Students with moderate initial CT levels show substantially greater self-efficacy gains than high-CT and low-CT peers, an 'Optimal Development Zone' effect¶

Claim · All claims
Evidence · 1 study (2 entries) · associational r2 · q2 · i3 large
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i3 K-means clustering identified three profiles, and ANOVA showed significant gain differences [F(2,84)=51.64, p<.001, η²=.55], with the Moderate-CT group significantly outperforming both High-CT (Δ=17.62, p<.001) and Low-CT groups (Δ=18.88, p<.001). → Yimeng Sun 2026 q2 i? High-CT and Low-CT groups showed similar, modest gains that did not differ significantly (mean difference=-1.27, p=.761). → Yimeng Sun 2026 (2)

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 · i3 · associational · r2

ANOVA on self-efficacy gains across three K-means-clustered CT-level profiles (High-CT n=37, Moderate-CT n=21, Low-CT n=26, N=84) with η² = .55; the moderate group's gains significantly exceeded both other groups (Δ=17.62 and Δ=18.88, p<.001).

"ANOVA revealed significant differences [𝐹(2,84)=51.64,𝑝<.001,𝜂²=.55]. Moderate-CT students demonstrated the greatest gains (Table 4), significantly outperforming both High-CT (𝛥=17.62,𝑝<.001) and Low-CT groups (𝛥=18.88,𝑝<.001)."

Yimeng Sun 2026 (2)¶

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

q2 · i? · causal · r1

Post-hoc comparison in the same ANOVA of gains by CT-level group: the Low-CT versus High-CT gain difference was not statistically significant, so equivalence was not tested; both extreme groups showed constrained gains.

"High-CT and Low-CT groups showed sim- ilar, modest gains (mean difference=-1.27, 𝑝=.761), despite vast differences in abso- lute ability levels."

Discussion¶

Related Claims¶

• Teaching a ChatGPT agent improves students' self-regulated learning, specifically self-efficacy and use of cognitive strategies — related
• Latent profile analysis of pre-service teachers' CT skills identifies three profiles (Novice, Developing, Proficient), revealing heterogeneous, non-linear skill acquisition — 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
• 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
• Exploratory K-means clustering identifies four heterogeneous student AI-user profiles differing in experience, competence, usefulness, trust, and control, but the solution's optimality and stability are not established — related
• Iterative testing engagement during AI agent creation predicts CT self-efficacy gains after controlling for initial self-efficacy — related