Case file
Conservatism Bias
- Filed under
- Too Much Information
- Also recorded as
- Conservatism in belief revision
The charge
We revise beliefs too slowly when new evidence arrives. Old views keep too much weight after the facts have moved.
How it operates
Priors, identity, and coherence make belief change uncomfortable, so updates are smaller than the evidence warrants. People cling to yesterday's model after today's data arrives.
Logged incidents
- Incident 01
A go-to-market plan remains mostly unchanged even after channel data shows the original acquisition thesis is failing.
- Incident 02
A manager keeps rating an employee as average because of an early impression despite a year of strong performance.
- Incident 03
An investor clings to a bullish thesis after repeated earnings disappointments and deteriorating fundamentals.
What to watch for
Ask yourself: 'What evidence would make me materially change my mind, and has it already happened?'
Recommended action
Use Bayesian updating, explicit trigger points for belief revision, and scheduled forecast updates. Commit in advance to what size of evidence should force a meaningful change.
Known associates
- AnchoringAn initial number, label, or piece of information pulls later estimates toward it, even when it is arbitrary…
- Contrast EffectSomething looks better or worse depending on what it is compared with immediately before or beside it.
- Distinction BiasWhen options are compared side by side, we exaggerate small measurable differences that matter little in…
- Focusing effectWhen one salient detail is in focus, it pulls judgment toward itself and crowds out other relevant factors.
- Framing EffectEquivalent information leads to different choices depending on how it is worded or packaged.
- Money IllusionWe think in nominal money terms and ignore inflation, purchasing power, or real value.
Source of record