Anchor research: evidence for an increase in the perceived similarity of stimuli obtained with a speeded-response paradigm.

A set of tiny (T) and huge (H) anchor circles and a set of small (S) and large (L) intermediate circles were used. Trials with TT, HH, TH, and HT pairs were immediately and unexpectedly followed by single trials with SS, LL, SL, or LS pairs. In addition, the anchor and intermediate sets were interchanged. Subjects took longer to decide that unexpected SL and LS pairs were different, and they made no errors to unexpected SS and LL pairs. In contrast, subjects took less time to decide that unexpected SS and LL pairs were the same, and they frequently made errors to unexpected SL and LS pairs. Nonetheless, subjects' responses to the unexpected pairs of anchor circles were not similarly affected. Decision-criterion explanations can probably be discounted. Instead, the findings accord with earlier research suggesting that memory involving anchor stimuli increases the perceived similarity of intermediate and/or relatively similar stimuli. However, the present findings strongly suggest that memory altered the quality of sensory information at a perceptual level. This memorial influence seemingly occurred within about 400 msec. In one study, a set of large and huge circles replaced the small and large set, and similar results were found. This study relates to theories of anchor effects.