Downloads & Free Reading Options - Results

Every day, we are confronted with numerous sources of information that compete for attention. This often poses an attentional challenge, since only a portion of these sources may be compatible with our goal. Often, paying attention to some sources of information has detrimental effects on performance, i.e., it generates interference. Hallmark of attentional control is the ability to adjust control settings in reaction to such events, to overcome future interference. We have recently shown that, in multi-dimensional environments, attentional control is achieved by modulating the processing of task-irrelevant information. Specifically, control adaptation is highly dimension-specific rather than dimension-general because it selectively suppresses or enhances task-irrelevant dimensions based on their history of compatibility with the instructed task. However, our prior findings also raised the question of whether people may adopt dimension-general control settings in specific circumstances. For example, we surmised that transfer of control across non-task-relevant dimensions may happen when their features systematically co-occur (i.e., statistically bound). We reasoned that such circumstances would strongly reduce the benefit of assigning dimension-specific control based on individual dimensions’ conflict history. In other words, it may be possible that co-occurring distractors may be clustered together during control adaptation. Therefore, with this study, we introduce a statistical relationship between non-task-relevant information, and measure cross-dimension transfer of attentional control. This preregistered study aims to examine if humans can learn statistical relationships of non-task-relevant task features, and whether this can lead to generalization of adjustments in attentional control towards them.