Self-reported LLM usage collapses at the first team milestone and recovers only partially on later coding-heavy team tasks¶

Claim · All claims
Evidence · 1 study · associational r1 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Usage dropped 42.7 percentage points from HW3 to TP1, then recovered to 75.0% (TP2) and 78.1% (TP3), still below the individual-homework average of 84%. → Sehrish Basir Nizamani 2026

Evidence¶

Sehrish Basir Nizamani 2026¶

Sehrish Basir Nizamani, Zannah Ziew, Saad Nizamani, and Khyati Goyal. (2026). Less Deliberate in Teams: Student LLM Use Across Individual and Collaborative Work. https://arxiv.org/abs/2606.30860 doi:10.48550/arxiv.2606.30860

q2 · i? · associational · r1

Post-task survey data from 96 consenting undergraduates in two sections of an upper-division data analytics course, with per-assignment usage rates computed over the full sample of 96. The quote reports the "drop of 42.7 percentage points" and the recovery rates; no standardized effect size is printed for the trajectory.

"Usage collapsed to 39.6% at TP1, a drop of 42.7 percentage points from HW3, before recovering substantially to 75.0% at TP2 and 78.1% at TP3."

Discussion¶

Related Claims¶

• Verification behaviors degrade most in team contexts, with test-based verification falling and never fully recovering — related
• Error timing and stakes shape trust in AI systems: later errors may be more damaging, early errors may matter more, and trust recovers quickly after good post-error performance, with high-stakes errors causing abrupt collapse — related
• LLM usage varied across assignments and tasks, increasing for later assignments (especially discussion points), while Low-Reliance students concentrated use early and reduced it after Midterm 2 — related
• Team code comprehension dropped in lec09 when Implementation-phase AI usage peaked, and recovered in lec10 following instructor intervention — related
• In SDPBL with SDD introduced, AI usage varies greatly by team and phase, with all teams using AI in Implementation and Debugging but large team differences in Investigation and Planning — related