Several studies have shown that people can selectively attend to stimulus c
olour. e.g., in visual search. and that preknowledge of a target colour can
improve response speed/accuracy, The purpose was to use a form-identificat
ion task to determine whether valid colour precues can produce benefits and
invalid cues costs. The subject had to identify the orientation of a "T"-s
haped element in a ring of randomly-oriented -"L"-s when either two or four
or the elements were differently coloured. Contrary to Moore and Egeth's (
1998) recent findings, colour-based attention did affect performance under
data-limited conditions: Colour cues produced benefits when processing load
was high; when the load was reduced, they incurred only costs. Surprisingl
y, a valid colour cue succeeded in improving performance in the high-load c
ondition even when its validity was reduced to the chance level. Overall. t
he results suggest that knowledge of a target colour does not facilitate th
e processing of the target, but makes it possible to prioritize it.