Machine learning models using whole-module fNIRS features predict transfer scores in all feedback groups, with the frontopolar area a predictor regardless of feedback type¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? The Extra Trees model using brain activity from the entire learning module predicted transfer scores best across all groups, with mean R² of 0.88 (MF), 0.90 (AF), and 0.87 (NF), and identified key predictive regions including FP, SMG, DLPFC, and MTG. → Yin 2025

Evidence¶

Yin 2025¶

Yin, J., Xu, H., Pan, Y., & Hu, Y. (2025). Effects of different AI-driven Chatbot feedback on learning outcomes and brain activity. npj Science of Learning. https://doi.org/10.1038/s41539-025-00311-8

q2 · i? · causal · r2

Machine learning prediction analysis (Extra Trees ensemble with SHAP feature importance) in each feedback group; models using the whole learning module outperformed single-phase models. Key regions "that predicted transfer scores included the FP, SMG, DLPFC, and fusiform gyrus" in the MF group; SMG, DLPFC, MTG, FP in AF; SMG, MTG, FP, DLPFC in NF.

"For the MF group, the results revealed that the mean R2 achieved 0.88, the mean MAE achieved 0.72, and the mean MSE achieved 0.93."

Discussion¶

Related Claims¶

• Feedback types elicit distinct fNIRS activation patterns: frontopolar and middle temporal gyrus higher under metacognitive feedback, supramarginal gyrus under affective feedback, DLPFC under neutral feedback — related
• In the metacognitive feedback group, frontopolar activation correlates with metacognitive sensitivity and middle temporal gyrus activation correlates with retention — related