Abstract
United States political affiliations run deep in modern times. Recent work has shown that even color preferences modulate on election day through a strengthening of Democrats' preference for blue, and Republicans' for red (Schloss & Palmer, 2014). These results were obtained through overt reporting of color preference and political affiliation and, therefore, may be subject to explicit attitudes and biases. Here, we measured whether color preferences reflect more implicit biases using a flanker task with irrelevant colors. We tested 451 subjects through Amazon's Mechanical Turk: approximately half in early July of 2016, and the other half on election day (November 8, 2016). Five alphanumeric characters appeared horizontally at display center. The middle letter always contained either an 'S' or 'H' target rendered in white, and subjects ignored the four flanking letters. Subjects identified the target via computer keyboard. We adopted a 2 x 3 factorial design with two flanker colors (red, blue) and three target-flanker shape compatibilities (compatible, neutral, incompatible), randomized within blocks. After the flanker task, subjects were queried about color preference and political party affiliation. On election day (versus non-election days), we found increased preference for Blue amongst Democrats and decreased preference for Blue amongst Republicans, while Red preferences were unchanged, consistent with previous findings. Interestingly, Independents showed increased Red preference and no change in Blue preference. On election day (versus baseline), both Democrats and Republicans revealed no change in flanker compatibility effects (incompatible – compatible RTs) for flankers rendered in their respective party colors. However, Democrats were significantly more affected by Red flankers (and Republicans by Blue flankers) on election day. Independents displayed increased flanker effects for both flanker colors on election day. Together, these results suggest that effects of political affiliation on color preference are implicit in nature and may dynamically alter fundamental perceptual and cognitive processes.
Meeting abstract presented at VSS 2017