Performance of Information Complexity Criteria in Structural Equation Models with Applications

Authors

  • Eylem Deniz Howe Mimar Sinan Fine Arts University
  • Hamparsum Bozdogan Department of Statistics, Operations, and Management Science, The University of Tennessee, Knoxville, TN, 37996, USA
  • Gulay Kiroglu Mimar Sinan Fine Arts University

Keywords:

Structural Equation Model, Information Criteria, AIC-type Criteria, ICOMP-type Criteria

Abstract

A common problem in structural equation modeling is that of model selection. Many researchers have addressed this problem, but many methods have provided mixed benefits until recently. Akaike’s well-known criteria, AIC, has been applied in the context of structural equation modeling, but the effectiveness of many other information criteria have not been studied in a convincing manner. In this paper, we compare the SEM model selection prowess of several AIC-type and ICOMP-type criteria. We also introduce two new large sample consistent forms of Bozdogan’s ICOMP criteria - one of which is robust to model misspecification. To study the empirical performance of the information criteria, we use a well-known SEM simulationprotocol, and demonstrate that most of the information-theoretic criteria select the “pseudo trueâ€Â model with very high frequencies. We also demonstrate, however, that the performance of AIC is inversely related to the sample size. Finally, we apply the new criteria to select an analytical model for a real dataset from a retail marketing study of consumer behavior. Our results show the versatilityof the new proposed method where both the goodness-of fit and the complexity of the model is taken into account in one criterion function.

Downloads

Issue

Section

Econometrics and Statistics

How to Cite

Performance of Information Complexity Criteria in Structural Equation Models with Applications. (2012). European Journal of Pure and Applied Mathematics, 5(3), 282-301. https://www.ejpam.com/index.php/ejpam/article/view/1675