*4.2. Participants' Profile*

The study's participants belonged to five different cities in Pakistan. In terms of gender, 218 of the 354 participants were male, comprising 61.6% of the total sample. Married individuals were 203, accounting for 57.3 percent. In terms of qualifications, 146 individuals had bachelor's degrees, with a percentage of 41.2. One hundred and forty-six participants earned roughly 172 US dollars per month, making a 41.2% representation. In Pakistan, the average salary of a person is around 229 US dollars (PKR 40,000) per month. Table 1 shows the participants' information.


**Table 1.** Participants' Profile.

#### *4.3. Reliability and Convergent Validity*

The internal consistency of the data is characterized as "reliability" [84]. Internal consistency was examined in this study using Cronbach's alpha (α) and composite reliability scores. The Cronbach's alpha (α) values greater than 0.60 suggest the internal consistency of the data. Cronbach's alpha (α) values between 0.70 and 0.80 are deemed credible. While the values between 0.80 and 0.90 are considered to be substantially reliable. The values of all constructions were more than 0.80, indicating internal consistency. Another way for establishing internal consistency is the composite reliability (CR) rating. Internal consistency is better measured by CR [85]. All constructions had CR values larger than 0.90, demonstrating internal consistency. The degree of resemblance of measurement constructs when assessed using diverse measuring methods is referred to as convergent validity [84]. We used three measures to determine convergent validity: composite reliability (CR), outer loadings, and average variance extracted (AVE). Table 2 shows that the values of CR ranged from 0.845 to 0.934, outer factor loadings values ranged between 0.696 to 0.933, and AVE values ranged from 0.646 to 0.826 confirming the presence of convergent validity [85,86]. Figure 2 shows the measurement model that depicts the strength of relationships (path coefficients) among the constructs.

#### *4.4. Descriptive Statistics and Discriminant Validity*

Descriptive statistics test was performed to assess the values of mean, median, mode and standard deviation as shown in Table 3. The degree to which measurement constructs differ from one another is referred to as discriminant validity [85]. The establishment of discriminant validity is required for the appropriateness of the statistical results [87]. The discriminant validity of this study was determined using two measures: the Fornell and Larcker criteria and the Heterotrait-Monotrait (HTMT) ratio. According to Fornell and Larcker's [86] criteria, a construct should have more variances with its components than other constructs [85]. As demonstrated in Table 3, the diagonal values, i.e., the square roots of AVE, were bigger than the inter-correlation among the constructs, indicating discriminant validity [88]. Second, the discriminant validity was determined using the HTMT ratio criteria. HTMT values less than 0.90 are required for the establishment of discriminant validity [87]. The HTMT value of all constructs shown in Table 4 was less than 0.90, confirming the discriminant validity.

### *4.5. Predictive Power of the Inner Model*

In the study, model fit criteria were assessed through the values of coefficient of determination (R2) and cross-validated redundancy (Q2) [85,87]. The variance explained by the independent constructs on dependent constructs is represented by the value of (R2). The value of (R2) for endogenous constructs was 10.5%, 17.3%, and 18.9% for word of

mouth, intention to use, and willingness to pay more respectively, which moderated the predictive accuracy of the studied model. Next, we assessed the value of cross-validated redundancy (Q2) through the blindfolding method. In this method, a (Q2) value above zero indicates that the model has an adequate predictive relevance. The values of (Q2) for word of mouth, intention to use, and willingness to pay were 6.6%, 10.9%, and 11.8% respectively, which showed that the model possessed an adequate predictive relevance.


**Table 2.** Reliability Testing and Convergent Validity.

Note: CR = Composite Reliability; AVE = Average Variance Extracted.
