A SEM–Neural Network Approach to Predict Customers’ Intention to Purchase Battery Electric Vehicles in China’s Zhejiang Province

Yueling Xu, Wenyu Zhang, Haijun Bao, Shuai Zhang, Ying XiangSustainability 2019 · DOI

TL;DR

This study combines Structural Equation Modeling (SEM) and Neural Networks (NN) to predict consumer purchase intention for Battery Electric Vehicles (BEVs) in China's Zhejiang Province. It identifies key drivers of BEV adoption, highlighting the critical role of perceived behavioral control and offering actionable insights for policymakers and businesses to boost BEV sales and improve air quality.

What Problem Does This Paper Address?

Motor vehicle pollution is a significant global concern, and China, a major vehicle producer and consumer, is actively promoting Battery Electric Vehicles (BEVs) as a solution. However, understanding what truly drives consumers to purchase BEVs is crucial for effective policy and marketing. Existing research often focuses on industry-level policies or uses traditional models that may oversimplify complex consumer behavior. This paper aims to bridge this gap by investigating the factors influencing BEV purchase intention from a consumer perspective, specifically within Zhejiang Province, a region with proactive environmental policies.

Methods

The researchers employed a two-stage analytical approach:

  1. Structural Equation Modeling (SEM): This statistical technique was used to test hypotheses derived from the Theory of Planned Behavior (TPB) and its extensions. SEM helped identify which independent variables (attitude, perceived behavioral control, subject norm, environmental performance, price value, non-monetary incentive policy, and monetary incentive policy) had a significant positive relationship with BEV purchase intention.
  2. Neural Network (NN) Analysis: Following the SEM analysis, a neural network model was used to rank the normalized importance of the significant predictors identified by SEM. This approach leverages the NN's ability to capture non-linear relationships and overcome the "black box" limitation of traditional SEM by providing a more nuanced understanding of predictor influence. The combination of SEM and NN aims to mitigate the oversimplification of SEM and the overfitting issues of NN.

Datasets

The study utilized data from a survey of 382 residents in Zhejiang Province, China, conducted in December 2018 and January 2019. The survey questionnaire was translated from English to Chinese. The data collected included demographic information and responses to Likert-scale questions measuring various constructs related to BEV purchase intention. No new benchmark datasets were introduced.

Key Insights and Ideas

  • Perceived Behavioral Control is King: The most striking finding is that a consumer's belief in their ability to purchase and own a BEV (perceived behavioral control) is the strongest predictor of their intention to buy one. This suggests that addressing practical barriers and making BEVs more accessible is paramount.
  • Beyond Just "Green": While environmental performance is important, it's not the sole driver. Consumers are also influenced by their personal attitudes towards BEVs and the perceived social norms around them.
  • Incentives Matter, But How? Monetary incentive policies (like subsidies and tax breaks) showed a significant positive impact, reinforcing the idea that financial considerations are key. However, non-monetary incentives (like preferential parking or lane access) surprisingly had no significant influence in this study, suggesting a need to re-evaluate their effectiveness or communication.
  • Price is Not Always Right (or Wrong): Price value, as a standalone factor, did not significantly influence purchase intention in this specific context, which might be counterintuitive but could be explained by the broader policy environment and other dominant factors.
  • The Power of the Hybrid Approach: Combining SEM for hypothesis testing and NN for importance ranking offers a robust way to analyze complex consumer behavior, capturing both linear and non-linear relationships and providing a more complete picture than either method alone.

Key Results

  • Significant Predictors: Attitude, perceived behavioral control, subject norm, environmental performance, and monetary incentive policy measures all had a significant and positive effect on BEV purchase intention.
  • Non-Significant Predictors: Price value and non-monetary incentive policy measures did not show a significant influence on purchase intention in this study.
  • Predictor Importance Ranking (from NN analysis):
    • Perceived behavioral control: 100%
    • Subject norm: 84.69%
    • Environmental performance: 77.80%
    • Attitude: 76.10%
    • Monetary incentive policy measures: 50%
  • Model Accuracy: The NN model demonstrated low RMSE values (average training RMSE of 0.018, testing RMSE of 0.022), indicating accurate and reliable predictions.

Limitations and Future Work

The authors acknowledge several limitations:

  • Limited Scope of Factors: The study focused on a specific set of predictors. Future research could explore additional factors that influence BEV purchase intention, potentially drawing from other consumer behavior theories.
  • Regional Specificity: The study was conducted solely in Zhejiang Province. The findings may not be generalizable to other regions in China or other countries due to differences in geographical locations, market conditions, and policy implementation. Future studies should expand to different provinces or cities to derive more robust and universally applicable insights.
  • Policy Implementation: The effectiveness of non-monetary incentives was not significant, suggesting a need for future research to investigate how these policies are designed, communicated, and implemented to maximize their impact.

The authors suggest that future work should expand the independent variables to create a more comprehensive model and conduct empirical analyses across different regions to inform policy design more effectively.

Summary updated

Abstract

As part of the increasing efforts toward the prevention and control of motor vehicle pollution, the Chinese government has practiced a range of policies to stimulate the purchase and use of battery electric vehicles (BEVs). Zhejiang Province, a key province in China, has proactively implemented and monitored an environmental protection plan. This study aims to contribute toward streamlining marketing and planning activities to introduce strategic policies that stimulate the purchase and use of BEVs. This study considers the nature of human behavior by extending the theory of planned behavior model to identify its predictors, as well as its non-linear relationship with customers’ purchase intention. To better understand the predictors, a substantial literature review was given to validate the hypotheses. A quantitative study using 382 surveys completed by customers in Zhejiang Province was conducted by integrating a structural equation model (SEM) and a neural network (NN). The initial analysis results from the SEM revealed five factors that have impacted the customers’ purchase intention of BEVs. In the second phase, the normalized importance among those five significant predictors was ranked using the NN. The findings have provided theoretical implications to scholars and academics, and managerial implications to enterprises, and are also helpful for decision makers to implement appropriate policies to promote the purchase intention of BEVs, thereby improving the air quality.
BibTeX
@inproceedings{xu2019semneural,
  author = {Yueling Xu and Wenyu Zhang and Haijun Bao and Shuai Zhang and Ying Xiang},
  title = {{A SEM–Neural Network Approach to Predict Customers’ Intention to Purchase Battery Electric Vehicles in China’s Zhejiang Province}},
  booktitle = {Sustainability},
  year = {2019},
  doi = {10.3390/su11113164},
}
Figure