Anchoring and Adjustment is a term used to refer to the way people form beliefs as well as usually referencing the common errors in this process as documented by Kahnemen and Tversky, as well as being one of the primary subjects in Philip Tetlock's Superforecasting.
Anchoring refers to the initial belief about some value. People often incorrectly anchor on
- Examples that came first
- Examples that are more easily remembered
- Examples used by high status people
- Examples related to current emotional state
- Examples that support the desired conclusion (related to writing the bottom line first)
And others. In contrast, correct anchoring is done via reference class forecasting to establish a base rate for the phenomenon.
Adjustment is the process by which the initial anchor value is modified to account for new data. People adjust incorrectly when they
- Ignore new evidence that conflicts with older evidence
- Make only very large or only very tiny adjustments
- Double count evidence (see: Consilience)
- Fail to update on mundane things like the passage of time as the date of a prediction draws closer