Built environment and travel satisfaction revisited: Differences between consonant and dissonant travelers

Abstract

The literature has extensively evaluated the built environment (BE) and travel satisfaction connections; however, there is a lack of research on their potential nonlinear connections and differences in these connections across various groups. The study applies a double machine learning method to scrutinize the nonlinear associations of travel satisfaction with BE, as well as how these connections vary across consonant and dissonant travelers. The findings indicate that BE variables at both residences and trip destinations play a significant role in predicting travel mode dissonance. Furthermore, the study highlights the differentiated impacts of BE variables at different locations on travel satisfaction, with relative contributions of 49.34% and 71.60% for consonant and dissonant travelers, respectively. Additionally, the analysis reveals nonlinear connections between BE variables and travel satisfaction, with variations in connection patterns between consonant and dissonant travelers. The findings deepen the understanding of travel satisfaction and offer implications for land use planning.