Abstract

Dimension set consists of stimuli at mutually exclusive levels on attribute dimension (e.g., orientation: '(' vs. ')' ). Using dimension sets, previous studies have obtained positive repetition effects (PRE) that detection rate was higher when a target was flanked by a distractor of the same identity (i.e., repeated) than by different one (Kim & Kwak, 1990). It was suggested that PRE was caused by the efficient processing of parallelism emerging from the repeated displays (Kwak, Kim, & Park, 1993). On the contrary, in the alternative displays that a target is presented with a distractor of different identity, there emerges a good form having symmetry / closure, which can be easily detected to contribute to negative repetition effect. If a variety of stimulus properties are confounded in dimension set and they behave incongruently, it is not likely to observe consistent pattern of repetition effects. This study tried to test the effect of parallelism (Exp. 2) and good form (Exp. 3) separately. The results showed that both of parallelism and good form were not perceptually salient under tachistoscopic situation. The problems of global configuration and task were discussed in relation to dimension set and repetition effects.