By December 2020, 29 countries (or jurisdictions within them) had legalized same-sex marriage. In Taiwan, the issue of same-sex marriage was not taken seriously in the past, but it became one of the most controversial issues in recent years. In 2017, the chief justice of Taiwan ruled that prohibiting same-sex marriage was unconstitutional, and this ruling required the Legislative Yuan to establish laws allowing same-sex marriage by 2019. However, the result of a referendum in 2018 opposed the recognition of same-sex marriage. This result reveals the tension between proponents and opponents of same-sex marriage. The current study investigated the effect of religion and support for traditional family values on the relationship between implicit and explicit attitudes toward homosexuality and opinion on same-sex marriage.

Some of the most influential organizations opposing same-sex marriage in Taiwan have religious backgrounds, especially Catholicism and other Christian denominations. Although such organizations do not present their religion as the main reason for opposing same-sex marriage; they claim to want to protect traditional family values (fulfilling familial obligations by having and raising children) and monogamy (in contrast with promiscuity and sexual openness). More importantly, religious individuals have strong implicit and explicit attitudes and may behave accordingly. The current study examined whether religious beliefs moderate the relationship of implicit and explicit attitudes with opinions on same-sex marriage. Specifically, the relationship between individual attitudes (both implicit and explicit) and opinion on same-sex marriage legislation in Taiwan may be enhanced by religious beliefs that do not support same-sex relationships (such as Christianity and Islam). In addition, this moderation may be mediated by the endorsement of traditional family values. Therefore, this study evaluated a moderated mediation model of the effects of religious beliefs and endorsement of family values on the relationships between implicit and explicit attitudes and opinions on same-sex marriage.

For Study 1, 2599 Taiwanese participants (69% female; mean age = 24.09 years, SD = 5.12) were recruited from an implicit association test (IAT) database. After providing their background information, including their gender, age, and religion, the participants were measured their implicit and explicit attitudes (two items, r = .27) toward homosexuality and whether they support same-sex marriage (only one item). The results revealed that religious beliefs moderated the relationships of both implicit (b = 0.16, p = .01) and explicit attitudes (b = 0.07, p < .001) toward homosexuality with opinion on same-sex marriage. Among individuals following religions (e.g., Christianity or Islam) opposing same-sex relationships, implicit attitude toward homosexuality was more strongly associated with opinion on same-sex marriage (b = 0.21, p < .001), whereas, among those without religious beliefs opposing same-sex relationships, implicit attitude toward homosexuality was weaklier associated with opinion on same-sex marriage (b = 0.05, p = .01). Moreover, religious beliefs opposing same-sex relationships increased the strength of the relationship between explicit attitude toward homosexuality and opinions on same-sex marriage (b = 0.14, p <.001), whereas, for people with religious beliefs not opposing same-sex relationships, explicit attitude toward homosexuality had a lower association with opinion on same-sex marriage (b = 0.05, p = .01). In summary, the findings of Study 1 support our hypothesis that religious beliefs opposing same-sex relationships increase the correlation between both implicit and explicit attitudes toward homosexuality and opinions on same-sex marriage.

Study 2 further investigated the results of Study 1 by using comprehensive and validated measurements for both the IAT and survey scales. In addition, the IAT was modified: The original questions about abstract concepts (good vs. bad) were replaced with ones on sexual openness (promiscuity vs. monogamy). For Study 2, 192 participants (62% female; mean age = 23.36 years, SD = 3.31) were recruited. After providing their background information, including their gender, age, and religion, the participants were asked about their implicit and explicit attitudes toward homosexuality (five items, Cronbach’s α = .84), traditional family values (having and raising children, parenting with family members; 22 items, Cronbach’s α = .87), and their opinions on same-sex marriage (11 items, Cronbach’s α = .93).

The results of Study 2 revealed that religious beliefs moderated the relationship between opinions on same-sex marriage and implicit attitudes toward homosexuality and opinions on same-sex marriage (b = −1.24, p = .02) but not explicit attitudes (b = 0.02, p = .56). Simple slope analysis revealed that, for those with religious beliefs opposing same-sex relationships, promiscuity bias increased the correlation between implicit attitudes and opinion on same-sex marriage (b = 2.57, p < .01), whereas for those without religious beliefs opposing same-sex relationships, promiscuity bias was not related to the relationship between their attitudes toward homosexuality and opinion on same-sex marriage (b = 0.29, p = .06). Furthermore, a moderated mediation analysis was conducted in which belief in traditional family values was the mediator of the effect of religion on the relationship between implicit and explicit attitudes and same-sex marriage stance. The results of 5000-iteration bootstrapping indicated moderated mediation (b = −0.24, 95% CI, −0.53, −0.08). Follow-up analysis revealed that for those with religious beliefs opposing same-sex relationships and support for traditional family values, implicit attitudes continued to predict their opinions on same-sex marriage (b = −0.24, 95% CI, −0.53, −0.08). However, for those with religious beliefs not opposing same-sex relationships, this indirect path via traditional family values did not represent a mediation effect (b = 0.04, 95% CI, −0.01, 0.12). In summary, the findings of Study 2 support our hypothesis of the moderation effect of religion on the relationship between implicit attitudes toward homosexuality and opinions on same-sex marriage in the presence of promiscuity bias. Specifically, religious beliefs opposing same-sex marriage increased the strength of the relationship between implicit attitude and opinion on same-sex marriage. Furthermore, moderated mediation analysis demonstrated that traditional family values mediated the relationship between implicit attitude and same-sex marriage that was moderated by religious beliefs. These findings provide insights that may build a bridge between proponents and opponents of same-sex marriage and help them to communicate, thereby reducing tension in Taiwanese society.