An effort-based reward did not significantly reduce answer offloading or improve unaided performance¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r3 · q3
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q3 i? The reward's association with lower odds of answer offloading was not significant (OR = 0.66, p = .139). → Sebastian Maier 2026 q3 i? Participants who received the reward did not have higher odds of answering an unaided final test item correctly (OR = 0.76, p = .890). → Sebastian Maier 2026 (2)

Evidence¶

Sebastian Maier 2026¶

Sebastian Maier, Kai Schwabe, Manuel Schneider, and Stefan Feuerriegel. (2026). Designing Against Deskilling: Metacognitive Feedback Reduces Cognitive Offloading to LLM Assistants. arXiv. https://arxiv.org/abs/2609.20143 doi:10.48550/arxiv.2609.20143

q3 · i? · causal · r3

Confirmatory item-level mixed-effects logistic regression (H3a) in the preregistered experiment, testing the effort-based reward main effect on answer offloading. The article reports the effect was "not significant" with OR = 0.66, 95% CI [0.31, 1.40], p = .139; no effect size beyond the OR is printed.

"The reward was associated with lower odds of answer offloading, but this effect was not significant (𝑏=− 0.42, SE= 0.39,𝑂𝑅= 0.66, 95% CI[0.31, 1.40],𝑧=− 1.08,𝑝=. 139)"

Sebastian Maier 2026 (2)¶

Sebastian Maier, Kai Schwabe, Manuel Schneider, and Stefan Feuerriegel. (2026). Designing Against Deskilling: Metacognitive Feedback Reduces Cognitive Offloading to LLM Assistants. arXiv. https://arxiv.org/abs/2609.20143

q3 · i? · causal · r3

Confirmatory test of H3b on unaided final-test performance in the same experiment. The article reports participants "did not have higher odds of answering an unaided final test item correctly" with OR = 0.76, 95% CI [0.50, 1.18], p = .890; equivalence was not tested.

"Participants who received the reward also did not have higher odds of answering an unaided final test item correctly (𝑏=− 0.27, SE= 0.22,𝑂𝑅= 0.76, 95% CI[0.50, 1.18],𝑧=− 1.23,𝑝=. 890)"

Discussion¶

Related Claims¶

• Feedback interventions show phase-specific effects: advice feedback reduces reminder bias, but ranking feedback alters beliefs without increasing offloading — the general claim this finding bears on, with the studies behind it
• Greater answer offloading during learning is associated with lower subsequent unaided performance (exploratory) — related
• Access to an LLM assistant during practice did not significantly lower subsequent unaided test performance — related
• The effort-based reward reduced how often participants used the assistant, whereas metacognitive feedback changed how they used it — related
• Unstructured LLM use risks cognitive offloading that displaces effortful thought and weakens argument construction — related
• Higher-value intentions are more likely to be offloaded than lower-value ones — related
• In one preregistered experiment, metacognitive feedback on learners' own LLM use lowered the odds of answer offloading and raised the odds of a correct unaided answer on one-sided tests, with 95% intervals reaching no effect — related