Abstract
In the composite-face effect subjects attempt to recognize one face-half when it is aligned or misaligned with the other half. In same/different experiments, subjects are less likely to perceive that one half is the same in the aligned than the misaligned condition, which others argue represents interference from the whole facial configuration in aligned stimuli. However, it is unclear whether this is due to reduced discriminability or a criterion shift. Furthermore, the contribution of focusing attention on one face-half is unknown. We had 18 healthy subjects perform two composite face tests, each with one block of aligned faces and another of misaligned faces. Each trial consisted of a composite face viewed for 200 ms, followed by a second composite face shown at 50% larger scale, also for 200 ms. In the first test, subjects indicated if one half of the face was the same or different in the two images, while disregarding the other face-half. Half of the subjects responded to the upper face-half and half to the lower face-half. In the second test, subjects indicated if EITHER the top or bottom half of the face was the same. The same composite faces were used in both tests, with order counterbalanced. We calculated hit rate, false alarm rate, d' and c' (criterion bias). Face-alignment had a significant effect on c' but not on d', an effect that derived mainly from the condition of attending to one face-half. When subjects attended to both halves, hit rates decreased and false alarms increased, resulting in a decrease in d'; also, in this condition alignment did not have an effect on any variable. We conclude that the composite effect is due to a shift in criterion bias rather than discriminability, and that focused attention on one face-half is critical.
This work was supported by operating grants from the CIHR (MOP-77615) and the NIMH (1R01 MH069898). JJSB was supported by a Canada Research Chair and a Senior Scholarship from the Michael Smith Foundation for Health Research.