Under controlled synthetic MNAR bias, TSDR-equipped models improve across the entire bias spectrum, with the RMSE advantage diminishing beyond gamma > 0.8¶

Claim · All claims
Evidence · 1 study (2 entries) · 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? On six synthetic datasets with MNAR degree gamma from 0.0 to 0.999, TSDR consistently yields improvements across the entire spectrum as baselines degrade. → Peilin Zhan 2026 q2 i? For RMSE, TSDR is highly effective at reducing error when MNAR is low, though this advantage diminishes slightly beyond gamma > 0.8; positive results at gamma = 0.0 indicate the framework is safe. → Peilin Zhan 2026 (2)

Evidence¶

Peilin Zhan 2026¶

Peilin Zhan, Wei Chen, Weilin Chen, Shuyi Pan, Ruichu Cai. (2026). Temporal Smoothness Doubly Robust Learning for Debiased Knowledge Tracing. https://arxiv.org/abs/2605.05958 doi:10.48550/arxiv.2605.05958

q2 · i? · design · r2

Simulation study (RQ2, Figure 3) on six synthetic datasets with controllable MNAR mechanism, gamma in {0, 0.2, 0.4, 0.6, 0.8, 0.999}, all six backbones with and without TSDR. The article reports TSDR "consistently yields improvements across the entire spectrum."

"we observe varying degrees of performance degradation across all baselines as the bias severity increases, particularly visible in AUC. However, equipping SOTA mod- els with TSDR consistently yields improvements across the entire spectrum."

Peilin Zhan 2026 (2)¶

Peilin Zhan, Wei Chen, Weilin Chen, Shuyi Pan, Ruichu Cai. (2026). Temporal Smoothness Doubly Robust Learning for Debiased Knowledge Tracing. https://arxiv.org/abs/2605.05958

q2 · i? · design · r2

Simulation study (RQ2) RMSE analysis across MNAR degrees. The article reports TSDR "is highly effective at reducing error when MNAR is low, although this advantage diminishes slightly beyondγ >0.8."

"Specifically for RMSE, with the exception of DKT, TSDR is highly effective at reducing error when MNAR is low, although this advantage diminishes slightly beyondγ >0.8."

Discussion¶

Related Claims¶

• TSDR consistently improves calibration (ACC and RMSE) even on datasets where AUC gains plateau — related
• TSDR improves the AUC of six state-of-the-art KT backbones across nine real-world benchmark datasets — related
• TSDR's AUC gains are largest on open-ended platforms and limited on EdNet, whose mandatory bundle completion policy constrains selection bias — related
• TSDR is robust to the smoothness coefficient, with a stable sweet spot for lambda in (0.3, 1.0) — related