Partial least squares structural equation modelling of destination brand equity with a higher-order construct: a worked example comparing visitors and non-visitors

This technical note documents a partial least squares structural equation modelling (PLS-SEM) analysis of a consumer-based brand equity (CBBE) model for a tourism destination, and uses it as a worked example of the analytical decisions these models require. Survey data from 430 respondents (215 previous visitors and 215 non-visitors) were used to estimate a model in which previous visitation predicts a higher-order brand equity construct (brand awareness, brand image and perceived quality) and loyalty outcomes (intention to visit and intention to recommend), which in turn relate to overall satisfaction. The higher-order construct was estimated with the repeated-indicators approach in PLS-Graph, and significance was assessed with 500 bootstrap resamples. After iterative item purification, all first-order constructs met conventional thresholds for composite reliability (0.80–0.92), convergent validity (average variance extracted 0.57–0.77) and Fornell–Larcker discriminant validity. Previous visitation was positively associated with brand awareness, brand image, perceived quality and intention to recommend, but not with intention to visit. All three brand equity dimensions were associated with both loyalty outcomes and with satisfaction. The note closes with practical lessons for applied users: checking indicator direction before purification, avoiding over-purification, supplementing Fornell–Larcker with the heterotrait–monotrait ratio, interpreting R² in repeated-indicator higher-order models, and assessing formative higher-order constructs appropriately. The original step-by-step analysis protocol is included as an appendix, with current equivalents for each step. The statistical analysis was conducted in 2006–2009 at the Christ College Academy for Management Education, OHIO University MBA Program, Bangalore, India.