An integrated examination of determinants shaping clothing purchase intentions: insights from PLS-SEM and ANN

Introduction Based on the cognitive–affective–conative (C-A-C) framework, this study adopts a hybrid approach combining PLS-SEM and artificial neural networks (ANN) to explore the factors that shape purchase intention in China’s apparel market. Methods A cross-sectional, self-reported online survey was conducted via Questionnaire Star using a convenience sampling method. Over a 6-week period, a total of 601 valid responses were collected, mainly from young consumers with relatively higher educational attainment. In the first stage, PLS-SEM with 5,000 bootstrap resamples was applied to test the structural paths and mediation effects. In the second stage, the significant predictors identified from the PLS-SEM analysis were input into an ANN model to capture potential nonlinear relationships and rank the importance of variables. Results The PLS-SEM results indicate that innovation, quality, and promotional attributes all exert a significant positive impact on perceived value. Perceived value, in turn, significantly enhances purchase intention and plays a mediating role in the effects of innovation, quality, and promotional attributes on purchase intention. The ANN analysis further confirms these findings, highlighting perceived value as the strongest predictor of purchase intention. It also reveals that quality and promotional attributes are the most influential drivers of perceived value, while innovation attributes can additionally stimulate consumers’ impulse-driven purchase intention. Discussion The combined results suggest that perceived value serves as both the primary pathway through which product and marketing attributes influence purchase intention and the most effective variable for predicting purchase intention. This study makes contributions by translating the C-A-C framework into a testable “attributes–value–intention” chain and integrating PLS-SEM and ANN to uncover both linear transmission mechanisms and nonlinear predictive contributions. The integrated evidence enhances the explanatory and predictive power of the research and provides practical guidance for Chinese apparel firms to optimize quality cues and value creation strategies.