Protocol
Make the first judgment before the group talks
The first voice in the room should not silently become part of everybody else's evidence.
When it fits
- Several people will judge the same case and early discussion could anchor everyone on the first confident opinion.
When to avoid it
- Do not force a judgment before reviewers have information that is genuinely necessary and only available collaboratively.
Why it matters
Give each reviewer the same available case information and ask for an initial judgment before discussion. Capture the score or recommendation plus one short reason. Only then reveal the spread and discuss. This preserves information about independent perspectives that disappears once everyone has heard the dominant view.
Steps
- All reviewers receive the same intended information before rating.
- Each initial judgment is recorded before group discussion.
- Reviewers give a brief reason or key evidence, not only a number.
- Initial ratings remain available after the final decision for audit.
An example
Before an architecture review call, each reviewer independently rates the proposed change's operational risk and notes the main reason.
Check your result
The team can see the pre-discussion distribution rather than reconstructing who supposedly thought what afterward.
Keep this limit in mind
- Do not force a judgment before reviewers have information that is genuinely necessary and only available collaboratively.
Connected ideas
Use beforeDiscuss the reasons, then re-rate privately
Evidence and sources
Workplace judgment research recommends collecting independent judgments before social discussion when the goal is to reduce unwanted variability and influence effects.
Independence is less important when the task genuinely requires shared information before any individual judgment is possible.
Improving Workplace Judgments by Reducing Noise: Lessons Learned from a Century of Selection Research · Independent judgments and aggregation
The 2026 AMP procedure asked analysts to make initial ratings independently before group discussion to reduce groupthink while preserving distinct perspectives.
The study involved five advanced music analysts and is a proof of concept rather than a general effect-size estimate.
Analysis from multiple perspectives (AMP): Applying decision hygiene to analysis of musical structure · Decision-hygiene procedure