Binary classification detects slope-height and block-slope difficulties earlier than other difficulties¶

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

Subclaims¶

q2 i? For the SH-rest split, MCC exceeds 0.7 after 15 s; for BS-rest, 0.7 is reached after 35 s; but GUE-rest cannot be recognized well (MCC fluctuates around 0). → Kennel 2025

Evidence¶

Kennel 2025¶

Kennel, K., Ishimaru, S., Küchemann, S., Steinert, S., Kuhn, J., & Ruzika, S. (2025). Gaze-Based Prediction of Students' Math Difficulties - A Time Dynamic Machine Learning Approach to Enable Early Individual Assistance. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-024-00447-5

q2 · i? · associational · r2

Binary classification results from the same eye tracking study with 143 students. The article reports that for SH-rest "the MCC already exceeds a value of 0.7 after 15 s" and for BS-rest "a value of 0.7 is reached after 35 s".

"For the first split, the MCC already exceeds a value of 0.7 after 15 s and after 30 s it is already above 0.8. For the second split, a value of 0.7 is reached after 35 s, 0.8 is achieved after 45 s and after 50 s the values remain constantly high between 0.85 and 0.9."

Discussion¶

Related Claims¶

• The five difficulty groups differ significantly on all 15 eye tracking metrics — related
• On one graphical differentiation item, eye tracking data predicted students' specific difficulties as a multiclass classification problem — related