An ACO-derived 12-item short form (GenAI-SRL-12) retains the full scale's factor structure, reliability, and construct equivalence¶

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 · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q2 i? CFA of the six-factor 12-item short form showed acceptable fit (CFI = 0.959, RMSEA = 0.068). → Xiaoqi Wang 2026 q2 i3 Short-form dimensions correlated 0.910–0.944 with their full-scale counterparts, indicating construct equivalence. → Xiaoqi Wang 2026 (2)

Evidence¶

Xiaoqi Wang 2026¶

Xiaoqi Wang, Lawrence Jun Zhang, Yanan Zhang. (2026). Generative artificial intelligence-supported self-regulated learning (GenAI-SRL) in L2 writing: scale development, validation, and short-form construction. Metacognition and Learning. https://doi.org/10.1007/s11409-026-09481-1

q2 · i? · associational · r2

Ant Colony Optimization on merged data split into training (n = 453) and validation (n = 194) subsets selected two items per dimension across 10 random seeds. CFA of the resulting GenAI-SRL-12 showed "acceptable fit" with CFI = 0.959 and RMSEA = 0.068; α ranged 0.657–0.849.

"CFA examined the structural validity of the 12-item short form. The six-factor model showed acceptable fit: χ²/df = 1.891, CFI = 0.959, TLI = 0.931, RMSEA = 0.068, SRMR = 0.043"

Xiaoqi Wang 2026 (2)¶

Xiaoqi Wang, Lawrence Jun Zhang, Yanan Zhang. (2026). Generative artificial intelligence-supported self-regulated learning (GenAI-SRL) in L2 writing: scale development, validation, and short-form construction. Metacognition and Learning. https://doi.org/10.1007/s11409-026-09481-1

q2 · i3 · associational · r2

Pearson correlations on the merged dataset between GenAI-SRL-12 and full-scale subscales. Matching dimensions correlated r = 0.910–0.944 (p < 0.001), which the authors read as "construct equivalence between the two versions."

"scale correlated highly with their respective full-scale counterparts ( r = 0.910–0.944, p < 0.001), and the moderate intercorrelations among dimensions ( r = 0.344–0.462, p < 0.001) followed the correlational pattern of the full scale."

Discussion¶

Related Claims¶

• GenAI-supported SRL in L2 writing has a six-factor latent structure spanning cognitive, metacognitive, motivational, affective, social-behavioral, and environmental regulation — a broader claim this one bears on
• All six GenAI-SRL dimensions correlate positively with learners' perceived writing improvement (r = 0.275–0.338) — related
• The GenAI-SRL scale shows satisfactory reliability, convergent validity, and discriminant validity across its six dimensions — related
• The ten-construct instrument showed strong psychometric quality, with the four ARCS dimensions forming a psychometrically sound higher-order motivation construct — related
• The self-report measure supports a refined 13-item, four-factor model with good fit — related