Determining the Number of Factors in Exploratory Factor Analysis: A Comparison of Several Methods

• In Event: Advanced Topics in Multiple Regression Analysis

Abstract

Perhaps the most important step in conducting exploratory factor analysis is deciding the number of factors to extract. Research shows that extracting an incorrect number of factors can have detrimental consequences in interpreting a rotated factor solution. Many methods have been proposed to estimate the correct number of factors. The purpose of this study was to expand upon previous research and examine the performance of several prominent methods under a wide variety of factor structures. Ten of the most commonly discussed methods for determining the number of factors were compared using a Monte Carlo approach to simulate a wide variety of factor models with extensive replication. In general, parallel analysis on the reduced correlation matrix performed well.

Authors

• Robert Pearson, University of Northern Colorado

• Daniel J. Mundfrom, Eastern Kentucky University

• Adam Piccone, Datalogix