Association as a cause of dating bias.

Two experiments investigated a dating bias in which people tend to estimate recent dates too remotely and remote dates too recently. Experiment 1 examined and upheld a prediction of the hypothesis that bias arises because events whose time of occurrence is unknown are dated through associated events for which some time information is available. Experiment 2 required some subjects to date prototype events while others dated specific events. Prototype events were dated more recently, and the errors in dating specific events were related to differences in the way the prototype events were dated. Both sets of results were predicted by the association hypothesis, according to which events whose dates are well known are dated with reference either to specific associated events or to prototype events.