Iterative testing engagement during AI agent creation predicts CT self-efficacy gains after controlling for initial self-efficacy¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2026 · checked against its sources 2026-10-17 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? Hierarchical regression identified iterative testing engagement as the sole behavioral predictor of CT self-efficacy gains (β = .20, p = .050, ΔR² = .035) after controlling for initial self-efficacy; no standardized effect size is printed for this finding. → Yimeng Sun 2026

Evidence¶

Yimeng Sun 2026¶

Yimeng Sun, Haiyang Xin, Qiannan Niu, Shuang Li, Lingyun Huang, Gaowei Chen. (2026). Computational Thinking Development in AI Agent Creation: A Mixed-Methods Study. https://arxiv.org/abs/2605.14330 doi:10.48550/arxiv.2605.14330

q2 · i? · associational · r2

Hierarchical regression on N=80 students using Days 4-5 platform behavioral logs; iterative testing engagement was the sole behavioral predictor of self-efficacy gains, a statistically significant regression coefficient with no standardized effect size printed, bootstrap validation (BCa 95% CI [1.10, 116.78]) and cross-validation (R² = .38) reported as confirming model stability.

"After controlling for initial self-efficacy (𝛽 = −.52,𝑝 < .001), iterative testing positively predicted gains (𝛽 = .20,𝑝 = .050,𝛥𝑅² = .035)."

Discussion¶

Learner Variables¶

• Motivation — predictor: learners who differ on it differ in outcomes
• Self-Efficacy — outcome: instruction changes it

Related Claims¶

• Students' argumentative essay quality improves substantially from initial to revised drafts regardless of feedback source — the general claim this finding bears on, with the studies behind it
• Five-day no-code AI agent creation improves students' CT self-perception, with large gains in abstract and algorithmic thinking and a medium gain in pattern recognition and generalization — related
• Markov transition analysis shows the three clusters differ in workflow dynamics: iterative optimization loops, rapid prototyping flows, and browsing-anchored exploration — related
• Teacher-AI learning task design objectives cluster into cognitive gains and thinking skills, behavioral changes and design efficiency, and adaptivity, engagement, self-efficacy, and inclusivity — related
• Problem decomposition, artifact creation, data practices, and testing/debugging appear in all 28 reviewed CT manuscripts, while iterative refinement appears in 17 — related
• Teaching a ChatGPT agent improves students' self-regulated learning, specifically self-efficacy and use of cognitive strategies — related
• Platform logs show behavioral shifts toward goal-oriented optimization between workshop Days 4 and 5 — related
• Students with moderate initial CT levels show substantially greater self-efficacy gains than high-CT and low-CT peers, an 'Optimal Development Zone' effect — related