The face-specific proportion congruency effect: social stimuli as contextual cues

We follow the design described in order to obtain a comparable pattern of results with previous research (Cañadas et al. 2013). However, an alternative design was also tested showing similar conclusions. Specifically, we ran an ANOVA 2 Group congruency (High vs. Low proportion of congruency associated with each category) × 2 Individual Congruency (High vs. Low proportion of congruency associated with each face) × 2 Trial congruency (Congruent vs. Incongruent) × 2 Instructions (Categorization vs. Individualization). Results showed that participants were faster on congruent trials (vs. incongruent, F (1, 38) = 10.45, p < .01; M = 767, SE = 11 vs. M = 777, SE = 11 ms, respectively) and on faces associated mainly to congruent than incongruent displays, although this was a just a tendency (F (1, 38) = 3.67, p = .06; M = 769, SE = 11 vs. M = 776; SE = 11.03 ms). The analysis also showed as significant the Individual congruency x Trial congruency interaction, F (1, 38) = 6.86, p = .01, indicating that the proportion of congruency individually associated to each face modulated the congruency effect. When responding to highly congruent faces, participants were faster in congruent than incongruent trials (M = 760, SE = 10 vs. M = 778; SE = 13 ms). However, this effect almost disappeared and was not significant when responding in the context of a face individually associated to a low proportion congruency (M = 775, SE = 11 vs. M = 776; SE = 10 ms). This effect was not modulated by group proportion congruency or the instructions received (Fs < 1).