Determinants of Green Smartphone Application Adoption for Sustainable Food Consumption Among University Students. A Combination of SEM and Machine Learning Methods

Abstract

Recently, many smartphone applications have been developed offering green food items. However, the number of active users on the applications is very limited. This study aims to investigate the determinants of green smartphone application adoption among users. The study employs content richness model and Unified Theory of Acceptance and Use of Technology (UTAUT) as well as extrinsic constructs such as customisation and environmental concerns. A quantitative approach using a survey is utilised by collecting 700 responses. The data is analysed using Structural Equation Modelling (SEM) and three machine learning techniques including Artificial Neural Networks (ANN), Classification Regression Tree (CRT) and Chi-Squared Automatic Interaction Detection (CHAID). The results indicate that UTAUT, customisation and environmental concerns positively impact the adoption of green applications. Further analysis revealed fitness of analytical methods and the importance of variables for the overall sample and the subsamples derived. The study provides theoretical and practical contributions to academics, marketers and software developers in understanding consumer behaviour in the field. The result assist developers and marketers to decipher consumer behaviour towards green applications for sustainable consumption. The research contributes to theory and practice by employing an integrative model to investigate the role of technology in sustainable consumption. Moreover, the findings revealed the fitness of three machine learning methods to analyse the data collected for green consumption and the importance of variables in the model. The data is collected by employing convenience sampling. Hence, the results cannot be generalised accurately. Furthermore, data collection is conducted using a cross-sectional approach. Future researchers can add to the findings using a probability sampling and/or longitudinal data collection to generalise the results and reveal the changes in consumer behaviour.