engleski

Conflict and facilitation in visual cognition: A dual-process approach

Modern dual-processing models propose that multiple Type 1 processes run simultaneously. When two or more processes generate different responses, conflict may be detected and, if so, needs to be resolved. We aimed to investigate the effects of conflict between highly automated visual cognition processes, namely object differentiation (counting) and grouping. In a within-subject design, participants (N = 55) were shown images of ungrouped (equidistant) or grouped circles after which a number cue was presented. Their task was to decide whether the cued number was same as the number of circles. The cue could be the true number of circles, the number of groups (when the circles were grouped) or a filler (+/- 1 of the number of circles). When the circles were grouped the true number of circles and the number of groups were conflicting results from the two processes (differentiation and grouping). We expected that, when circles were grouped, response times would be longer when rejecting the group cue than rejecting the filler cue and that both would be slower than accepting the true cue. Further, accepting true cues was expected to be slower in the grouped compared to the ungrouped condition, and that rejecting filler cues would be faster in the grouped when compared to the ungrouped condition. A one-way ANOVA partially confirmed the first hypothesis (F(2, 108) = 11.88, p < .01, ηp2 = .18). Participants were slower to reject group and filler cues than to accept true cues, but there was no difference between the first two. A 2(grouping) by 2(true vs. filler) ANOVA showed participants were more accurate when accepting true cues in the ungrouped condition while being more accurate when rejecting fillers in the grouped condition (interaction effect - F(1, 54) = 18.93, p < .01, ηp2 = .26). Response times revealed that participants were equally fast when accepting true cues regardless of grouping, but were significantly slower to reject filler cues when stimuli were ungrouped (interaction effect - F(1, 54) = 35.34, p < .01, ηp2 = .40). We can conclude that manipulating different visual cognition processes to produce conflict is possible, but it produces a slightly different effect compared to reasoning research. Of particular note is the fact that grouping does not have a large impact on accepting true cues but does facilitate rejecting fillers. The results show parallelism of automated Type 1 processes in visual cognition discussed within a dual-process framework.

Gestalt grouping ; visual cognition ; counting ; dual-process theory

nije evidentirano