Student Perceptions of the Use of Gen-AI in a Higher Education Program in Spain

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript digital-3672139 investigates teacher-training students' perceptions of Gen-AI in a higher education program in Spain, using a mixed-methods approach (N=407). The findings highlight perceived benefits, including enhanced engagement, motivation, and personalized learning, alongside challenges such as dependency, ethical concerns, and privacy issues. While the work provides valuable insights into student perspectives, its novelty is limited. The methodology and conclusions closely mirror existing studies like Student Perceptions of ChatGPT Use in a College Essay Assignment, which similarly explores student attitudes toward AI tools, ethical dilemmas, and trust dynamics. Both studies emphasize mixed-method designs, Likert-scale surveys, and thematic analysis of open-ended responses. However, this paper lacks unique theoretical or practical contributions, such as novel frameworks for AI integration or longitudinal assessments of learning outcomes. The quantitative results (e.g., post-test acceptance scores ≈4/5) align with prior literature but fail to demonstrate significant advancements. Visualizations (e.g., word clouds) are poorly rendered and offer little interpretative depth. Overall, the study reinforces established themes in Gen-AI research without addressing critical gaps, such as causal relationships between AI use and academic performance or scalable strategies for mitigating dependency risks.

 

Review Comments

1. The English needs to be revised throughout. The authors should pay attention to the spelling and grammar throughout this work. I would only respectfully recommend that the authors perform this revision or seek the help of someone who can aid the authors.
2. (Section 1 Introduction) The reviewer hopes the introduction section in this paper can introduce more studies in recent years. The reviewer suggests authors don't list a lot of related tasks directly. It is better to select some representative and related literature or models to introduce with certain logic. For example, the latter model is an improvement on one aspect of the former model.
3. The paper’s focus on student perceptions of Gen-AI mirrors Student Perceptions of ChatGPT Use in a College Essay Assignment, which also examines ethical concerns, trust, and engagement. Both studies rely on pre-post surveys and thematic analysis, but the current work lacks differentiation. For instance, while the comparison study uniquely explores ChatGPT’s role in grading and evolving trust dynamics, this paper’s findings (e.g., dependency risks, privacy concerns) are well-documented in prior literature. The authors should clarify how their work advances beyond these precedents—e.g., by proposing actionable frameworks for AI literacy or evaluating pedagogical interventions—to justify its novelty.
4. The quantitative results (Tables 2–5) report mean scores but omit critical statistical parameters (e.g., p-values, effect sizes, confidence intervals). For example, the claim that "post-test acceptance improved" (e.g., "I enjoy Gen-AI lessons" rising from M=3.21 to 3.91) lacks significance testing, making it unclear whether changes are meaningful or due to chance. Additionally, the study does not control for confounding variables (e.g., prior AI experience, digital literacy levels). Incorporating ANOVA or regression analyses would strengthen causal inferences and align with rigorous educational research standards.
5. (Section I, Introduction) The reviewer suggest to revise the original statement as … As far as the application of Gen-AI is concerned, students' beliefs and attitudes towards its integration and use can influence their degree of involvement and the learning outcomes they are expected to achieve, …; ldcnet: limb direction cues-aware network for flexible human pose estimation in industrial behavioral biometrics systems; ehpe: skeleton cues-based gaussian coordinate encoding for efficient human pose estimation; mmatrans: muscle movement aware representation learning for facial expression recognition via transformers.
6. Figures 2–5 (word clouds and category networks) are low-resolution and lack interpretative annotations. For instance, Figure 2’s "Benefits of Gen-AI" lists generic terms like "speed" and "creativity" without contextualizing their frequency or thematic relationships. Comparatively, the cited ChatGPT study uses clear heatmaps to map trust evolution. The authors should enhance visualizations with hierarchical clustering or co-occurrence networks to reveal deeper insights, such as how "dependency" correlates with "critical thinking" challenges.
7. The study’s cross-sectional design limits its impact. Unlike the ChatGPT paper, which tracks perceptions before and after a specific assignment, this work does not evaluate how sustained Gen-AI exposure influences long-term attitudes or learning outcomes. Including longitudinal data (e.g., tracking students across semesters) or comparing results with non-teacher-training cohorts (e.g., STEM students) would strengthen generalizability and address gaps in AI adoption literature.
8. The original sentence is suggested to revise as …Research on the application of Gen AI in higher education systems has revealed explicit benefits...; from engagement to performance: the role of effort regulation in higher education online learning; understanding learner continuance intention: a comparison of live video learning, pre-recorded video learning and hybrid video learning in covid-19 pandemic; exploring the relationship between teacher talk supports and student engagement from the perspective of students’ perceived care.
9. While the paper suggests adapting curricula to "learners’ needs," it offers no concrete strategies. For example, how might educators balance personalized learning with dependency risks? The ChatGPT study proposes "calibrated trust" models involving instructor oversight—a actionable takeaway absent here. The authors should propose evidence-based interventions, such as AI literacy modules or hybrid assessment frameworks, to translate findings into practice. Additionally, expanding the scope to vocational training or adult education could demonstrate scalability beyond teacher-training contexts.
10. The study acknowledges privacy concerns but does not detail how data security was ensured (e.g., anonymization protocols, encryption). Furthermore, the convenience sampling (N=407 from a single Spanish region) introduces selection bias, limiting representativeness. Replicating the study in diverse geographic or socio-economic settings (e.g., rural vs. urban institutions) would enhance external validity.
11. The paper lacks a robust theoretical framework (e.g., Technology Acceptance Model, Self-Determination Theory) to contextualize For instance, the observed "motivation" gains could be analyzed through the lens of intrinsic vs. extrinsic motivators, but such connections are absent. Grounding the study in established theories would strengthen its academic contribution and provide a foundation for future research.

 

My overall impression of this manuscript is that it is in general well-organized. The work seems interesting and the technical contributions are solid. I would like to check the revised manuscript again.

Comments on the Quality of English Language

The English needs to be revised throughout. The authors should pay attention to the spelling and grammar throughout this work. I would only respectfully recommend that the authors perform this revision or seek the help of someone who can aid the authors.

Author Response

Comments 1: This manuscript digital-3672139 investigates teacher-training students' perceptions of Gen-AI in a higher education program in Spain, using a mixed-methods approach (N=407). The findings highlight perceived benefits, including enhanced engagement, motivation, and personalized learning, alongside challenges such as dependency, ethical concerns, and privacy issues. While the work provides valuable insights into student perspectives, its novelty is limited. The methodology and conclusions closely mirror existing studies like Student Perceptions of ChatGPT Use in a College Essay Assignment, which similarly explores student attitudes toward AI tools, ethical dilemmas, and trust dynamics. Both studies emphasize mixed-method designs, Likert-scale surveys, and thematic analysis of open-ended responses. However, this paper lacks unique theoretical or practical contributions, such as novel frameworks for AI integration or longitudinal assessments of learning outcomes. The quantitative results (e.g., post-test acceptance scores ≈4/5) align with prior literature but fail to demonstrate significant advancements. Visualizations (e.g., word clouds) are poorly rendered and offer little interpretative depth. Overall, the study reinforces established themes in Gen-AI research without addressing critical gaps, such as causal relationships between AI use and academic performance or scalable strategies for mitigating dependency risks.

Response 1: Thank you for pointing this out. We agree with this comment and with the importance of acknowledging and writing effectively about limitations in research. We have therefore included all your comments in the limitations section, as you have described them in your review.

Comments 2: The English needs to be revised throughout. The authors should pay attention to the spelling and grammar throughout this work. I would only respectfully recommend that the authors perform this revision or seek the help of someone who can aid the authors.

Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have proofread the textt, paying attention to the spelling and grammar throughout the work.

Comments 3: (Section 1 Introduction) The reviewer hopes the introduction section in this paper can introduce more studies in recent years. The reviewer suggests authors don't list a lot of related tasks directly. It is better to select some representative and related literature or models to introduce with certain logic. For example, the latter model is an improvement on one aspect of the former model.

Response 3: Thank you for pointing this out. We agree with this comment. Therefore, we have presented some models that allow for better integration of AI into higher education, as requested.

Comments 4: The paper’s focus on student perceptions of Gen-AI mirrors Student Perceptions of ChatGPT Use in a College Essay Assignment, which also examines ethical concerns, trust, and engagement. Both studies rely on pre-post surveys and thematic analysis, but the current work lacks differentiation. For instance, while the comparison study uniquely explores ChatGPT’s role in grading and evolving trust dynamics, this paper’s findings (e.g., dependency risks, privacy concerns) are well-documented in prior literature. The authors should clarify how their work advances beyond these precedents—e.g., by proposing actionable frameworks for AI literacy or evaluating pedagogical interventions—to justify its novelty.

Response 4: Thank you for pointing this out. We agree with this comment. Therefore, we've included that information in the conclusions section.

Comments 5: The quantitative results (Tables 2–5) report mean scores but omit critical statistical parameters (e.g., p-values, effect sizes, confidence intervals). For example, the claim that "post-test acceptance improved" (e.g., "I enjoy Gen-AI lessons" rising from M=3.21 to 3.91) lacks significance testing, making it unclear whether changes are meaningful or due to chance. Additionally, the study does not control for confounding variables (e.g., prior AI experience, digital literacy levels). Incorporating ANOVA or regression analyses would strengthen causal inferences and align with rigorous educational research standards.

Response 5: Thank you for pointing this out. We agree with this comment. Therefore, we have applied non-parametric tests (Wilcoxon signed-rank test and Kruskal Wallis tests) to make clear that some changes are significant and not due to chance, as requested.

Comments 6: (Section I, Introduction) The reviewer suggest to revise the original statement as … As far as the application of Gen-AI is concerned, students' beliefs and attitudes towards its integration and use can influence their degree of involvement and the learning outcomes they are expected to achieve, …; ldcnet: limb direction cues-aware network for flexible human pose estimation in industrial behavioral biometrics systems; ehpe: skeleton cues-based gaussian coordinate encoding for efficient human pose estimation; mmatrans: muscle movement aware representation learning for facial expression recognition via transformers.

 

Response 6: Thank you for pointing this out. We agree with this comment. We have therefore revised the statement as suggested by the reviewer.

Comments 7: Figures 2–5 (word clouds and category networks) are low-resolution and lack interpretative annotations. For instance, Figure 2’s "Benefits of Gen-AI" lists generic terms like "speed" and "creativity" without contextualizing their frequency or thematic relationships. Comparatively, the cited ChatGPT study uses clear heatmaps to map trust evolution. The authors should enhance visualizations with hierarchical clustering or co-occurrence networks to reveal deeper insights, such as how "dependency" correlates with "critical thinking" challenges.

Response 7: Thank you for pointing this out. We agree with this comment. Therefore, we have modifed and enhanced visualizations as requested.

Comments 8: The study’s cross-sectional design limits its impact. Unlike the ChatGPT paper, which tracks perceptions before and after a specific assignment, this work does not evaluate how sustained Gen-AI exposure influences long-term attitudes or learning outcomes. Including longitudinal data (e.g., tracking students across semesters) or comparing results with non-teacher-training cohorts (e.g., STEM students) would strengthen generalizability and address gaps in AI adoption literature.

Response 8: Thank you for pointing this out. We agree with this comment and with the importance of acknowledging and writing effectively about limitations in research. We have therefore included all your comments in the limitations section, as you have described them in your review.

Comments 9: The original sentence is suggested to revise as …Research on the application of Gen AI in higher education systems has revealed explicit benefits...; from engagement to performance: the role of effort regulation in higher education online learning; understanding learner continuance intention: a comparison of live video learning, pre-recorded video learning and hybrid video learning in covid-19 pandemic; exploring the relationship between teacher talk supports and student engagement from the perspective of students’ perceived care.

Response 9: Thank you for pointing this out. We agree with this comment. Therefore, we have taken your comment into consideration.

Comments 10: While the paper suggests adapting curricula to "learners’ needs," it offers no concrete strategies. For example, how might educators balance personalized learning with dependency risks? The ChatGPT study proposes "calibrated trust" models involving instructor oversight—a actionable takeaway absent here. The authors should propose evidence-based interventions, such as AI literacy modules or hybrid assessment frameworks, to translate findings into practice. Additionally, expanding the scope to vocational training or adult education could demonstrate scalability beyond teacher-training contexts.

Response 10: Thank you for pointing this out. We agree with this comment and with the importance of including concrete strategies as the one suggested in your review to adapt curricula to learners’ needs. We also agree with the importance of acknowledging and writing effectively about limitations in research. That is why, we have included all your comments in the limitations section.

 

Comments 11: The study acknowledges privacy concerns but does not detail how data security was ensured (e.g., anonymization protocols, encryption). Furthermore, the convenience sampling (N=407 from a single Spanish region) introduces selection bias, limiting representativeness. Replicating the study in diverse geographic or socio-economic settings (e.g., rural vs. urban institutions) would enhance external validity.

Response 11: Thank you for pointing this out. More information has been provided on anonymization protocols, as requested. In addition, we agree with the importance of acknowledging and writing effectively about limitations in research. That is why, we have included all your comments in the limitations section.

Comments 12: The paper lacks a robust theoretical framework (e.g., Technology Acceptance Model, Self-Determination Theory) to contextualize For instance, the observed "motivation" gains could be analyzed through the lens of intrinsic vs. extrinsic motivators, but such connections are absent. Grounding the study in established theories would strengthen its academic contribution and provide a foundation for future research.

 Response 12: Thank you for pointing this out. We agree with this comment. Therefore, we have provided a more robust theoretical framework, as requested.

Comments 13: My overall impression of this manuscript is that it is in general well-organized. The work seems interesting, and the technical contributions are solid. I would like to check the revised manuscript again.

Response 13: Thank you for your comment. Therefore, we have included the suggestions you have commented in the manuscript to the extent possible.

Comments 14: Comments on the Quality of English Language. The English needs to be revised throughout. The authors should pay attention to the spelling and grammar throughout this work. I would only respectfully recommend that the authors perform this revision or seek the help of someone who can aid the authors.

Response 14: Thank you for pointing this out. We agree with this comment. Therefore, we have proofread the text, as requested.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This study surveyed 407 pre-service teachers majoring in early childhood and elementary education. However, the overall conclusion drawn is “students' perceptions of the use of Gen-AI in higher education programs,” which may suggest a mismatch between the surveyed population and the intended research objectives. Why were early childhood education majors specifically chosen? It may be more appropriate to narrow the scope of the conclusion to focus on the development of early childhood education professionals.

The title mentions a “case study,” a term typically used to refer to qualitative research methods involving in-depth analysis of individual cases. However, this paper employs quantitative analysis with a large sample. Therefore, it is recommended to revise the title accordingly and adopt a structured abstract format, including sections such as Purpose, Methodology, Analytical Approach, Results, and Contributions.

The paper lacks a literature review section.

The full content of the questionnaire should be provided, at least in the appendix.

The survey includes four dimensions, yet only one Cronbach’s alpha is reported. Rather than conducting a reliability analysis on the entire item set, reliability should be assessed separately for each dimension.

Please include both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Common method bias should be tested using the Unmeasured Latent Method Variable (ULMV) approach.

Please provide results for both convergent and discriminant validity.

In Section 2.4, why are students the ones responsible for developing the learning plan?

What is the basis for Figure 1?

What was the effective response rate of the survey? How were invalid responses excluded?

Why was only descriptive statistical analysis conducted? The use of basic statistical methods limits the theoretical contribution of the paper. Consider applying more advanced methods such as MANOVA or fsQCA.

What is the rationale behind the paths depicted in Figures 3 and 5?

Separate the Discussion and Conclusion into two distinct sections. The Discussion should critically analyze the findings of the study and highlight discrepancies with prior research. The Conclusion should briefly summarize the overall findings, theoretical contributions, practical implications, limitations, and directions for future research.

Author Response

Comments 1: This study surveyed 407 pre-service teachers majoring in early childhood and elementary education. However, the overall conclusion drawn is “students' perceptions of the use of Gen-AI in higher education programs,” which may suggest a mismatch between the surveyed population and the intended research objectives. Why were early childhood education majors specifically chosen? It may be more appropriate to narrow the scope of the conclusion to focus on the development of early childhood education professionals.

Response: Thank you for pointing this out. We agree with this comment. Indeed, the participants in our study were all pre-service teachers specializing in early childhood and elementary education. This focus was intentional, as our objective was to explore how future educators in these foundational stages of learning perceive and engage with Gen-AI tools within the context of their own higher education experience.

However, we recognize that the phrasing of our conclusions may have inadvertently suggested broader generalizability beyond our target population. In response to this valuable feedback, we have revised both the abstract and conclusion sections to explicitly clarify that our findings pertain to pre-service teachers in early childhood and elementary education programs. This adjustment ensures alignment between our sample population and the stated implications of the study.

We have also added a sentence to the “Participants” section to explain why this particular group was selected: due to their unique position as both learners in higher education and future educators responsible for shaping early learning environments, their perceptions of Gen-AI carry particular relevance for understanding how such technologies might influence teaching practices and pedagogical beliefs from the ground up.

 

Comments 2: The title mentions a “case study,” a term typically used to refer to qualitative research methods involving in-depth analysis of individual cases. However, this paper employs quantitative analysis with a large sample. Therefore, it is recommended to revise the title accordingly and adopt a structured abstract format, including sections such as Purpose, Methodology, Analytical Approach, Results, and Contributions.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have changed the title removing “a case study”.

Comments 3: The paper lacks a literature review section.

Response: Thank you for pointing this out. However, we respectfully disagree with the assertion that the manuscript lacks a literature review. A review of the relevant literature is indeed included within the introduction section, as is common in studies of this nature, particularly when the literature is directly integrated to support the rationale and context for the study.

Nonetheless, in response to the reviewer’s concern, we have further expanded and clarified this section to ensure that the theoretical background and relevant prior studies are more explicitly presented and easier to identify as fulfilling the function of a literature review. We believe this revision strengthens the framing of the study while maintaining a cohesive structure.

Comments 4: The full content of the questionnaire should be provided, at least in the appendix.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have provided the questionnaire in the results section.

Comments 5: The survey includes four dimensions, yet only one Cronbach’s alpha is reported. Rather than conducting a reliability analysis on the entire item set, reliability should be assessed separately for each dimension.

Response: Thank you for pointing this out. We acknowledge that the ideal approach would be to report separate reliability coefficients for each dimension of the instrument. However, in the current stage of the study, we conducted and reported the reliability analysis for the primary dimension most central to our research objectives.

Conducting separate reliability analyses for all four dimensions would require reprocessing the dataset and partially redesigning the analysis plan, which would go beyond the current scope and timeline of this work. That said, we have clarified this decision in the manuscript and indicated the limitation explicitly in the “Limitations and Future Research” section. We also acknowledge that a more detailed reliability breakdown could enhance the instrument’s validation and will consider this in future studies or follow-up publications.

We appreciate the reviewer’s suggestion, which helps us reflect critically on the methodological scope of the study.

Comments 6: Please include both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA). Common method bias should be tested using the Unmeasured Latent Method Variable (ULMV) approach.

Response: Thank you for pointing this out. We agree that the use of both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), along with testing for common method bias using the Unmeasured Latent Method Variable (ULMV) approach, would certainly strengthen the psychometric validation of the instrument.

However, we respectfully note that the primary aim of this study was not the development or full validation of a new instrument, but rather the exploration of perceptions using an existing or adapted instrument in the context of pre-service teacher education. As such, only a basic reliability check (Cronbach’s alpha) and descriptive analyses were conducted to ensure the internal consistency and relevance of the data for answering the research questions.

We recognize the value of a more comprehensive validation process and have noted this as a limitation in the revised manuscript. We also state that future studies will aim to include both EFA and CFA, as well as formal checks for method bias, to more robustly validate the measurement model.

Comments 7: Please provide results for both convergent and discriminant validity.

Response: Thank you for pointing this out. We fully agree that assessing convergent and discriminant validity is an important step in establishing the psychometric soundness of a measurement instrument.

However, as noted in our response to the previous comment, the focus of this study was not on the development or full-scale validation of a new instrument, but rather on exploring pre-service teachers’ perceptions using a targeted set of items related to generative AI in higher education. Given the exploratory nature of the study and the structure of the instrument, we did not conduct a full construct validation process, including tests of convergent and discriminant validity.

We acknowledge this as a limitation and have now included a statement in the manuscript’s “Limitations and Future Research” section, indicating the need for further validation of the instrument—including tests for convergent and discriminant validity—in future work.

Comments 8: In Section 2.4, why are students the ones responsible for developing the learning plan?

Response: Thank you for pointing this out. In Section 2.4, students are described as responsible for developing the learning plan to emphasize the learner-centered approach adopted in the educational program. This approach encourages active engagement, autonomy, and critical thinking by having pre-service teachers design their own learning pathways, which aligns with contemporary pedagogical theories promoting self-regulated learning.

Moreover, empowering students to create their own learning plans reflects the goal of preparing future educators to take ownership of their professional development and to apply similar learner-centered strategies in their future classrooms.

Comments 9: What is the basis for Figure 1?

Response: Thank you for pointing this out. The basis of Figure 1 is just a description of the resources used in the intervention program.

Comments 10: What was the effective response rate of the survey? How were invalid responses excluded?

Response: Thank you for pointing this out. While we recognize the importance of transparency regarding response rates and data cleaning procedures, we did not consider it necessary to include these specific details in the manuscript, as the focus of the study is on the analysis of the valid responses obtained. Nevertheless, all responses underwent standard quality checks to ensure data reliability.

If the editor or reviewers deem it essential, we are willing to provide additional information or include a brief explanation regarding response rates and exclusion criteria in a revised version.

Comments 11: Why was only descriptive statistical analysis conducted? The use of basic statistical methods limits the theoretical contribution of the paper. Consider applying more advanced methods such as MANOVA or fsQCA.

Response: The decision to focus on descriptive statistical analysis was guided by the exploratory nature of our study, which aimed primarily to map and understand pre-service teachers’ perceptions regarding the use of generative AI in their educational programs.

While we acknowledge that more advanced statistical methods such as MANOVA or fsQCA could provide deeper insights and stronger theoretical contributions, applying these techniques would require a different research design and additional data considerations beyond the scope of the current study.

We have noted this as a limitation in the manuscript and suggested that future research could build upon our findings by employing such advanced analytical methods to further explore the relationships and complexities inherent in this topic.

Comments 12: What is the rationale behind the paths depicted in Figures 3 and 5?

Response: Thank you for pointing this out. We have changed figures 3 and 5.

Comments 13: Separate the Discussion and Conclusion into two distinct sections. The Discussion should critically analyze the findings of the study and highlight discrepancies with prior research. The Conclusion should briefly summarize the overall findings, theoretical contributions, practical implications, limitations, and directions for future research.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have separated both sections.

Reviewer 3 Report

Comments and Suggestions for Authors

add Generative AI, primary education and early childhood education to key words

The intro seems a bit positively skewed towards AI use.

Spell out Gen-AI the first time you use it. and check for consistent spelling (e.g., l. 119)

l. 53-4: cite your research

Clarify what you mean about a Gen-AI-based training program (l. 146): do you mean that it is developed by AI or do you mean that the subject matter is Gen-AI?

Fun way to record open-ended questions (word cloud)

State how the program was developed. Was it pre-existing one? Did you develop it or locate it (if the latter what was your basis for selection)?

Good details about the program.

Likert scale isn't scalar so you shouldn't do means and SD, just median/mode/range

It is typical that one would learn something after taking a course....   There are non-parametric tests to tell if the change was significant.

Nice quotes.

I like the qual figures.

I am surprised that bias was only mentioned in small letters in the word cloud.

I'm not seeing any part of the training that had students examine AI limitations/negative aspects (or even about prompts, how AI works). It seems these aspects should be part of the program as participants themselves noted issues - the training can also then address how to deal with these issues. An analogy is learning how to drive a car; you tell learners about car dangers or bad driving habits and then how to drive well and avoid accidents; you don't tell them to avoid cars.

 

Author Response

Comments 1: add Generative AI, primary education and early childhood education to key words

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have added the suggested proposals to key words.

Comments 2: The intro seems a bit positively skewed towards AI use.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have modified the introduction and we have described previous literatura that demonstrate significant advancements in this field.

Comments 3: Spell out Gen-AI the first time you use it. and check for consistent spelling (e.g., l. 119)

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have

Comments 4: l. 53-4: cite your research.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have cited our research.

Comments 5: Clarify what you mean about a Gen-AI-based training program (l. 146): do you mean that it is developed by AI or do you mean that the subject matter is Gen-AI?

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have specified that the use of Gen-AI resources is one of its characteristics, which is why it is developed with these tools, with the teaching of social sciences being the main matter.

Comments 6: Fun way to record open-ended questions (word cloud).

Response: Thank you for pointing this out. Thank you very much for liking it.

Comments 7: State how the program was developed. Was it pre-existing one? Did you develop it or locate it (if the latter what was your basis for selection)?

Response: Thank you for pointing this out. We agree with this comment. We have therefore provided information on the pre-existing programme and how we adapted it for the purposes of social science teaching.

Comments 8: Good details about the program.

Response: Thank you for pointing this out. Thank you very much for liking it.

Comments 9: Likert scale isn't scalar so you shouldn't do means and SD, just median/mode/range.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have done median, mode and range. M/Mo/R.

Comments 10: It is typical that one would learn something after taking a course....   There are non-parametric tests to tell if the change was significant.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have applied non-parametric tests and we have described the significant changes.

Comments 11: Nice quotes.

Response: Thank you for pointing this out. We agree with this comment.

Comments 12: I like the qual figures.

Response: Thank you for pointing this out. We agree with this comment.

Comments 13: I am surprised that bias was only mentioned in small letters in the word cloud.

Response: Thank you for pointing this out. We agree with this comment. Participants have focused their attention to other issues such as dependency, overreliance or loss of creativity. That is also remarkable in this study.

Comments 14: I'm not seeing any part of the training that had students examine AI limitations/negative aspects (or even about prompts, how AI works). It seems these aspects should be part of the program as participants themselves noted issues - the training can also then address how to deal with these issues. An analogy is learning how to drive a car; you tell learners about car dangers or bad driving habits and then how to drive well and avoid accidents; you don't tell them to avoid cars.

Response: Thank you for pointing this out. We agree with this comment. Therefore, we have described in the programme what is highlighted in comment # 16, that is, part of the training was to make students think over potential risks of AI abuse and how to deal with them. To illustrate this, careful instructions were provided to student teachers not only to carry out suggested tasks but how to deal with potential challenges, i.e., how prompts can be better defined.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

It is not meaningful to conduct reliability analysis on all items as a whole. I believe this statistical strategy may significantly undermine the rigor of the paper. I strongly recommend conducting the analysis by dividing the items into different dimensions, as this is a fundamental statistical principle.

The survey process is also critically important and should be described in detail in the methodology section. This includes: how the survey was conducted, how respondents’ attention was assessed, how invalid samples were excluded, the specific proportion of exclusions, the methods used to encourage participation, and how common method bias was avoided, among others.

Lastly, in my view, even when using established scales, it is still necessary to analyze convergent and discriminant validity when applying them in different contexts. Of course, this issue is not as serious as the two mentioned above. However, if the authors do not face substantive constraints, I would still recommend conducting these validity analyses.

Author Response

Comment 1: It is not meaningful to conduct reliability analysis on all items as a whole. I believe this statistical strategy may significantly undermine the rigor of the paper. I strongly recommend conducting the analysis by dividing the items into different dimensions, as this is a fundamental statistical principle.

Response 1: Thank you for pointing this out. We agree with this comment. Therefore, we have applied several measures of sampling adequacy in the factor analysis (KMO, Bartlett) and we have also applied EFA and CFA. In addition, we have also divided the items into different dimensions as requested.

Comment 2: The survey process is also critically important and should be described in detail in the methodology section. This includes: how the survey was conducted, how respondents’ attention was assessed, how invalid samples were excluded, the specific proportion of exclusions, the methods used to encourage participation, and how common method bias was

Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have provided information on how the survey was conducted, how respondents' attention was assessed, how invalid samples were excluded, the specific proportion of exclusions, the methods used to encourage participation, and how common method bias was minimized as requested.

Comment 3: Lastly, in my view, even when using established scales, it is still necessary to analyze convergent and discriminant validity when applying them in different contexts. Of course, this issue is not as serious as the two mentioned above. However, if the authors do not face substantive constraints, I would still recommend conducting these validity analyses.

Response 3: Thank you for pointing this out. We agree with this comment, we have provided information on internal validity, as requested.

Reviewer 3 Report

Comments and Suggestions for Authors

Keep you crossed-out keywords.

l. 48: (PBL) model.

Usually the intro is shorter and provides the query within the first couple paragraphs, which focuses the reader. Most of your intro should be in a lit review section.

Why did you add PBL in the article? It isn't in the abstract, and it doesn't seem to be the intent of your investigation; your adding it in the article ids distracting. Add a RQ on PBL, although your data gathering didn't include PBL. So do omit the PBL.

RQ2: avoid yes/no Q: what are possible differences in participants' perceptions in using GAI?

Proofread for tense use (e.g., l. 188P; in most cases, it should be past tense. And proofread for use of capitals in headings.

Table 2: define M Mo and R. State the range of values (1-5? and what they mean).  Survey seems positive skewed. 

l. 362 (p< .001). Italicize p

Table 4: left-justify statements.

You didn't ask about digital literacy, so your conclusion shouldn't push on digital literacy.

Hyped/over-generalization: The exploration of engaging web-based classroom resources integrated into various 734 formats and applications is paramount in today's education. (the PARAMOUNT part cannot be asserted with just one study, particularly in light of your instruments and method.

 

 

 

 

 

Author Response

Comment 1: Keep you crossed-out keywords.

Response 1: Thank you for pointing this out. We agree with this comment. Therefore, we have erased the crossed-out keywords, as requested.

Comment 2: l. 48: (PBL) model.

Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have followed your comment, as requested.

Comment 3: Usually the intro is shorter and provides the query within the first couple paragraphs, which focuses the reader. Most of your intro should be in a lit review section.

Response 3: Thank you for pointing this out. We agree with this comment. Therefore, we have provided the query within the first couple paragraphs and we have included a literature review section, as requested.

Comment 4: Why did you add PBL in the article? It isn't in the abstract, and it doesn't seem to be the intent of your investigation; your adding it in the article ids distracting. Add a RQ on PBL, although your data gathering didn't include PBL. So do omit the PBL.

Response 4: Thank you for pointing this out. We agree with this comment. Therefore, we have omitted the PBL, as requested.

Comment 5: RQ2: avoid yes/no Q: what are possible differences in participants' perceptions in using GAI?

Response 5: Thank you for pointing this out. We agree with this comment. Therefore, we have changed the RQ2, as suggested.

Comment 6: Proofread for tense use (e.g., l. 188P; in most cases, it should be past tense. And proofread for use of capitals in headings.

Response 6: Thank you for pointing this out. We agree with this comment. Therefore, we have revised the use of verb tenses and capitalization in titles, as requested.

Comment 7: Table 2: define M Mo and R. State the range of values (1-5? and what they mean).  Survey seems positive skewed. 

Response 7: Thank you for pointing this out. We agree with this comment. Therefore, we have specified the medians, modes and ranges, as well as the range of values, as requested.

Comment 8: l. 362 (p< .001). Italicize p

Response 8: Thank you for pointing this out. We agree with this comment. Therefore, we have italicized p, as requested.

Comment 9: Table 4: left-justify statements.

Response 9: Thank you for pointing this out. We agree with this comment. Therefore, we have left-justified statements in Table 4, as requested.

Comment 10: You didn't ask about digital literacy, so your conclusion shouldn't push on digital literacy.

Response 10: Thank you for pointing this out. We agree with this comment. Therefore, we have omitted the digital literacy from the conclusion

Comment 11: Hyped/over-generalization: The exploration of engaging web-based classroom resources integrated into various 734 formats and applications is paramount in today's education. (the PARAMOUNT part cannot be asserted with just one study, particularly in light of your instruments and method.

Response 11: Thank you for pointing this out. We agree with this comment. Therefore, we have changed this sentence to avoid hyper/overgeneralization, as requested.

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

I have no further comments