Modelling Student Retention in Tutorial Classes with Uncertainty—A Bayesian Approach to Predicting Attendance-Based Retention
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
For the manuscript entitled "Modelling Student Retention in Tutorial Classes with Uncertainty – A Bayesian Approach to Predicting Attendance-based Retention." please consider the following suggestions for revision:
- The abstract should act as a succinct yet comprehensive summary of your paper. In this sense, it is important to mention the data used in your research in the abstract, to provide readers with a quick understanding of the empirical basis for your research.
- To improve the clarity of the introduction, it should include a brief overview of each section of the paper. This not only aids in reader comprehension but also serves as a guide to the logical flow of your argument.
- The literature review could be more detailed in order to provide a more comprehensive understanding of the retention literature and the various recent models used to address it. Consider discussing the strengths and weaknesses of different approaches and where your research fits within this spectrum. This section should also highlight the theoretical and empirical contributions that your paper makes to the existing body of knowledge.
- Please ensure that the references are formatted consistently throughout the text according to the guidelines of the publication. There are some inconsistencies in the text regarding the presentation of the references [for instance: in line 49 the reference is (Wakelin, 49 2022) but in line 26 the reference is [1].]
Author Response
Comment 1:
The abstract should act as a succinct yet comprehensive summary of your paper. In this sense, it is important to mention the data used in your research in the abstract, to provide readers with a quick understanding of the empirical basis for your research.
Response:
Updated the abstract to have more information on the data used in the study
Comment 2:
To improve the clarity of the introduction, it should include a brief overview of each section of the paper. This not only aids in reader comprehension but also serves as a guide to the logical flow of your argument.
Response:
Added an overview section at the end of the introduction section (page 2 line 84-93)
Comment 3:
The literature review could be more detailed in order to provide a more comprehensive understanding of the retention literature and the various recent models used to address it. Consider discussing the strengths and weaknesses of different approaches and where your research fits within this spectrum. This section should also highlight the theoretical and empirical contributions that your paper makes to the existing body of knowledge.
Response:
Added more literature and provided more explanations on Bayesian modelling under the Bayesian Modeling section (page 4)
Comment 4:
Please ensure that the references are formatted consistently throughout the text according to the guidelines of the publication. There are some inconsistencies in the text regarding the presentation of the references [for instance: in line 49 the reference is (Wakelin, 49 2022) but in line 26 the reference is [1].
Response:
Updated the reference list and in-text citations across the paper (all pages)
Reviewer 2 Report
Comments and Suggestions for Authors
The research has been well conducted and is methodologically sound. The article addresses an important issue in higher education: student retention, using a new Bayesian approach that adds value to the existing literature.
The paper is based on an exhaustive literature review, which provides a solid theoretical basis and reveals clear gaps that the research aims to fill.
As for the methodology, the adoption of the KDD framework ensures rigor, and the application of BART compared to RFR is well conducted. The detailed results and insightful analyses, in particular the discussion of the HDI to quantify uncertainty, are pertinent.
The study has value for educational institutions and policymakers in that it demonstrates its practical relevance. The discussion of the results is balanced; it recognises limitations and makes suggestions for future research directions.
However, some areas could be improved to ensure a stronger paper:
1. A brief discussion of the computational efficiency and practical implementation aspects of the BART model.
2. Evidence of a case study or practical application of the model would strengthen the validation and demonstrate its usefulness in real-world scenarios.
3. Greater fluidity of the narrative is needed, and concern that complex statistical concepts are clearly explained to a wider audience, including those who may not be experts in Bayesian methods, benefits understanding and replication of this study in other contexts.
Overall, the robustness of the methodology, the rigour of the analysis, and the practical implications make this article a positive contribution.
Author Response
Comment:
A brief discussion of the computational efficiency and practical implementation aspects of the BART model.
Response:
Added more detailed explanations on the Bayesian modelling and an introduction section for Bayesian modelling (page 4)
Comment 2:
Evidence of a case study or practical application of the model would strengthen the validation and demonstrate its usefulness in real-world scenarios.
Response:
Added a brief discussion on the BART models computational efficiency and practical implementation at the end of the results section (page 12).
Reviewer 3 Report
Comments and Suggestions for Authors
The presented analysis is focused on selected student retention rates and selected prediction models. It is worth considering other models and variables when discussing the results. Discussion of the results should precede conclusions.
Author Response
Comment 1:
The presented analysis is focused on selected student retention rates and selected prediction models. It is worth considering other models and variables when discussing the results. Discussion of the results should precede conclusions.
Response:
The manuscript has been revised to ensure that the conclusion provides detailed ending remarks that is in line with the results section (page 12 and 13)
