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¶

Claim · All claims
Evidence · 1 study · causal r1 · q2 · i2 medium
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i2 After a five-day AI agent creation workshop, students showed significant pre-post improvements in all three CT self-perception dimensions, with d = 0.71 for abstract thinking, d = 0.70 for algorithmic thinking, and d = 0.35 for pattern recognition and generalization. → 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 · i2 · causal · r1

Paired-samples t-tests (N=84) comparing pre-post CT Self-Perception Scale scores after the five-day workshop; all three dimensions improved significantly, with d = 0.71 and d = 0.70 described as "large effect sizes" and d = 0.35 as a medium effect.

"Abstract thinking (𝑑 = 0.71) and algorithmic thinking (𝑑 = 0.70) showed large effect sizes, while pattern recognition and generalization demonstrated a medium effect ( 𝑑 = 0.35)."

Discussion¶

Related Claims¶

• Iterative testing engagement during AI agent creation predicts CT self-efficacy gains after controlling for initial self-efficacy — related
• SDT-driven activity system redesign in Cycle 2 produced significant pre-post gains in AI-TPACK (d = 0.70), IPACK (d = 0.86), and AI attitude (d = 0.58) — related
• Students reported significant perceived gains in abstract and critical thinking, with other higher-order thinking effects ambiguous — related
• High school novices show large CT learning gains after using RoboBlockly Studio, improving Bebras performance from 1.28 to 2.88 of 4 tasks — 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
• Secondary students report near-maximum confidence in passive digital tasks but sharply lower confidence in active creation and algorithmic logic — related