*4.2. Structural Equation Model*

We first used a descriptive analysis of the demographic characteristics of the sample. Second, we evaluated the indicators related to model quality. Finally, the proposed hypotheses were tested.

The covariance-based structural equation model (CB-SEM) and variance-based partial least squares structural equation modeling (VB-SEM) can both be used to analyze structural equation models. However, the following may be noted. (1) Partial least squares structural equation modeling (PLS-SEM) is more suitable than CB-SEM for measuring structural equation models with more than six latent variables [53]. (2) PLS-SEM is suitable for a wider range of data characteristics than CB-SEM, especially for handling non-normally distributed data [53]. (3) PLS-SEM is more suitable for small-sample measurements and exploratory studies [53].

This is an exploratory study with six latent variables in the research model and a small, effective sample size. Additionally, a multivariate normality analysis was performed on the data collected in this study using a web calculator to measure the distribution of the data (https://webpower.psychstat.org/, accessed on 13 July 2022). The results show Mardia's multivariate skewness (β = 40.707, *p* < 0.001) and multivariate kurtosis (β = 473.530, *p* < 0.01) that suggest multivariate non-normality. In summary, PLS-SEM is more suitable for data analysis in this study [54,55].
