Optimization of Model Specification Search in Structural Equation Modeling With Mixed Ant Colony Optimization Algorithm

• In Event: 18.018 - Advances in Structural Equation Modeling

Abstract

Model misspecification is a crucial effect of model fit and parameter estimation in structural equation modeling (SEM). Heuristic algorithm is a popular approach to handle model misspecification by searching for optimum model from all possible subset models. This study provides a new searching algorithm, mixed ACO, which is modified based on the original Ant Colony Optimization (ACO) algorithm and Tabu search feature. A Monte Carlo study shows that the two model extraction strategies of mixed ACO can provide both population model and optimum fit model across all simulation conditions. This method can be used for applied research to detect misspecification with prior knowledge.

Authors

• Zeyuan Jing, University of Florida

• Walter L. Leite, University of Florida

• Huan Kuang, University of Florida