Teams whose perception of who uses more AI reversed between t1 and t3 achieved lower scores than teams with stable perceptions¶

Claim · All claims
Evidence · 1 study · associational r2 · q2 · i2 medium
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i2 Reversal teams scored lower (medians 94.5 vs. 98 percentage points; Mann–Whitney U = 192, p = .011, r_rb = .427), and the association held in robust regression controlling for initial misalignment and prior performance (b = −1.90, t = −2.32, p = .021). → Laura Graf 2026

Evidence¶

Laura Graf 2026¶

Laura Graf, Ramona Beinstingel, Stephan Krusche, Oleksandra Poquet. (2026). Students' Perception Accuracy of Partners' AI Use and its Relation to Collaboration Performance. Preprint, Technical University of Munich. https://arxiv.org/abs/2606.23237 doi:10.48550/arxiv.2606.23237

q2 · i2 · associational · r2

Mann–Whitney U comparison of teams with versus without a t1→t3 perception reversal in the 103-pair study; the article prints the rank-biserial effect size "r_rb = .427" (with U = 192, p = .011). Only 25 teams experienced a reversal, so the authors call the result indicative rather than conclusive.

"Teams experiencing a direction reversal achieved lower scores than teams with stable perceptions (medians: 94.5 vs. 98 percentage points; Mann–Whitney U = 192, p = .011, r_rb = .427)."

Discussion¶

Related Claims¶

• Perception alignment improves mainly in the early weeks and does not consistently decrease through the project — related
• Greater early-project misalignment in partners' perceptions of each other's AI use is associated with lower final project scores — related
• Prior programming performance moderates the misalignment–performance association, with the cost concentrated in lower-performing teams — related
• Week 3 performance differs by perception-transition group, with Aligned teams increasing and Multifaceted teams decreasing in scores — related