Q003 — Self-Audit¶


Research	R0029 — Plural Voice Attribution
Run	2026-03-27
Query	Q003

ROBIS 4-Domain Audit¶

Domain 1: Eligibility Criteria¶

Rating: Low risk

Criterion	Assessment
Defined what counts as "formal policy"	Yes — official editorial policy, call-for-papers guidelines, or published policy statements
Applied consistently	Yes — same threshold for all venues

Notes: Clear eligibility criteria held throughout.

Domain 2: Search Comprehensiveness¶

Rating: Low risk

Criterion	Assessment
Multiple search strategies	Yes — publisher-focused and conference-focused searches
Named venues covered	Yes — all 6 named venues (Nature, Science, ACM, IEEE, ICML, NeurIPS) plus 5 publishers
All results dispositioned	Yes — 50 results across all searches
Source diversity	Yes — journals, conferences, publishers, professional societies

Notes: 5 targeted searches, 50+ total results, 6 sources selected.

Domain 3: Evaluation Consistency¶

Rating: Low risk

Criterion	Assessment
All sources scored	Yes — 6 scorecards
ACH matrix applied	Yes
Diagnosticity analysis	Yes

Notes: Consistent application across all venue types.

Domain 4: Synthesis Fairness¶

Rating: Low risk

Criterion	Assessment
All hypotheses tested	Yes
Science's policy reversal noted	Yes — evolution documented
Variation acknowledged	Yes — disclosure spectrum explicitly mapped
Gaps acknowledged	Yes — ICML gap, enforcement gap, language gap noted

Notes: The two-layer finding (consensus on principle, variation on implementation) emerged from evidence rather than being imposed.

Overall Assessment¶

Overall risk of bias: Low risk

This query had the most comprehensive evidence base of all five queries, with primary policy documents from all named venues. The main risk was availability bias (English-language, Western-centric venues), which was acknowledged.

Researcher Bias Check¶

Confirmation bias (low risk): The query names specific venues, creating an expectation of finding policies. This was appropriate — the venues were chosen because they are major. No bias from the venue selection.
Completeness bias (low risk): ICML's limited AI-specific policy was noted as a gap rather than ignored.