Abstract
Previous research has shown that color preference is affected by the environment and often changes due to social settings (sports teams, college allegiances, etc; cf. Schloss & Palmer, 2017). In this experiment, we utilized the United States two-party political system as a social backdrop upon which to measure the effect of color-party association on both explicit and implicit biases. We tested 451 subjects through Amazon’s Mechanical Turk; approximately half during summer of 2016, and the other half on the 2016 presidential election day (November 8). We implicitly measured the effect of color on attentional filtering of distractors by having participants complete an Eriksen flanker task with a white target letter and red or blue flankers (blocked by color). They then explicitly rated their color preference on a 16-color array, which included the same red and blue colors from the flanker task. Finally, subjects completed a political demographic survey, where they indicated their political affiliation and voting history. Color preference data showed that, on election day, Republicans experienced a decrease in their preference for blue but were more distracted by red flankers; while Democrats showed no change in color preference on election day but were more distracted by blue flankers. Due to this dichotomy between implicit flanker task and explicit color preference results, a random forest analysis was used to determine the relative importance of factors in the determination of political party affiliation. The blue flanker effect had the largest impact on accurately determining political affiliation. Second most important was the red flanker effect and third was the blue color rating. The red color rating actually decreased the accuracy of the random forest model. These results suggest that political affiliation dynamically changes fundamental perceptual & cognitive processes that are best captured by measures of implicit biases rather than explicit surveys.