CDM-AIR declares four falsification conditions, including moderation nullity and context redundancy¶

Claim · All claims
Evidence · 1 study · theoretical r3 · q1
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¶

q1 i? If the H12 and H13 interaction terms are repeatedly indistinguishable from zero in adequately powered studies, the model reduces to TPB with a re-labelled antecedent; if facilitating conditions predict outcomes directly while paths through SN and PBC are null, context should be modelled as a direct determinant as in UTAUT. → Mnguni L 2026

Evidence¶

Mnguni L 2026¶

Mnguni L (2026) A Capability–Decision Model of teacher readiness for AI integration in teaching. Front. Educ. 11:1897632. https://doi.org/10.3389/feduc.2026.1897632

q1 · i? · theoretical · r3

Section 5.3 of the conceptual article specifies four disconfirmation conditions: mediation failure against an additive specification, discriminant collapse, moderation nullity, and context redundancy. Reporting against these conditions, not just path significance, is what would allow cumulative evaluation.

"if the H12 and H13 interaction terms are repeatedly indistinguishable from zero in adequately powered studies (N ≥ 400 with latent interaction estimation), the model’s account of the intention–behaviour gap fails, and CDM-AIR reduces to TPB with a re-labelled antecedent."

Discussion¶

Related Claims¶

• Capability and PBC must be separated by a priori discriminant-validity criteria including HTMT below .85 — related
• Detecting that students are tracked to teachers does not necessarily imply bias in teacher effectiveness estimates — a broader claim this one bears on
• The Rothstein falsification test can reject value-added models even when estimated teacher effects are unbiased — a broader claim this one bears on
• Five testable predictions (H1-H5) are specified with falsification conditions but await empirical validation — related
• Prior comparative evidence shows small intention–norm and intention–control associations and a TPB model significant in Thailand but not South Africa — related
• In its traditional treatment, germane cognitive load is essentially indistinguishable from intrinsic load and may be a redundant concept — related
• Germane cognitive load can be defined in terms of intrinsic cognitive load, associating it with element interactivity — related