 underlying factors. T-PA and R-PA are conceptualized as stepwise hypothesis-testing procedures and, thus, are alternatives to sequential likelihood ratio test (LRT) methods. We assessed the accuracy of T-PA, R-PA, and LRT methods using a Monte Carlo approach. Although no method was uniformly more accurate across all 180 conditions, the PA approaches outperformed LRT methods overall. Relative to T-PA, R-PA tended to perform better within the framework of hypothesis testing and to evidence greater accuracy in conditions with higher factor loadings.

KW - factor analysis

KW - parallel analysis

KW - revised parallel analysis

UR - http://www.scopus.com/inward/record.url?scp=84930526296&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84930526296&partnerID=8YFLogxK

U2 - 10.1177/0013164414546566

DO - 10.1177/0013164414546566

M3 - Article

VL - 75

SP - 428

EP - 457

JO - Educational and Psychological Measurement

JF - Educational and Psychological Measurement

SN - 0013-1644

IS - 3

ER -