The naive Bayes model achieved the highest precision (0.75) for ethnic bias detection on the test set¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q2 i? Among the evaluated classifiers, naive Bayes demonstrated the highest precision (0.75 on the test set), making it suitable when precision in bias detection is prioritized. → Josmario Albuquerque 2026

Evidence¶

Josmario Albuquerque 2026¶

Josmario Albuquerque, Bart Rienties, Martin Hlosta, Wayne Holmes. (2026). Learning Analytics to Uncover Ethnic Bias in Educational Texts: An Ensemble Learning Approach. Journal of Learning Analytics 13(1). https://doi.org/10.18608/jla.2026.8905

q2 · i? · design · r2

Test-set evaluation of fine-tuned models on the 69-sample held-out portion of the OER dataset; the results section reports that naive Bayes "demonstrated the highest precision (0.75 on the test set)", though its overall accuracy was low (0.49 per Table 8).

"In contrast, the naive Bayes (NB) model demonstrated the highest precision (0.75 on the test set)."

Discussion¶

Related Claims¶

• SVM and random forest classifiers showed consistent performance in classifying ethnic bias, with F1-scores of 0.71 and 0.70 on the test set — related
• A stacking ensemble of LG, SVM, NB, KNN and XGB achieved the best fine-tuning F1-score (0.88), outperforming individual models such as SVM (0.86) — related
• Simple classifiers (Logistic Regression, Naive Bayes) outperformed tree-based models on this small dataset — related
• The fine-tuned BERT model's classification performance is consistent across stereotype-threatened and non-threatened students and other demographic subgroups — related