Table 1 shows the demographic details of 399 respondents from Klang Valley, Malaysia. The respondents were 197 scientists and 202 public, where 51.1% were female, and 48.9% were male. More than 70% of them were less than 40 years old. Approximately 42.4% of respondents were Malays, which reflected the actual population ratio in the Klang Valley, where most of them are Malays . Table 2 shows the overall mean scores for religiosity (6.07), intention to dengue vaccine (5.71), trust in key players (5.51), attitude to dengue vaccine (5.71), and perceived benefit (5.38) were rated high. The stakeholders responded that they were entirely dedicated to their religion, trusted the key players, viewed the dengue vaccine as incredibly beneficial, and had a positive attitude and intention to accept it. Nevertheless, the stakeholders were rated moderate for attitude to technology (Mean = 4.74, above the mid-point of 4.0) and perceived risk (Mean = 3.58, below the mid-point of 4.0). The findings imply that the stakeholders were more attracted to technology and believed that the dengue vaccination had limited risk.
MEASUREMENT MODEL ANALYSIS
The analysis of the convergent reliability and validity of the variables is shown in Table 3. Convergent validity can be determined if the factor loadings are larger than 0.7 [48, 49], the composite reliability (CR) is more than 0.70 , and the average variance extracted (AVE) is larger than 0.50 [51, 52]. The findings indicated that the factor loadings of the items were higher than 0.7, except for several items (PBV1 = 0.693; PBV5 = 0.692; ATT1 = 0.698). Nonetheless, according to Byrne et al. (2010), if the total AVE exceeded 0.50, the factor loadings below 0.70 were retained . Therefore, all the variables had AVE values exceeding 0.50, and the values of CR were greater than 0.70, which is considered acceptable.
The discriminant validity analysis also found that the variables have met the requirements (Refer to Table 4). In the Fornell-Larcker criterion assessment, each variable has a more excellent square root value of AVE than the other variables . The value of the Heterotrait-monotrait (HTMT0.90)correlation for each of the variables was acceptable because the values were less than 0.85 [54, 55]. The measurement model analysis was also measured by standardised root mean square residual (SRMR) and normed fit index (NFI) as suggested by Lohmoller (1989) . Good fit model must have SRMR value below than 0.08  and the NFI value closer to 0.9 [58, 59]. In this study, the SRMR value was 0.074, and the NFI value was 0.71, considered an acceptable fit (Refer to Table 5). The variance inflation factor (VIF) values for all the variables were lower than 5.0, suggesting no collinearity concerns the inner model .
STRUCTURAL MODEL ANALYSIS
The structural model analysis started with the coefficient of determination (R2) testing. The R2 value for the intention (0.564) suggested that exogenous variables in the model could explain 56.4% of the variance in intention to dengue vaccine. The R2 value of the attitude was 0.371, suggesting that the exogenous variables explain 37.1% of the factor. Furthermore, the exogenous variables explained 19.6% of the variance in perceived benefit and 18.9% of the variance in perceived risk. The analysis continued with the blindfolding procedure to measure the predictive accuracy of the model predictions (Q2), where the value must be beyond zero .
The Q2 values for the perceived benefit (0.109), perceived risk (0.111), attitude (0.198), and intention to dengue vaccine (0.383) confirmed that the predictive relevance of the model was adequate for the exogenous variables. According to Cohen (1988) , attitude (f2 = 0.465) has a large effect size on intention to dengue vaccine compared with perceived benefit (f2 = 0.141). Perceived benefit has a medium effect size on attitude (f2 = 0.184), while the effect size of religiosity (f2 = 0.067), attitude to technology (f2 = 0.012), and trust in key players (f2 = 0.011) was small. The findings also showed that trust in key players (f2 = 0.112), attitude to technology (f2 = 0.051), and religiosity (f2 = 0.026) have a small effect size on perceived benefit. Lastly, attitude to technology (f2 = 0.113) and trust in key players (f2 = 0.082) have a small effect on perceived risk. Table 6 illustrates the results of R2, Q2, and f2 values.
DIRECT RELATIONSHIPS ANALYSIS
The relationship between exogenous and endogenous variables was evaluated by examining the path coefficients’ size in the structural model. Attitude (β = 0.544, t = 11.322, p < 0.001) was the most important direct predictor of intention to dengue vaccine, followed by perceived benefit (β = 0.299, t = 6.377, p < 0.001) (Refer to Table 7 and Figure 2). The findings indicated that when the respondents were inclined to have a good attitude to the dengue vaccine and viewed that it has higher benefits, they would have a positive intention to accept it. Attitude is an important factor in influencing intention whether they express likes or dislikes and support or reject anything . Arham et al. (2021b) showed that attitude was the most important factor in expressing support for the use of Wolbachia techniques to control dengue . Besides, perceived benefit also plays a role in determining intention. Mustapa et al. (2019) explained that the acceptance of new technology, especially in the field of health, disclosure of important benefits in determining intention .
Perceived benefit (β = 0.459, t = 10.415, p < 0.001) was the most significant direct predictor of attitude to dengue vaccine followed by religiosity (β = 0.211, t = 4.996, p < 0.001), attitude to technology (β = 0.095, t = 2.076, p = 0.019), and trust in key players (β = 0.095, t = 1.872, p = 0.031) (Refer to Table 7 and Figure 2). The results suggested that when stakeholders perceived higher benefits, clung to their religion, acknowledged that the benefits of technology outweigh risks on nature, and had a high level of trust in the key players involved in the dengue vaccine, they expressed a good attitude and accepted it. These findings indicate differences between the study of Arham et al. (2021b) and Amin and Hashim (2015) [12, 18]. Arham et al. (2021b) pointed out that perceived benefit and risk influenced acceptance towards Wolbachia techniques . In contrast, Amin and Hashim (2015) showed that perceived benefit and trust in key players were the factors influencing stakeholders’ attitudes towards genetically modified mosquito techniques in an effort to control dengue .
Nevertheless, the stakeholders will manifest a positive attitude towards dengue control techniques when they feel the benefit. According to Amin et al. (2018), the Malaysian community has firm religious beliefs, and the acceptance of the new technologies depends on their spiritual level . Conclusively, the stakeholders in this study have firm religious beliefs and do not feel that the dengue vaccine extends beyond religion. Trust in key players, such as implementers and researchers, will balance good relationships among stakeholders . This notion is clearly shown in this study, where stakeholders trust key players and accept new technologies beyond the values of nature. Dengue vaccine possibly does not pose any danger to environmental health if the authorities carry out their duties properly.
Trust in key players (β = 0.310, t = 6.554, p < 0.001), attitude to technology (β = 0.207, t = 4.319, p < 0.001), and religiosity (β = 0.147, t = 3.195, p = 0.001) have a positive association with perceived benefit (Refer to Table 7 and Figure 2). This finding suggests that when stakeholders trust people who play significant roles in the dengue vaccine, are deeply attached to their religion and are more inclined to technology (negative), they benefit from the dengue vaccine. Nevertheless, attitude to technology (β = -0.317, t = 5.896, p < 0.001) and trust in key players (β = -0.280, t = 6.157, p < 0.001) had a negative association with perceived risk (Refer to Table 7 and Figure 2). Although they have a tendency towards technology compared to nature values, they put higher trust in key players as they feel less risk on the dengue vaccine.
The study’s findings clearly show a bipolar relationship between predictor factors with perceptions of benefit and risk, as Alhakimi and Slovic (1994) described . Mustapa et al. (2021) discovered that stakeholders' acceptance of new technology is significantly influenced by high perceived benefits and low perceived risks . Therefore, the finding is further elucidated by previous studies, who showed an inverse relationship between general predictor factors such as belief in priorities, attitudes towards nature, and religion with perceptions of benefit and risk in determining the acceptance of Wolbachia and Outdoor Residual Spraying techniques [17, 18]. In conclusion, general predictor factors positively influence stakeholders’ benefits if they feel the benefits outweigh the risks. According to scholars, perceived benefit and risk are difficult to conceptualise separately because of their complex relationships that have inverse relationships [31, 32, 33].
Mediation analysis assesses the indirect effect of the path coefficients that have a positive sign between exogenous and endogenous variables via the mediator. As shown in Figure 2, attitude and perceived benefit were expected to act as a mediator. According to Zhao et al. (2010), the indirect effect can be calculated by dividing the direct effect on the exogenous and endogenous variables by the total mediating effect to measure the significance of the variance accounted for (VAF) value . If the VAF value exceeds 20%, the mediating effect is present.
Table 7 and Figure 2 shows that attitude mediates the positive relationship between perceived benefit and intention to dengue vaccine (β = 0.299, t = 6.377, p < 0.001) with the VAF value of 45.51%. Thus, the results confirmed that attitude act as a mediator, indicating that when the stakeholders perceive the higher benefits of the dengue vaccine, they expressed a positive attitude, which translated to a positive intention to accept dengue vaccine.
Hence, the findings also concluded that perceived benefit act as the mediating role in the relationship between trust in key players and attitude (β = 0.095, t = 1.872, p = 0.031) with the VAF value of 59.97%, attitude to technology and attitude (β = 0.095, t = 2.076, p = 0.019) with the VAF value of 50.0%, and religiosity and attitude (β = 0.211, t = 4.996, p < 0.001) with the VAF value of 24.24%. These findings demonstrate that when the respondents have higher trust in key players, rated positively on technology, and clung to their religion, they tend to assess the benefit of dengue vaccination as being elevated. They endorse it with a positive attitude.