A Hybrid Three-staged SF-AHP, PLS-SEM and ANN Model to Predict Vaccination Intention against COVID-19 Pandemic
Description
Vaccine program plays a vital role in building herd immunity in the worldwide population during the spread of the COVID-19 pandemic. This study aimed to identify the key factors affecting vaccination intention against COVID-19 using a new three-staged approach combining both Spherical Fuzzy Analytic Hierarchy Process (SF-AHP), Structural Equation Model (SEM), and Artificial Neural Network (ANN) to determine the relative weight and importance of the factors as well as to test the proposed hypotheses in the research model. Using online survey questionnaires, data were collected from 474 respondents throughout Vietnam. The findings of SF-AHP demonstrated that Perceived COVID-19 vaccines (PVC), Trust in government intervention strategies (TRS), and Social media (SOM) were the most important factors predicting vaccination intention against COVID-19 pandemic, based on fifteen experts’ points of view. The results of the SEM showed that individuals’ vaccination intention was significantly and directly influenced by Perceived COVID-19 vaccines (PVC), Trust in government intervention strategies (TRS). Although Perceived severity of COVID-19 (PSC) did not perform a direct influence on vaccination intention, its indirect influence through Perceived COVID-19 vaccines (PVC), and Social media (SOM) had no direct effect on the intention to receive COVID-19 vaccines, however, indirectly affecting COVID-19 vaccination intention moderating through trust in government strategy. However, Social Influence (SOI) was found no have a significant direct effect on the intention to take a vaccine against COVID-19. Finally, the ANNs’ findings were consistent with the SF-AHP and SEM models, also revealed that the best predictors of COVID-19 vaccination intention were Perceived COVID-19 vaccines (PVC), Trust in government intervention strategies (TRS), and Social media (SOM). The ANN model can predict vaccination intention with an accuracy of 90 %. This research proposed an innovative new approach employing quantitative and qualitative techniques to understand COVID-19 vaccination intention during the global pandemic. Furthermore, the proposed method can be applied and extended to evaluate the perceived effectiveness of COVID-19 measures in other countries currently dealing with the COVID-19 outbreak.
Files
Steps to reproduce
The proposed research framework consists of 3 phases. In Phase 1, assigning fuzzy weights to criteria based on pairwise comparisons is done using the SF-AHP model. In Phase 2, the PLS-SEM approach is used to validate the hypotheses as indirect/direct effects. In Phase 3, significant predictors from PLS-SEM analysis were taken as the ANN model’s input neurons. According to the normalized importance obtained from the multilayer perceptrons of the feed-forward-back-propagation ANN algorithm, we can find significant effects of vaccination intention and determine the accurate prediction rates.