Effective Education System for Athletes Utilising Big Data and AI Technology

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. Figure 2: Deep learning should be a subset of machine learning

2. Some specific application cases of AI in the field of education are worth analyzing

[1] Sentiment analysis of online course evaluation based on a new ensemble deep learning mode: evidence from Chinese

[2] Artificial intelligence in education: A review 

3. The formats of the tables need to be adjusted, such as Table 5 and Table 6

4. Provide the limitations and future plans of this paper

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

To solve that athletes have difficulty integrating into society after the end of their sports careers, this paper proposes a concept of effective education for athletes based on modern technologies (especially big data and AI), and proves the feasibility of this concept by describing specific examples of the technologies and conducting questionnaires. It is suggested that the author make revisions to the following parts of the article to improve the quality of the paper.

1. Some key words in the abstract mainly focus on technical tools (big data, AI) and educational management, but do not directly reflect one of the core goals of the paper - the career development of athletes after retirement. The core demand mentioned in the research question, "providing an effective education system for athletes to improve their career prospects after retirement", suggests adding keywords such as "career planning", which can enhance the clarity of the paper's theme and the visibility of academic contributions.

2. The introduction part of the paper has the problem of insufficient depth and breadth of the literature review. It does not mention the limitations of the traditional athlete education model and lacks discussion on the "particularity of athlete education", which may lead to readers' incomplete understanding of the research background and make it difficult to highlight the necessity of this study. It is suggested to supplement the classic theories and empirical research in the field of athlete education, clarify the theoretical significance of the research questions, and briefly describe the specific mechanism of AI and big data in educational scenarios.

3. The operational definitions of key variables (such as "educational importance" and "AI attractiveness ") are not clear, and mature scales have not been cited or their reliability and validity have not been verified. It is recommended to cite or develop mature scales to report reliability and validity indicators (such as Cronbach's Alpha, factor analysis results).

4. Although it was mentioned that the Cronbach's Alpha value was 0.722, it was not explained whether this value met the research requirements (it is usually recommended to be ≥0.7), nor was it explained how to handle the problem of low reliability. It is suggested to clarify the selection criteria and analysis logic of secondary cases, and supplement the pre-test and reliability and validity test processes of questionnaire design.

5. Only mentioning the "Informed Consent Statement" but not explaining the specific process (such as how to obtain the consent of the participants) poses ethical risks and may affect the legality and academic reputation of the paper. It is recommended to describe the informed consent process in detail.

6. The verification of some hypotheses (such as H2, H4) is only based on univariate analysis (such as chi-square test), without controlling for the interference of other variables. For example, H2 (exercise success is not related to the importance of education) does not consider the moderating effects of factors such as exercise type and gender. Moreover, the Effect Size or statistical Power Analysis was not reported, making it difficult to evaluate the practical significance of the results. It is recommended to adopt multiple regression analysis to control confounding variables and calculate the effect size and statistical power.

7. When analyzing the relationship between categorical variables (such as exercise success level, gender, and exercise type) and the perception of educational importance, the observation frequency of some cells was relatively low (for example, in the "Unsuccessful" group in Table 4, each option of educational importance was 0 or a very low value). According to the requirements of the chi-square test, the expected frequency of each cell should be ≥5; otherwise, the credibility of the result will be reduced. The paper does not mention the processing of low-frequency cells (such as merging categories or using Fisher's exact test), which may violate the test hypothesis. It is recommended to combine the categories (such as combining "Unsuccessful" and "Below-averagely successful" into "Less Successful"), or use the Fisher's exact test.

8. Hypothesis H5's verification relies on subjective inference (" Cramer's V = 0.172 indicates a weak correlation "), but does not combine theoretical explanations to explain why there is a weak correlation between AI attractiveness and educational importance. It is suggested to explain the weak correlation between the attractiveness of AI and the importance of education in combination with theories (such as self-determination theory).

9. Some of the results did not directly respond to the research questions. For example, RQ3 (" How does big data help athletes learn? ") It was only mentioned in secondary data analysis (such as the Khan Academy case), but the practical application effect of big data technology was not verified through questionnaire data. RQ5 (" Do athletes need AI/ big data skills?" The conclusion was not supported by quantitative data (such as skill mastery level), but only relied on subjective perception (such as "attractiveness" score). It is suggested that the research findings be combined with the career transition theory and gender research literature to propose potential explanations for the contradictions (for example, "high educational importance but low AI usage rate" may reflect the issue of tool availability).

10. The content layout of Table 5 (Gender's Dependence on the Perception of Educational Importance) and Table 6 (The perception of educational Importance depends on the type of movement) is unbalanced and the margins are unreasonable, which undermines the overall aesthetic appeal and reading fluency of the papers. It is suggested to streamline redundant content (reduce the row spacing of the table) and arrange the layout reasonably (center variables and avoid cross-page display of the table).

11. Some charts are disconnected from the textual descriptions or lack necessary textual explanations. For example, Figures 3 (athletes' perception of their own sports success) and 4 (the importance of education) do not clearly state the data distribution characteristics (such as mean and standard deviation) in the main text. Figure 5 (the attractiveness of educational forms) uses a radar chart but does not explain the scoring criteria (such as numerical range and option weights). It is suggested to supplement the statistical indicators (mean, standard deviation, sample size) of the table; Choose a more intuitive chart form (such as a bar chart instead of a radar chart), and clearly mark the data source and unit.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

As coming in a second round of review as a new reviewer I do not feel I should add new comments to address.

The paper reads well, and the research design is sound and in general well written and described.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I do not think this is a good revision, the response seems too simple and crude, which can not see the improvement related to suggestion clear. I want to reject the paper for its careless and impolite.