E01¶


Research	R0002 — Research Standards for AI-Assisted Writing
Run	2026-03-13
Claim	C004
Source	SRC02
Evidence	SRC02-E01
Type	Factual

Mulrow Examined 50 Reviews Against Eight Criteria

URL: Not captured — experimental run

Extract¶

Mulrow examined 50 review articles published in four major medical journals (1985-1986) against eight explicit scientific criteria. Key findings:

None of the 50 reviews met all eight criteria.
Only 1 satisfied six criteria.
32 satisfied four to five criteria.
17 satisfied just three criteria.
Only one had clearly specified methods for identifying, selecting, and validating included information.

Relevance to Hypotheses¶

Hypothesis	Relationship	Strength
H1	Supports	Strong — confirms "most reviews failed basic criteria" is accurate
H2	Neutral	Findings are factual regardless of causal chain interpretation
H3	Contradicts	Strong — the claim is an understatement, not an exaggeration

Context¶

The quantitative findings are devastating: a zero-out-of-fifty result on the full criteria set. The claim that "most reviews failed basic criteria" is technically an understatement — ALL reviews failed to meet the full set. Even the best-performing review only satisfied six of eight criteria. The characterization of "abysmal reporting quality" is editorially strong but substantively defensible given these numbers.

Notes¶

Full text was behind a paywall. Findings were assembled from the PubMed abstract and the James Lind Library curated summary. The quantitative breakdown (1/6, 32/4-5, 17/3) is from the James Lind Library account.