Academic Journal
Using Interpretable Machine Learning Approaches to Predict and Provide Explanations for Student Completion of Remedial Mathematics
| العنوان: | Using Interpretable Machine Learning Approaches to Predict and Provide Explanations for Student Completion of Remedial Mathematics |
|---|---|
| اللغة: | English |
| المؤلفون: | Thomas Mgonja (ORCID |
| المصدر: | Education and Information Technologies. 2024 29(16):22287-22312. |
| الاتاحة: | Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/ |
| Peer Reviewed: | Y |
| Page Count: | 26 |
| تاريخ النشر: | 2024 |
| نوع الوثيقة: | Journal Articles Reports - Research |
| Education Level: | Higher Education Postsecondary Education |
| Descriptors: | Higher Education, Remedial Mathematics, Artificial Intelligence, Predictor Variables, Comparative Analysis, Grade Point Average, Educational Background, Teacher Effectiveness |
| DOI: | 10.1007/s10639-024-12647-6 |
| تدمد: | 1360-2357 1573-7608 |
| مستخلص: | The successful completion of remedial mathematics is widely recognized as a crucial factor for college success. However, there is considerable concern and ongoing debate surrounding the low completion rates observed in remedial mathematics courses across various parts of the world. This study applies explainable artificial intelligence (XAI) tools to interpret predictions on whether students will complete mathematics remediation. Various machine learning models are compared, with random forest emerging as superior in predicting non-completion. Global interpretations using correlation analysis, logistic regression, feature importance, permutation importance, and SHapley Additive exPlanations (SHAP) summary plots identify significant predictors such as college grade point average (G.P.A), high school G.P.A, starting point in the remedial sequence, number of failed remedial courses, delay in remediation, Rate My Professor scores, and age. Additionally, local interpretations using Local Interpretable Model-Agnostic Explanations (LIME) and Diverse Counterfactual Explanations (DiCE) analyses were utilized to garner tailored advise for at-risk students. It was observed that instructor attributes cannot be overlooked, especially, when exploring local interpretations. Future research should consider other features such as a students' socio-economic status (SES), employment status, and placement exam scores. Future studies could also involve data from multiple institutions and examine user experience in implementing these models. |
| Abstractor: | As Provided |
| Entry Date: | 2024 |
| رقم الانضمام: | EJ1450655 |
| قاعدة البيانات: | ERIC |
| تدمد: | 1360-2357 1573-7608 |
|---|---|
| DOI: | 10.1007/s10639-024-12647-6 |