ThinkCritical.eu · Manipulation atlas
Group decisionsAveraging away disagreement
Differing estimates are averaged or set aside instead of asking why they differ – and what only one person knew gets lost.
When people's estimates differ widely, that is valuable information: someone knows or sees something the others don't. Averaging or setting aside the “extremes” throws that knowledge away and leaves the group with a false sense of agreement.
Five colleagues estimate how many people will come to an open day: 80, 90, 100, 110 and 400. The manager says: “400 is way off, let's plan for a hundred.” Nobody asks why their colleague wrote 400. She knew that a regional radio station would cover the event this year and that schools from the area were coming.
How to spot it
- The estimates are far apart, yet the average is calculated straight away.
- The extreme values are dismissed as “way off” without anyone asking why.
- The person with the different estimate never got to speak.
- The result is presented as consensus, though there wasn't any.
How to check and what to do
- Ask whoever gave the lowest and the highest estimates to explain their reasoning.
- Write down the reasons, not just the numbers – and find out what information the others were missing.
- After the discussion, let everyone estimate again; only the second round is worth averaging.
- Put a range in the plan (for example 80–400) and prepare for both ends.
Full explanation and exercises:
thinkcritical.eu/en/manipulation-atlas/ignored-spread