Infrastructure access and routines shape heterogeneous public EV charging preferences

Abstract

Existing research models EV charging choice as tradeoffs among station attributes, obscuring qualitatively distinct behavioral profiles that govern how tradeoffs are formed. Using survey data from 365 BEV owners, we combine factor analysis, latent profile analysis, and inverse propensity weighting (IPW) to identify preference dimensions, segment users, and estimate causal drivers of profile membership. Factor analysis reveals eight latent dimensions spanning station-facing choice attributes and user-facing contextual factors. Latent profile analysis identifies two groups: Efficiency-Oriented Users (76%), prioritizing accessibility and cost, and Information-Responsive Users (24%), emphasizing provider trust, amenities, and situational compatibility. IPW estimates show that private charger access (OR = 2.39), business use (OR = 2.70), and public charging frequency (weekly OR = 7.43) increase Information-Responsive membership, while commuting use decreases it (OR = 0.47). These results demonstrate that preference heterogeneity is shaped by infrastructural access and behavioral routines, offering a preference-based explanation for station underutilization and underscoring the need for differentiated infrastructure planning.