Comparing Accuracy of Parallel Analysis and Fit Statistics for Estimating the Number of Factors with Ordered Categorical Items in Exploratory Factor Analysis
Abstract
Determining the number of factors in factor analysis is crucial in that it affects the rest of the analysis procedure and can eventually mislead theory development when the factor retention decision is inaccurate. Researchers have recommended various methods for deciding the number of factors to retain in exploratory factor analysis (EFA), but this remains one of the most difficult decisions in the EFA. This study aims to compare the parallel analysis which is considered the most reliable method to the performance of fit indices that researchers have started used as another strategy for determining the optimal number of factors in EFA. The Monte Carlo simulation study was conducted with ordered categorical items because there is a lack of simulation studies that are conducted with ordered items despite their popularity in social sciences. The results indicate that the parallel analysis and the RMSEA performed well in most conditions, followed by TLI and then by CFI. Implications, limitations of this study, and future research directions are discussed.
Subject Area
Quantitative psychology
Recommended Citation
Lee, Hyunjung, "Comparing Accuracy of Parallel Analysis and Fit Statistics for Estimating the Number of Factors with Ordered Categorical Items in Exploratory Factor Analysis" (2022). ETD Collection for Fordham University. AAI29320495.
https://research.library.fordham.edu/dissertations/AAI29320495
