Abstract
Color has often been used as a coding device to attract attention or to make stimuli conspicuous in search tasks. Some early studies suggested that color may be more effective than other visual feature dimensions in this regard. Recent models of visual search suggest that observers can attend to signals from several stimuli and that signals in different feature coding mechanisms may be combined linearly in mechanisms under attentional control to facilitate search for a known target. We have conducted studies of an observers ability to combine signals in different cardinal color mechanisms in search tasks. Preliminary experiments were consistent with the assumption that the cardinal directions represent independent color mechanisms under the observation conditions used. Initial experiments using a yes-no task and briefly flashed displays supported the idea that signals in different cardinal mechanisms could be combined to facilitate search. Subsequent experiments using a two-alternative forced choice procedure were aimed at determining whether signals in different cardinal color mechanisms were linearly summed. Results do not support the linear summation hypothesis. Results are consistent with the hypothesis that signals in a cardinal mechanism that cannot be used to discriminate the target from possible distractors, can be used to select a subset of potential target stimuli for attention. Decisions about the target are then based only on signals generated by the selected stimuli in a cardinal mechanism that can be used to discriminate the target from possible distractors. Further studies show that varying distractor signals that provide no useful information can impede the detection of a target that differs from distractors along a different cardinal axis, suggesting that the variation may interfere with attending to signals from several stimuli when necessary.