Effects of Cross-loadings on Determining the Number of Factors to Retain. Issue 6 (1st November 2020)

• Record Type: Journal Article
• Authors: Li, Yujun; Wen, Zhonglin; Hau, Kit-Tai; Yuan, Ke-Hai; Peng, Yafeng
• Is Part Of: Structural equation modeling. Volume 27:Issue 6(2020)
• Page Start: 841
• Page End: 863
• Publication Date: 2020-11-01
• Subjects: Exploratory factor analysis -- cross-loading -- number of factors -- non-normality; Multivariate analysis -- Periodicals; Social sciences -- Statistical methods -- Periodicals; 519.535
• DOI: 10.1080/10705511.2020.1745075
• Languages: English
• ISSNs: 1070-5511

Abstract

ABSTRACT: In exploratory factor analysis (EFA), cross-loadings frequently occur in empirical research, but its effects on determining the number of factors to retain are seldom known. In this paper, we analyzed whether and how cross-loadings affected the performance of the parallel analysis (PA), the empirical Kaiser criterion (EKC), the likelihood ratio test (LRT), the comparative fit index (CFI), the Tucker-Lewis index (TLI), and the root mean square error of approximation (RMSEA) in determining the number of factors to retain. A large-scale simulation study was also conducted. A few conclusions can be drawn: (1) overall, PA provides the most accurate performance, especially when data are non-normally distributed; (2) cross-loadings noticeably affect the performance of PA, CFI, and TLI with different patterns, and they virtually have no effect on EKC, LRT, and RMSEA; (3) no method is immune to the sizable detrimental effect of normality assumption violation. Several recommendations were provided.