Incorporating Uncertainty Into Parallel Analysis for Choosing the Number of Factors via Bayesian Methods

• In Event: Methods for Multilevel, Categorical, and Multidimensional Data

Abstract

Parallel analysis (PA) tends to yield more accurate results in determining the number of factors, in comparison with other statistical methods. Nevertheless, PA can suggest an incorrect number of factors, particularly in statistically unfavorable conditions. Thus, when the number of factors is chosen based on PA, uncertainty exists. We propose a Bayesian parallel analysis (B-PA) method to incorporate the uncertainty with decisions about the estimated number of factors. B-PA yields a probability distribution for the various possible numbers of factors. Results of a simulation study show that B-PA provides relevant information regarding the uncertainty in determining the number of factors, even if the indicated number of factors with the highest probability is incorrect.

Authors

• Roy Levy, Arizona State University

• Yan Xia

• Samuel B. Green, Arizona State University