Understanding Pro-Environmental Behavior in Sustainable Mobility: An Integrated Framework for Electric Vehicle (EV) Purchase Intentions

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript studies Taiwanese consumers’ intention to purchase electric vehicles (EVs) by integrating the Norm Activation Model (NAM) and Theory of Planned Behavior (TPB). Using survey data (n=421), the authors apply Structural Equation Modeling (SEM) and Artificial Neural Networks (ANN) to test relationships and rank predictor importance. The findings suggest that attitude, perceived behavioral control, and personal norms strongly influence EV purchase intention, while subjective norms are less direct. The study aims to contribute by offering a hybrid theoretical and methodological framework.

Evaluation of methodology, analyses, and conclusions

• The use of convenience sampling limits representativeness. The sample is skewed toward highly educated respondents, and generalization beyond Taiwan is uncertain.
• Reliability and validity are reported, but details on survey items, loadings, and discriminant validity checks are incomplete. Some statistics appear inconsistent and need clarification.
• Most hypothesized paths are significant, but causal language should be avoided since the design is cross-sectional. Model comparisons use Δχ² tests, which are not valid for non-nested models; AIC/BIC or predictive checks are preferable.
• No statistical checks are reported, despite reliance on self-report survey data.
• While innovative, reporting is vague. Key details on model setup, validation, and performance are missing. The claim of predictive improvement is not convincingly demonstrated.
• The section largely repeats results. Contributions beyond confirming known predictors (attitude, norms, PBC) need to be clarified. Policy and managerial implications remain generic.

Key recommendations for improvement

• Clearly explain how your integration of NAM and TPB differs from prior work and why the SEM–ANN approach adds value.
• Provide full item details, reliability/validity checks, and correct inconsistencies in the tables.
• Use appropriate model comparison metrics (AIC/BIC), report bootstrap details for indirect effects, and address common method bias.
• Document preprocessing, validation, and baseline comparisons to show why ANN improves prediction.
• Go beyond restating results: explain surprising findings (e.g., weak subjective norm effect), relate them to cultural/policy context, and offer more specific implications.
• Correct grammar, unify abbreviations, and streamline the conclusion to focus on contributions, limits, and future research.

Author Response

Please check the uploaded Word file.

Thank you for your suggestions.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The study aims to identify the key factors influencing Taiwanese consumers' intention to purchase electric vehicles. To this end, the author combines the Theory of Planned Behaviour (TPB) with the Norm Activation Model (NAM), creating a comprehensive framework that takes into account both rational decision-making processes and moral obligations that shape environmentally oriented behaviour.

The data was collected through a survey in which 421 Taiwanese citizens participated. The analysis was conducted in two stages: first, structural equation modelling (SEM) was used to test hypotheses and identify significant relationships between variables; then, artificial neural networks (ANN) were used to determine the relative importance of predictors and identify potential nonlinear interactions.

Combining two theoretical models (TPB and NAM) provided a more complete picture than when they were applied separately. The combined use of SEM and ANN strengthened the reliability of the results and allowed us to identify both linear and nonlinear relationships. The results have practical value: they show that information campaigns aimed at increasing environmental responsibility and forming a positive attitude among consumers are necessary to promote electric vehicles.

However, the results of the study should be interpreted with caution due to certain methodological limitations. The sample was limited to five major cities in Taiwan, which limits the possibility of generalising the results to other regions. In addition, the use of convenience sampling may have reduced the representativeness of the data set. However, despite these shortcomings, the study has practical value at the local level.

To improve the quality of the article, the author should:

- supplement paragraph 2 (Review of Literature and Development of Hypotheses) with references to additional sources of information from the last 2 years.

- in paragraph 6 (Implications), focus on the practical value of the study rather than its theoretical value.

Author Response

Please check the uploaded Word file.

Thank you for your suggestions.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper proposed an integrated framework for EV purchase intentions and the use of neural network models for predictive modelling. Strengths of the paper includes addressing the limitations of existing frameworks, provides insights into purchase intentions and ways to encourage the adoption of EV from the findings.

Some areas of concerns/improvements are:

• For Table 3: Other parameters shown in the table should also be discussed and recheck RMSEA 90% CI values
• The R² values used in line 407 should be shown and/or discussed before being used to support the proposed model’s fit is better. Moreover, the little difference between 0.58 and 0.56 may not be significantly meaningful. Possibly, provide other ways of showing that the proposed model is better.
• The small sample size used for training the ANN models limits their use.
• A table of comparison can be provided comparing the authors’ results and from other researchers to show the similarities and/or differences in their findings.

Some other minor errors:

• Line 297 mentioned about NAM but the table showed NAT
• RMSEA 90% CI mentioned in table 3 for proposed model (0.70) does not match with the table 5 description (0.066)

Author Response

Please check the uploaded Word file.

Thank you for your suggestions.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

1.                      The study focuses exclusively on Taiwanese consumers and the results may not be generalizable to other countries with different cultural, economic or regulatory contexts regarding electric vehicles. The authors should discuss these as limitations or explore in the article a few words about how to replicate this study.

2.                      The research combines two distinct theoretical models (NAM + TPB). Integration can generate conceptual overlap (e.g., between personal norms and attitudes), making it difficult to clearly interpret individual effects. The choice of constructs and how they complement each other without redundancy should be better justified.

3.                      Uses SEM for causal analysis and ANN for prediction. Although innovative, the use of ANN can be seen as a "black box" — difficult to interpret theoretically. In addition, there may be inconsistency between the results of the two methods. The authors should further explain how the ANN results complement the SEM findings and discuss interpretive limitations.

Author Response

Please check the uploaded Word file.

Thank you for your suggestions.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have revised manuscript very well and now the manuscript can be accepted for publication.

Author Response

Thank you for your recommendation.