Tversky and Kahneman: people estimate quantities by starting from an initial value (anchor) and adjusting insufficiently.

Asked "Is the population of Turkey more or less than 35 million?" then "What IS the population?" people give lower estimates than those asked about 100 million. The arbitrary anchor contaminates the estimate.

The mechanism is still debated (insufficient adjustment, selective accessibility, or both), but the effect is extremely robust. Doctors adjust dosages relative to the first number they see. Negotiators benefit from making the first offer.

See: [[loss-aversion]], [[@alice/the-collector-fallacy]]