Extending a tourism causality network model

Abstract

This study provides insights into a powerful statistical technique to test model invariance across multiple groups. In doing so, it provides insights for policymakers and contributes to the literature on tourism demand forecasting by validating and extending previous results on how supply-side factors influence tourism demand. Specifically, this study considered an a priori validated structural model for relationships among the economy, society, environment and tourism constructs in 162 countries, examining the extent to which the model was invariant across two groups of isolated observations: developed and less-developed countries. Upon testing, the a priori model did not replicate across both groups. Instead a reduced model, incorporating the society, environment and tourism constructs only was used to compare and test for variances in parameters across the two groups using multigroup analysis sampling in AMOS 16.0. The results indicated that path coefficients were equal across groups. A positive relationship existed between the society and environment constructs, while a similar positive significant relationship existed between the society and environment constructs from one side and tourism, respectively.