Empirical Investigation of Multilayered Framework for Predicting Academic Performance in Open and Distance Learning
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsAlthough this article adopts a novel approach to understanding artificial intelligence usage in open design learning settings in higher education, the article is actually a systematic literature review that does not adequately express and explain its content analysis for coding articles, from which quantitative data was derived and quantitative processes were conducted. After reading the article, I am very unclear about how each article within the systematic review was quantified or coded to generate quantitative information in order to develop a structural equation model. There needs to be much more detail, included regarding how the articles were selected, how the articles were eliminated from the review, based on systematic processes, and how each article was quantified to produce numeric measurements in order to establish a structural equation model.
Comments on the Quality of English LanguageModerate editing of English is necessary--there are many areas where grammar and mechanics must be addresssed.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper presents a sophisticated multilayered framework integrating multiple AI technologies to predict academic performance. This approach enhances existing models by combining various analytical methods such as Structural Equation Modeling (SEM) and Support Vector Machine (SVM), offering a robust alternative to single-method analyses.
One of the strengths of this paper is its commitment to inclusivity, considering gender and regional disparities in the application of AI in education. This alignment with UNESCO’s 2030 educational goals demonstrates a profound understanding of the need for equitable education. Furthermore, the paper doesn't just propose a theoretical model; it includes preliminary empirical testing and validation, which adds to its practical relevance and applicability in real-world scenarios. I suggest to add some literature review about the use of AI in educational models (https://doi.org/10.3390/app13116716 or https://doi.org/10.1108/ITSE-11-2023-0218).
However, the complexity of the proposed framework, while a technical strength, could pose accessibility challenges for practitioners not well-versed in advanced analytical techniques. The successful implementation of this framework heavily depends on the availability of high-quality data and advanced AI technologies, which might not be readily available in all educational settings. Moreover, the current empirical validation focuses only on initial layers of the framework, indicating that comprehensive validation involving all layers is still pending.
To enhance the practicality and accessibility of the framework, it would be beneficial to simplify it or provide a tiered approach that allows educational institutions to implement components based on their specific needs and capabilities. Expanding empirical testing to include all layers and using a more diverse dataset could further validate the framework's effectiveness across various global educational environments.
Additionally, there is a potential risk of overfitting due to the use of multiple analytical methods and the complexity of the framework. To mitigate this, the study could focus on developing training programs and user-friendly tools that assist practitioners in implementing the framework. Strengthening the focus on ethical considerations, particularly regarding data privacy and algorithm biases, is also crucial.
In conclusion, while the paper makes substantial contributions to the theoretical and practical understanding of AI's role in education, further work is needed to ensure its adaptability, effectiveness, and ethical application across diverse educational settings. Longitudinal studies could also be beneficial to observe the long-term effects and sustainability of the AI-driven interventions proposed by the framework.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsWith respect to the starting point of this work, based on reference [5], it is advisable to provide more information on the characteristics of multilayer studies and how they help to deepen the introduction of AI in the educational field.
The problem statement is clear and well presented. However, the authors should provide a more detailed discussion of why and how a knowledge gap exists in the work's study area. This is very important to help readers understand why a new solution is needed.
The objectives are clearly outlined, but could have more impact if specific performance objectives or benchmarks were provided.
The bibliographical references are valuable and completely current, including a total of 39 references, most of them being published in the last five years, except for two references. The geographical scope of all of them is varied and they have international impact.
The methodological section describes the procedures and processes developed in the literature review. However, some aspects are noted that should be included to improve the quality of the review.
It is advisable to delve into the “comprehensive search strategy” used, including results obtained, evaluated and eliminated from each of the databases, being able to use the PRISMA method. Likewise, it would be appropriate to include the keywords used for the searches, the Boolean operators included and the search equations used in each of the query databases.
It is necessary to justify why such a long search period is used (last 8 years), since the impact generated by AI has had its greatest impact in the last 3 or at most, 5 years, making it more relevant to focus attention on publications in these years.
Likewise, it is advisable to incorporate the measures that have been followed to guarantee validity and reliability in carrying out the systematic search, as well as in the use of the model applied for data analysis.
The results adjust to the processes developed in the methodological section. They are represented in a structured way and respond to the proposed research objectives. The description and interpretation of the results found is carried out in the discussion section, offering adequate depth. However, it would be advisable to generate lines of discussion with the findings found, based on previous models, which allows readers to see the improvements of the model used.
The conclusions section is relevant, providing the most relevant findings of the study and including recommendations for educational institutions. It also delves into both the limitations found and the future implications, being one of the strongest and best prepared sections of this document.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThank you for adhering to the reviewer feedback as now this paper is poised to make a substantial contribution to the literature.
Comments on the Quality of English LanguageMinor edits required
