Organic food purchasing: dissecting with data science

This research develops an integrated Machine Learning (ML)-Structural Equation Modelling (SEM) approach to unravel the complex dynamics of organic food purchasing decisions. The study aims to bridge a gap in consumer behaviour understanding by juxtaposing the insights derived from both methods. Employing the Random Forest algorithm (handling highdimensional, unstructured data) on survey data from 1003 Sydney residents, key factors impacting organic food purchasing decisions, such as taste, trust in certification, environmental consciousness, and health benefits, were discovered. Following the ML phase, a SEM approach was used to confirm and understand the interplay of these factors. SEM results emphasized 'perceived trust' as a significant determinant of organic food consumption intention, illustrating SEM's capability in interpreting ML results and capturing the interaction between factors. This methodological interplay enables the generation of insights with high predictive power and provides comprehensive understanding, valuable for marketing strategies and policymaking in the organic food industry.