Predicting pro-environmental behavioral intention using interpretable machine learning

Abstract

Abstract This study employs machine learning techniques to investigate the determinants of pro-environmental behavioral intention (PEBI) using the survey data from the 2021 wave of the Chinese General Social Survey (CGSS). By comparing the predictive performance of nine prevalent machine learning algorithms, the results show that the Random Forest model achieves the highest predictive accuracy in predicting PEBI. Furthermore, we develop an interpretable model based on the Random Forest and Shapley additive explanations (SHAP) analysis to examine the relative importance of 20 key predictors. The findings reveal that environmental concern is the most influential factor and is positively related to individual PEBI, followed by environmental attitude, connection to nature, environmental knowledge, government performance perception, and environmental responsibility. Additionally, we also examine the heterogeneity across different subsamples and explore interaction effects between key variables. Overall, this study adds novel knowledge to environmental sustainability literature by employing machine learning methods to predict the influencing factors of PEBI.