Automated CTML metrics show significant improvement in temporal contiguity and coherence for CTML-informed videos, while modality, redundancy, and image quality show no significant difference¶

Claim · All claims
Evidence · 1 study · causal r1 · q2
Currency · no dated study recorded · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? Across 14 videos (7 topics × 2 conditions), temporal contiguity (0.294 vs. 0.273, p=.021) and coherence (0.729 vs. 0.646, p=.011) improved significantly; the other three metrics showed no significant difference. → metric-eval-1

Evidence¶

metric-eval-1¶

Yearim Kim, Injun Baek, and Nojun Kwak. 2026. When Saying No Makes Better Videos: Designing Dual Gatekeeping for Pedagogically Grounded AI Content Creation. In Proceedings of Understanding and Engaging Critical Resistance to AI in Education (CHI '26 Workshop). ACM, New York, NY, USA. https://arxiv.org/abs/2608.19812 doi:10.48550/arxiv.2608.19812

q2 · i? · causal · r1

Algorithmic assessment of a corpus of 14 videos (7 topics × 2 conditions) from established science and philosophy curricula, generated with identical structure so CTML-informed generation and review together were the sole variable; the design cannot separate the two. The article reports "temporal contiguity (0.294 vs. 0.273,𝑝=. 021) and coherence (0.729 vs. 0.646,𝑝=. 011)"; no standardized effect size is printed.

"Two of the five metrics showed significant improvement: temporal contiguity (0.294 vs. 0.273,𝑝=. 021) and coherence (0.729 vs. 0.646,𝑝=. 011). The remaining three showed no significant difference"

Discussion¶

Related Claims¶

• Human ratings and automated metrics converge on coherence and temporal alignment as the dimensions most enhanced by the dual gatekeeping pipeline — related
• A narrative review cites one study in which simultaneous notes, worked example and explanation aided understanding, and another in which temporally contiguous instruction lowered perceived cognitive load — related
• CTML-guided educator review of AI video scripts yields statistically significant improvements across all 12 CTML principles and overall instructional validity — related
• A deterministic slide-image override converts a 0/9 corpus-grounding failure into 9/10 successful slide matches on the same topic — related
• The pipeline feasibly produces coherent, synchronized educational videos from textbook queries, integrating retrieval, image generation, animation, and speech — related