Modeling mobile payment adoption toward a cashless society using integrated TAM TRA and TPB with SEM and ANN

The study investigates mobile payment system (MPS) adoption in Bangladesh by identifying the factors that shape users’ attitudes and behavioral intentions. The Technology Acceptance Model (TAM), Theory of Reasoned Action (TRA), and Theory of Planned Behavior (TPB) are integrated to evaluate the user behavior. A quantitative survey of 411 MPS users, conducted via purposive sampling, was administered online. Data analysis followed a two-stage strategy including Partial Least Squares Structural Equation Modeling (PLS-SEM) with artificial neural networks (ANNs). PLS-SEM was used for theory-driven linear modeling and hypothesis testing in SmartPLS 4.0, supported by descriptive and preliminary analyses in SPSS 27.0, while ANNs were applied to explore potential non-linear patterns and compare predictive accuracy. Main results show that attitude (AT), performance expectancy, and perceived ease of use significantly drive behavioral intention (BI) to use MPS. However, ANN analysis reveals interesting findings in the nonlinear model evaluation, where AT is the dominant predictor, and PBC is the second most significant predictor of BI. This study advances theory by integrating TAM, TRA, and TPB in the context of MPS user behavior in an emerging economy. The findings highlight the importance of user-centered interface design, usability, and manageable personalized services for fostering continuous MPS adoption.