HIDDEN-COVARIATION DETECTION AND IMAGERY ABILITY

In studies on hidden-covariation detection (HCD), it is demonstrated that humans may be influenced by covariations between stimuli, without acquiring any conscious or ''explicit'' knowledge about the crucial relations (e.g. Lewicki, Hill, & Sasaki, 1989). Even though not considered to be of any consequence, in the majority of studies the crucial X-Y covariation is embedded in a propositional structure of the type ''X is a Y''. In the experiments reported here, however, the possibility of HCD was studied in a situation implying mere spatio-temporal co-occurrence between two nonverbal stimuli X and Y Also, we investigated whether high imagery ability facilitates performance on an evaluative HCD task. While performing an attention-consuming distraction task, good and poor imagers were first exposed to a covariation between the line thickness (thick/thin) of a geometric figure and the valence (positive/negative) of a subsequently presented slide. In the test phase, subjects were required to express their subjective evaluation of a series of stimuli which differed in line thickness. In a series of three experiments, it was demonstrated that (a) even though none of the subjects was evidenced to have acquired any explicit knowledge about the X-Y covariation, (b) poor imagers' evaluations of the test stimuli were clearly influenced by the hidden acquisition covariation, whereas contrary to expectations, good imagers only evidenced a non-significant tendency for HCD; (c) moreover, the HCD effect in poor imagers manifested itself in the form of an evaluative contrast, mirroring the acquisition contingency. The data further suggested that (d) the evaluative contrast phenomenon is a performance rather than a learning issue. Finally, it is argued that their may exist a fundamental conceptual similarity between the HCD paradigm used in this study and the standard Pavlovian or Evaluative Conditioning paradigm.