Sequential Bayesian SEM for Task Technology Fit

The Task Technology Fit (TTF) model is a key framework in information systems research that examines the relationship between user task needs and technological capabilities. Structural Equation Modeling (SEM) and Bayesian Structural Equation Modeling (BSEM) are effective tools for analyzing the TTF model. SEM reveals complex relationships between observed and latent variables, while BSEM is particularly useful for dynamic analyses, incorporating prior information and updating the model in sequential steps. This study compared the performance of SEM, BSEM, and sequential Bayesian SEM in analyzing the TTF model, using Normal and Beta prior distributions. The Bayesian Information Criterion (BIC) assessed model fit, and the Root Mean Square Error (RMSE) evaluated coefficient accuracy. The results indicate that sequential BSEM effectively analyzes models like TTF in sequential conditions. The Beta distribution, known for its stability, is more suitable for sequential Bayesian models. This study introduces a new analytical framework to aid future research in information systems and sequential Bayesian analysis.