ABSTRACT
This chapter reports the results of a study examining the performance of two heuristic algorithms commonly used for conducting specification searches in structural equation modeling (SEM): the ant colony optimization (ACO) and the Tabu search algorithms. A secondary goal of the study was to determine which goodness of fit criteria perform best with the ACO and Tabu algorithms in specification searches. A number of fit indices were examined: the chi-square statistic, the Bayesian Information Criterion (BIC), the Comparative Fit Index (CFI), the Tucker–Lewis Index (TLI), and the Root Mean Square Error of Approximation (RMSEA). These indices were selected because of their widespread popularity in evaluating model goodness of fit within SEM. Using data with known structure, the algorithms and fit criteria were examined under different model misspecification conditions. In all conditions, the Tabu search procedure outperformed the ACO algorithm. Recommendations are provided for researchers to consider when undertaking specification searches.
