E01¶


Research	R0002 — Research Standards for AI-Assisted Writing
Run	2026-03-13
Claim	C005
Source	SRC02
Evidence	SRC02-E01
Type	Factual

"Other Bias" Domain Removed in RoB 2

URL: Not captured — experimental run

Extract¶

The original RoB tool included an "other bias" domain where funding source (15.6% of assessments) and conflict of interest (5.8% of assessments) were frequently evaluated. RoB 2 removed this domain entirely, replacing it with the five specific domains. The removal means there is no mechanism within RoB 2 for assessing funding or COI as sources of bias.

Relevance to Hypotheses¶

Hypothesis	Relationship	Strength
H1	Supports	Strong — documents the specific removal and its impact
H2	Neutral	Documents the removal but does not address whether it was deliberate
H3	Contradicts	Strong — the "other bias" domain was the mechanism, and it was removed

Context¶

The quantitative data on RoB 1 usage (15.6% funding, 5.8% COI in "other bias") demonstrates that the community was actively using the RoB tool for COI assessment. The removal of this capability in RoB 2 is therefore not merely theoretical — it eliminates a function that was being used in practice.

Notes¶

The 15.6% and 5.8% figures are from a systematic analysis of published Cochrane reviews, providing empirical evidence of how the "other bias" domain was used.