Bifactor-ESEM shows the highest fit among CFA, ESEM, bifactor-CFA and bifactor-ESEM models, supporting AITAQS as one general factor plus four specific dimensions.¶

Claim · All claims
Evidence · 1 study · design r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? The Bifactor-ESEM model exhibited the highest fit among all four compared models. → Zhang 2026

Evidence¶

Zhang 2026¶

Zhang, Z., Katayoun, M., Tan, L., & Wei, G. (2026). Development and Validation of a Multidimensional Scale for AI Tutoring Acceptance in Higher Education. npj Science of Learning. https://doi.org/10.1038/s41539-026-00431-9

q3 · i? · design · r2

Model comparison of CFA, ESEM, bifactor-CFA and bifactor-ESEM on separate formal datasets, summarized in Table 6. The Bifactor-ESEM "exhibited the highest fit among all models"; four-factor CFA was only marginally acceptable.

"the Bifactor-ESEM model exhibited the highest fit among all models. Although the 4factor -CFA model was marginally accep table and supported the dimensional structure of the AITAQS, the ESEM model clearly performed better."

Discussion¶

Related Claims¶

• All AITAQS items load above 0.50 on their target factor with cross-loadings below 0.30, in the hypothesized assignment (X1–X3 Feedback Quality through X10–X12 User Trust). — related
• Perceived social-emotional competence is best modeled as a bifactor structure with one global factor and five specific factors among Australian adolescents — a broader claim this one bears on
• Exploratory structural equation modeling (ESEM) differentiates victimization components better than traditional CFA, which positively biases factor correlations — a broader claim this one bears on
• Higher order cognitive skills (inference, perspective taking, comprehension monitoring) fit a bifactor structure with a general cross-language factor and language-specific factors in English and Spanish — a broader claim this one bears on
• The SRL-O supports both ten lower-order subscales and two superordinate factors (motivational beliefs and learning strategies) — a broader claim this one bears on
• The six achievement emotion regulation strategies are empirically separable constructs, and corresponding strategies across emotions are sufficiently distinct to warrant emotion-specific measures — related
• A bifactor model of oral language and reading fits better than one-, three-, and four-factor models in every grade from 4th through 10th — a broader claim this one bears on
• The AITAQS item pool shows sampling adequacy for factor analysis, with KMO = 0.758 and a significant Bartlett test (χ² = 438.75, df = 66, p < 0.001). — related
• Principal component analysis of the AITAQS extracts four factors accounting for 65.421% of variance, with item loadings from 0.64 to 0.85 matching the four hypothesized dimensions. — related