Abstract
Recent research has suggested that when people are in a positive mood they show reduced attentional selectivity and broadened attentional filters. Rowe, Hirsh, and Anderson (2007) investigated this effect by inducing happy and sad moods with happy or sad music while participants were completing attention tasks. Participants completing a flanker interference task were more influenced by response-incompatible flanker letters when they were in a positive mood. This effect occurred even as flanker eccentricity increased, suggesting positive mood broadened attentional scope resulting in impaired selective attention to the target. The current study extends these findings by manipulating mood with film clips played before, not during, the attention task, and by inducing an anxious rather than sad mood. Sixty-nine undergraduate participants completed a flanker task in positive and anxious moods, with the order of the film clips counterbalanced between participants. Mood ratings suggested that the film clips effectively induced happy and anxious moods that lasted through the subsequent flanker task. Participants showed a greater flanker interference effect in the positive mood condition than in the anxious mood condition. Planned contrasts revealed that participants responded more quickly to the targets with response-congruent flankers in the positive mood than in the anxious mood. However, this effect only occurred when the flankers were one letter width away from the target and not when flankers were two letter widths from the target. Thus, the effect of positive mood on visual attention was restricted to facilitating responses to trials with response-congruent flankers only within a limited spatial extent. This study suggests that the broadening of participants' attentional filters during positive mood states causes greater flanker interference because participants are more able to use the response-congruent information when in a positive mood, and not because they are unduly influenced by the response-incongruent flanking information.