Dataset on Online User-Generated Content, Tourist Engagement, Revisit Intention, and Destination Brand Evangelism in Vietnam (2025–2026)

Description

This dataset contains survey data examining the influence of online user-generated content (UGC) characteristics on tourist engagement, self-congruity, revisit intention, and destination brand evangelism in Vietnam. The study investigates how perceived informativeness, authenticity, vividness, and relevance of UGC affect engagement behavior and subsequent loyalty-related outcomes, while also incorporating travel constraints as a contextual factor. Data were collected from domestic and international tourists at major tourism destinations across Northern, Central, and Southern Vietnam between December 2025 and February 2026. A structured questionnaire using five-point Likert scales was administered on-site with the support of trained research assistants. A total of 1,500 tourists were approached, resulting in 1,347 valid responses after screening. The dataset includes raw survey responses, recoded variables, and processed outputs used for Partial Least Squares Structural Equation Modeling (PLS-SEM). The data can be used for replication studies, comparative tourism research, and further investigation into digital engagement, destination branding, and behavioral intention models in emerging tourism markets

Steps to reproduce

1. Download the raw dataset file from Mendeley Data. 2. Import the dataset into statistical software (e.g., SmartPLS 4, SPSS, or R). 3. Recode items where necessary according to the codebook (all constructs measured using five-point Likert scales). 4. Specify the reflective measurement model including the following constructs: Informativeness (INF), Authenticity (AUT), Vividness (VID), Relevance (REL), UGC Engagement (ENG), Self-Congruity (SC), Travel Constraints (TCN), Revisit Intention (RET), and Destination Brand Evangelism (EVA). 5. Assess the measurement model by evaluating outer loadings (>0.70), Cronbach’s alpha (>0.70), Composite Reliability (>0.70), and Average Variance Extracted (>0.50). 6. Evaluate discriminant validity using the HTMT criterion (<0.85). 7. Assess the structural model using bootstrapping (5,000 resamples) to obtain path coefficients, t-values, and p-values. 8. Examine R², f² effect sizes, and Stone–Geisser Q² values for predictive relevance. 9. Run PLSpredict to compare PLS-SEM prediction errors (RMSE, MAE) with linear model benchmarks.

Institutions

• Foreign Trade University

  Hanoi, Hanoi

Categories

Licence