The change-of-standard effect: distorted standards and adjusted impressions.

Previous accounts of the memory distortion known as the change-of-standard effect hypothesize that participants form a relative impression of a target at encoding and later use that impression with the average of all items to recall the target (Higgins & Lurie, 1983). In three experiments, we investigated the standard and the integration of the standard with the relative impression. Experiments 1 and 2 show that participants' subjective average at recall is distorted toward recent stimuli: It is computed when required and is therefore affected by the items' accessibility at that time. Furthermore, the impression's influence on recall is relatively small when the context changes between encoding and decoding. Experiment 3 shows that this change in the impression's influence occurs only when the participant integrates information across sessions, suggesting that such tasks make participants aware of the changed context and cause them to adjust the use of their impression in recalling the target.