US tab



Two-hundred and sixty-one students took part in this study. Sixty-one percent of participants were female and 39% of participants were male. The participant ethnicity was comprised of 57% Caucasians, 19% Hispanics/ Latinos, 11% African Americans, 9% Asian/Island Pacifiers, 4% other ethnicity, and less than 1% Native Americans/American Indians. All participants were over the age of 18. Participants were recruited from University of Central Florida through the Psychology Department's SONA System, a cloud-based research and participant recruitment system. Participants received SONA credit (course credits) for participating in this study.

Results were analyzed using SPSS version 24. The analyses began with multiple regression models and used backward regression: non-significant variables were systematically eliminated until the most efficient model was obtained. Variables were eliminated as long as the resulting F-ratio for the overall model continued to increase. Modeling stopped when the overall F-ratio no longer increased. Correlation and multiple regression analyses examined the relationship between hostile sexism and the potential predictors and benevolent sexism and the potential predictors.

The results of the regression analysis for hostile sexism indicate that 10 predictors (harm, ingroup, fairness, authority, purity, agreeableness, compassion, spiritualty, altruism, and openness) produced a significant model (F(10, 257) =14.21, p < .001, R2 = .356,. When hostile sexism was predicted, it found that the harm dimension, ingroup dimension, authority dimension, agreeableness, and spiritualty were significant predictors. The MFT fairness dimension, MFT purity dimension, compassion, altruism, and openness were not significant predictors of hostile sexism.

I found that agreeableness (β = -0.33, p < .01) and MFT harm (β = -0.21, p < .01) had a significant negative association with hostile sexism, such that the participants who were more agreeable expressed less hostile sexism.

Table 1. Results of the Multiple Regression Analyses

Only Christian, agnostic, and atheist participants were included in the results because the size of other groups (Jewish, Muslim, Buddhist, Hindu, and other) were too small. An analysis of variance showed that the effect of religious identification on benevolent sexism was significant (F(2, 205) = 20.83, p < .01). Likewise, an analysis of variance showed that the effect of religious identification (as measured by the demographic question) on hostile sexism was also significant (F(2, 205) = 10.56, p < .01). Participants who identify as Christian have a higher level of hostile sexism (M= 2.78) than participants who identify as agnostic (M= 2.31), participants who identified as atheists (M= 2.13), and participants who identify as another religion not listed (M= 2.53). Additionally, a main effect of Christian identification with MFT ingroup (F(2, 205) = 7.45, p < .01), a main effect of Christian identification with MFT authority (F(2, 205) = 5.66, p < .01), and a main effect of Christian identification with MFT purity (F(2, 205) = 21.11, p < .01). Religious identification is related a higher rating of sexist attitudes and binding MFT dimensions.

Bivariate correlations were computed among all pairs of variables. I found a significant positive correlation between hostile sexism and spirituality/religiousness, between hostile sexism and MFT ingroup dimension, between hostile sexism and MFT purity dimension, and between hostile sexism and benevolent sexism. I also found significant negative correlations between hostile sexism and altruism, between hostile sexism and compassion, between hostile sexism and MFT harm dimension, between hostile sexism and MFT fairness dimension, between hostile sexism and openness, and between hostile sexism and agreeableness. Refer to Table 2 in Appendix A for a full correlation matrix.