Decoding the patterns of critical factor associations driving electric vehicle recommendations

Abstract

This study aims to identify the patterns of key factor combinations that collectively influence consumers to recommend distinct Electric Vehicle (EV) types, namely hydrogen-powered Fuel Cell EVs (FCEVs) and Plug-in EVs (PEVs). To achieve this goal, we apply a non-parametric Association Rules Mining (ARM) approach which unearths latent relationships between different factors in large datasets for analysis. The study utilized the comprehensive 2019 EV user survey data from California, covering socio-economic and demographic characteristics, EV driving and charging habits, housing and parking options, the impact of incentives on EV purchase decisions, and willingness to recommend EVs, among other factors. The results revealed that individuals who find incentives important and experienced a reduced need for refuelling and waiting to refuel their EVs, as well as those who did not have to make special trips to refuel, were more likely to recommend FCEVs. Besides, the availability of home charging facilities and variable electricity rates highly influenced consumers to recommend PEVs. Also, individuals who frequently drove and had positive experiences with both EV types were highly likely to recommend them. Although FCEV users are concerned about station availability when travelling, and PEV users worry about the high cost of EVs, they are still willing to recommend them. The findings highlight critical factors that can enhance EV recommendations and increase adoption rates. Understanding these patterns enables policymakers and industry players to accelerate the transition to sustainable energy and mitigate climate change impacts.