Abstract
Purpose. The purpose of this quantitative study was to test facial appearance voter heuristics by comparing the prediction rates of competence judgments of opposite-sex candidate pairs and the prediction rates of competence judgments of same-sex candidate pairs. Methodology. This quantitative study gathered data from an online survey completed by 207 students at Bakersfield College. Inferential and descriptive statistics were used to compare predictive rates based on competence judgments for male and female respondents as well as compare predictive rates for opposite-sex and same-sex candidate pairs. Findings. Data analysis confirmed competence judgments as a predictor of overall election results, results for opposite-sex candidate pairs, and from female respondents. Differences between prediction rates were not statistically significant. A moderate relationship between the gender of the winning candidates and the gender of the candidates chosen as more competent was found when controlling for the gender of the respondents. Conclusions. Though gender bias was not confirmed, it was not rejected either. Gender bias may exist at a statistically insignificant rate. It is important to remember that real election outcomes can be won by statistically insignificant amounts, and the amount may be a result of gender bias. Recommendations. It is recommended that future research study possible gender bias in limited-information elections further. Suggestions include examining judgments for same-sex pairs with equal numbers of male-male and female-female pairs, testing competence judgments across cultures, and simulating real elections to examine gender bias interactions with election results.