STDC.1
Description
1) The main hypothesis indicates that tourism destination competitiveness exerts a mediating effect between community social capital and social prosperity perception, at indigenous Mexican communities. The sample was made up of 11 indigenous communities with touristic activities, associated with the Indigenous Tourism Network of Mexico (RITA). To collect data we use the non-probabilistic snowball technique and the number of observation units was 113. 2) To obtain data, a survey was developed, theoretical concepts were operationalized and Bourgaduos social distance scale was used. Each latent variable was made up of its underlying dimensions and in turn, these were formed with their respective items. To build the factors of each variable, the factorial analysis was used with the varimax extraction method (SPSS). Reliability was measured by Cronbach's alpha coefficient. The relationship between constructs was analysed with the Partial Least Squares Structural Equation Modeling (PLS-SEM) technique. According to pls theory, we used a reflective model and to run the model, the PLS algorithm was used. The lost values were identified as -0.99 and the algorithm that replaced them is called Case Wise Replacement. The PLS algorithm was used for Weighting Scheme were Path Weighting Scheme; Data Metric, mean 0, Var 1; maximum iterations were 500 as this is consistent with the recommendations made by Hair et al., (2019). 3) The interpretation of the results for the study context indicates that community social capital does not have a direct relationship with the perception of social prosperity, however, the indirect relationship through competitiveness indicates a mediating effect.
Files
Steps to reproduce
The data are the perception answers of tourism destination residents. We identified missing values as -99 We use factorial analysis in SPSS ver. 22 We used all data to construct each variable and the analysis was single for each variable. After that, we export data to smartpls ver. 2, we use a reflective model and run the pls algorithm. After the first analysis, we take out the latent variables that did not obtain the recommended score. It was re-estimated and we obtain the final model.