The topic-page groundedness gap between wiki and vector RAG was the study's only comparison not reaching significance at the 95% level¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? On topic-page questions, the 11-point grounded-rate gap between RAG (87%) and wiki (98%) had a 95% CI of [0, 22] that just touches zero at n=45, so RAG's shortfall there was directionally consistent but not statistically established. → Brian Wright 2026

Evidence¶

Brian Wright 2026¶

Brian Wright. (2026). Potential for Enhanced Learning in Machine Learning Classes by Using Wiki LLM Indexing. https://arxiv.org/abs/2609.25303 doi:10.48550/arxiv.2609.25303

q2 · i? · causal · r2

Bootstrap analysis of the groundedness endpoint on the 45 topic-page questions in the head-to-head evaluation. The article reports an 11-point gap with "95% CI [0, 22]" touching zero, and no effect size is printed for this comparison.

"the topic page grounded rate gap (11 points, 95% CI [0, 22]) just touches zero atn= 45, the one comparison in this study that does not reach significance at the 95% level"

Discussion¶

Related Claims¶

• Retrieval failure explained only 2 of vector RAG's 11 ungrounded answers; 9 answers retrieved the correct material and were still ungrounded, suggesting chunked context invites fabrication — related
• The wiki's advantage over vector RAG is roughly three times larger on cross-page questions than on topic-page questions, in both score and groundedness — related