Claims

Unit-test-based evaluation shows consistently negative correlations with human judgment, disproportionately penalizing complex high-quality interactive designs

Claim · All claims
Evidence · 1 study · design r2 · q2
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

Reading the evidence codes

q is the design tier, i the impact, n the sample, and the kind with its r the rigour judged within that kind: what each code means.

This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.

q2 i? The unit test baseline correlates negatively with human ratings across all dimensions (interactivity r = −0.600), suggesting brittle test scripts penalize complex but high-quality interactive designs. → Xiaozao Wang 2026

Evidence

Xiaozao Wang 2026

Xiaozao Wang, Zhewei Wang, Hongyi Wen. (2026). Evaluating Interactivity: Toward Automated Assessment of AI-Generated Explorable Explanations. https://arxiv.org/abs/2606.31012 doi:10.48550/arxiv.2606.31012

q2 · i? · design · r2

Authors' interpretation of the balanced-dataset results in the paper, where the unit test baseline using LLM-generated Playwright tests showed negative correlations with human judgments on all four dimensions, including r = −0.600 on interactivity per Table 1.

"Unit testing shows consistently negative correlations, suggesting that brittle test scripts disproportionately penalize complex but high-quality interactive designs."

Discussion

Page metadata
Field Value
type claim
id unit-test-negative-correlation-human-judgment
title Unit-test-based evaluation shows consistently negative correlations with human judgment, disproportionately penalizing complex high-quality interactive designs
description Unit-test-based evaluation shows consistently negative correlations with human judgment, disproportionately penalizing complex high-quality interactive designs
status review
evidence_strength weak
newest_evidence 2026
recheck_by 2029
generated by process:wiki-ingest on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — xiaozao-wang-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.