Does the attentional state determine processing dominance in to-be-ignored compound stimuli?
Previous studies (Paquet, 1992, 1994; Paquet and Merikle, 1988) have shown that identification of the local aspect of an attended compound stimulus is delayed by the presence of the global distractors within a to-be-ignored compound stimulus, whereas identification of the global aspect is unaffected by local distractors. This asymmetrical pattern of interference indicates that processing of to-be-ignored compound stimuli is characterized by global dominance. In this study, I investigated the hypothesis that these findings were the outcome of subjects adopting an attentional state which favoured processing of global information. Following a procedure developed by Ward (1982), subjects were asked to perform two consecutive identification judgements of the global or the local aspects of two successive relevant compound stimuli, each presented with a to-be-ignored compound stimulus. The first and second judgements were performed either at the samelevel (i.e., both global or both local), or required switching attention across levels (i.e., global on the first judgement and local on the second, or local on the first judgement and global on the second). The results confirmed that the attentional state (switched vs. unswitched) influences global/local processing speed on the second judgement (Ward, 1982). However, no evidence for local dominance was found when the attentional state favored processing of local information. Instead, global distractors always affected local target identification, whereas interference effects of local distractors upon global target identification were restricted to the attention switching conditions. These findings suggest that the pattern of dominance in to-be-ignored stimuli is not determined by the attentional state, but may be related to fundamental differences in the strength of global and local information channels.