Experiment 4 Priming by Contrast Inverted Distractor Faces

Prior research shows that contrast-inverted faces impair both perceptual discrimination and explicit recognition (Galper 1970; Phillips 1972; Luria and Strauss 1978; Hayes et al. 1986; Hayes 1988; Kemp et al. 1990; Johnston et al. 1992; Bruce and Langton 1994). Recognition deficits cannot be explained by any two-dimensional descriptions because negatives do not differ from positives in terms of zero-crossings and the spatial location of features.

Studies using contrast-inverted faces generally elicit explicit responses from observers. There

Figure 11.8

Observers viewed a row of five faces that consisted of contrast-inverted distractors and positive-contrast targets.

Figure 11.8

Observers viewed a row of five faces that consisted of contrast-inverted distractors and positive-contrast targets.

is now substantial evidence that conscious and implicit processes may be independent (Jacoby 1991; Schacter et al. 1993; Treisman and DeSchepper 1996; Tulving and Schacter 1990). The processing of contrast-inverted faces may be similar in that explicit responses may not be adequate predictors of underlying processing mechanisms. One hint at the validity of this point of view comes from physiological recordings in the superior temporal sulcus of macaque monkeys. Little decrement is measured in the responses of cells as a function of contrast inversion of face stimuli (Rolls and Baylis 1986). One may now ask whether the representations established by the contrast-inverted distractors provide any priming of their positive-contrast target counterparts.

Method

Contrast-inverted versions of faces were constructed. Observers viewed a row of five faces that consisted of negative-contrast distractors and positive-contrast targets (figure 11.8). Eighteen different observers participated.

Results and Discussion

Observers were significantly faster (15.4 msec) on related trials compared with unrelated trials: ¿(17) = 2.41, p < .025. They were also significantly more accurate on the related trials, committing 2.8 percent errors, compared with 4.4 percent on unrelated trials: t(17) = —2.41, p < .025. Contrast-inverted faces did not com pete for response selection of positive-contrast targets, as predicted by the behavioral studies of recognition; rather, in consonance with physiological recordings, they facilitated processing of their positive-contrast counterpart in the role of a target.

Was this article helpful?

0 0

Post a comment