5.3.4.2. Results and discussion

The trimming procedure for correct RTs was the same as in Experiment 1. Scores were analyzed in a three-way ANOVA. The main effect of Stroop condition was significant, F (1, 41) = 14.59, p < .001, with mean RTs greater in the incongruent condition (618 ms) than in the congruent condition (603 ms). The expected two-way interaction between emotional face and emotional induction was significant, F (4, 164) = 2.67, p < .05 (see Figure 12). There was also a significant two-way interaction between Stroop condition and emotional face, F (2, 82) = 6.60; p < .01; the Stroop effect was greater with the negative and positive emotional valence conditions than that with the neutral face. No other significant effects were found.

Analysis of the questionnaire data revealed that 52% of the participants reported that there were three types of faces, 26% reported two types only, 10% reported one type of face, and the remainder did not respond. Sixty-six percent of the participants evaluated the pictures as generally positive rather than negative, and 60% of the participants noticed the positive pictures first in a free-recall test of all pictures. All participants have described correctly at least three pictures corresponding to each category of emotional induction.

The processing style exhibited toward emotional stimuli was modified across emotional events presented in the pictures (Figure 12). After viewing the positive pictures, participants’ colour-naming with negative face priming speeded up, as in Experiments 1 and 3. These results suggest that negative participants processed task-relevant information more efficiently with negative distractor by allocating more attention to target processing after viewing the positive pictures. The allocation pattern of attention across the emotional valence of face in Experiments 1 and 3 was more similar to that resulting from an emotionally positive induced state (rather than from a neutral state). One possible explanation is concerned in the fact that the welcoming laboratory atmosphere experienced by the volunteer participants could induce some positive emotions in Experiments 1 and 3. Nevertheless, participants induced by positive pictures were more likely to use avoidant processing in response to negative information. In contrast, participants exhibited slower colour-naming latencies with the negative face after viewing the negative pictures. It seems more difficult to ignore negative information after a negative emotional state has been induced; this particular sequence of events, therefore, may be capturing more attentional resources and producing greater target processing interference. This pattern is similar to the vigilant processing mode explained by Mathews & MacLeod (2002) in which attention is readily captured by threatening cues. Finally, this provides support for the idea that emotional contexts can bias attentional allocation, at least in a transitory manner.