The Selection Criteria and Indexes for Identifying Factor Numbers and Structures in Multilevel Exploratory Factor Analysis

• In Event: 69.046 - Latent Growth/Transition Modeling and Model Selection

Abstract

In single-level EFA, the commonly model fit indexes, such as ICs, RMSEA, CFI, TLI, SRMR and parallel analysis, were used to extract the latent factors underlying observed items. No research or simulation works have been done to show whether these indexes also perform well in multilevel EFA (MEFA). The purpose of this study is to examine the accuracy of the above indexes in identifying the correct number of factors in MEFA. After data generation, we used both the design-based and the model-based approaches for the analysis. The preliminary results just showed the traditional IC indexes can recover the correct number of factors except for AIC. The performance of RMSEA and SRMR was poor, and CFI and TLI was very good.

Authors

• Melissa Y. Ji, Texas A&M University - College Station

• Wen Luo, Texas A&M University - College Station

• Oiman Kwok, Texas A&M University - College Station

• Yanyun Yang, Florida State University