The Contact Hypothesis and Millennial Evangelical Protestants’ Attitudes toward Same-Sex Families

Millennials

Millennials are the youngest generation of adults and are often compared with Generation X (born 1965 – 1979), Baby Boomers (born 1946 – 1964), and the Silent Generation (born before 1946). It has been found that Millennials have similar opinions to older generations regarding certain social issues such as abortion and gun control, yet are more liberal in their attitudes towards interracial marriage, marijuana legalization, and same-sex marriage (Pew Research Center 2014b).

Scholars attempting to explain the swift change in attitudes toward same-sex relationships often credit favorable attitudes among Millennials as the explanatory factor. They note that, as the younger Millennial cohort replaces the older cohorts, the population is overall more likely to support LGBQ rights (Baunauch 2011). A 2013 survey found that 70 percent of Millennials support same-sex marriage; at that time, Millenials comprised 27 percent of the adult population. This finding marked a considerable increase from 2003, when only 51 percent of Millennials supported same-sex marriage and they comprised nine percent of the population (Dimock, Doherty, and Kiley 2013). N. Silver’s (2013) analysis of eight national opinion polls notes that the steady increase in support for same-sex marriage is consistent with the idea that younger Americans are replacing older ones in the electorate. It is estimated that about half of the increase in support is attributed to this generational turnover (Silver 2013). D.M. Baunach (2011) explicitly investigated whether cohort replacement explained attitude change towards gay marriage in the U.S. by analyzing data from multiple years of the General Social Survey. Cohort succession was determined to be responsible for one-third of the change (Baunauch 2011).

M. Dimock, C. Doherty, and J. Kiley (2013) note that attitude change regarding same-sex relationships is occurring among older generations as well. The Silent Generation’s support for same-sex marriage increased from 17 percent in 2003 to 31 percent in 2013 (14% increase). Baby Boomers exhibited the lowest portion of growth from 33 percent in 2003 to 38 percent in 2013, and Generation X moved from 41 to 49 percent (i.e., increases of 5% and 8% respectively). Yet the greatest increase in support for same-sex marriage occurred among the Millennial generation, whose jump from 51 to 70 percent was a 19 point increase (Dimock, Doherty, and Kiley 2013). The combination of these attitude changes within each generation indicates that not all of the growth in support for same-sex marriage can be explained by cohort replacement. In fact, Baunach’s work identified attitudinal change as the explanatory factor accounting for two-thirds of the increase in support for gay marriage (Baunauch 2011).

Although much research on same-sex relationships focuses on same-sex marriage, attitudes toward other aspects of same-sex-headed family creation have also been explored. For instance, in 2013, two-thirds of Americans agreed that gay and lesbian couples could be as good as heterosexual couples as parents (68%). Millennials were the generation most likely to agree (79%), almost 30 percentage points greater than the Silent Generation (50%; Jones, Cox, and Navarro-Rivera 2014). Almost 60 percent of Americans approved of gay and lesbian couples adopting children (58%). Support among the generational groups was consistent with previously identified trends, with 70 percent of Millennials, 58 percent of Generation Xers, 52 percent of Baby Boomers, and 42 percent of the Silent Generation favoring LGBQ persons adopting (Jones, Cox, and Navarro-Rivera 2014).

Evangelical Protestants

Evangelical Protestantism is one among many U.S. religious traditions. Evangelical Protestant denominations are distinguished as being more theologically and socially conservative than mainline Protestantism (The Association of Religion Data Archives 2010). Evangelicals are also known for greater separation from the broader culture and stricter adherence to religious traditions (Streensland, Park, Regnerus, Robinson, Wilcox, and Woodberry 2000). Evangelicals are more likely than other religious groups to identify religion as the factor that most influences their thinking about government and public affairs (Smidt 2013, 186). As of 2007, evangelical Protestantism was the largest religious tradition in the U.S., with 26.3 percent of Americans indicating affiliation (Smidt 2013, 70).

Research on the sexuality attitudes of evangelical Protestants indicates that they are the tradition least supportive of same-sex relationships (Jones, Cox, and Navarro-Rivera 2014; Rowatt, LaBouff, Johnson, Froese, and Tsang 2009). An analysis of the 2007 Pew Forum Religious Landscape Survey showed that less than one-third of evangelical Protestants (28%) agreed that “homosexuality is a way of life that should be accepted by society,” compared to 58 percent of mainline Protestants and 62 percent of Roman Catholics (Smidt 2013, 188). Results from the 2007 Baylor Religion Survey revealed that mainline Protestants, Catholics, “other” religious groups, and the religiously unaffiliated are more likely to support same-sex marriage than evangelical Protestants. Among evangelical Protestants, the question of same-sex marriage “elicits a strong, unfavorable view regardless of their views toward the cause of homosexuality,” (Whitehead 2010, 74).

The Pew Center’s more recent survey on LGBT Issues and Trends (2014) included racial distinctions in religious affiliation. Most of the religious affiliations demonstrated majority support for same-sex marriage (e.g., Hispanic Catholic 56%, White Catholic 58%, White mainline Protestant 62%, Jewish 83%). White evangelical Protestants were among the three religious affiliations in which less than half of the participants supported same-sex marriage (White evangelical Protestants 27%, Black Protestants 35%, Hispanic Protestants 46%; Jones, Cox, and Navarro-Rivera 2014). This survey included questions regarding additional aspects of same-sex-headed family creation. The results showed that, with the exception of White evangelical Protestants, the majority of religious respondents agreed that gay and lesbian couples can be as good as heterosexual couples as parents. White evangelical Protestants, along with Hispanic and Black Protestants, are also the religious traditions in which less than 50 percent support allowing gay and lesbian couples to adopt children (33%, 46%, and 40% respectively; Jones, Cox, and Navarro-Rivera 2014).

These increasingly tolerant attitudes among people in certain religious traditions towards same-sex relationships is evident in some of the trends of associated religious institutions. In July 2015, a few days after the U.S. Supreme Court legalized same-sex marriage nationwide, the Pew Research Center outlined the standings of major religious traditions (Masci and Lipka 2015). They note that the Conservative and Reformed Jewish movements had approved of same-sex marriage for a number of years. Some Mainline Protestant denominations, such as the Episcopal Church and the Presbyterian Church (U.S.A), recently sanctioned blessings or same-sex marriage ceremonies. However; most religious institutions have remained firmly against allowing same-sex marriages. These include the Roman Catholic Church, the Orthodox Jewish movement, the Southern Baptist Convention, and other evangelical Protestant denominations (Masci and Lipka 2015).

The Contact Hypothesis and Influence on Attitudes

The contact hypothesis posits that intergroup contact can increase tolerance and empathy while decreasing stereotypes, prejudice, and conflict. Since its inception in the early 1950s, researchers have found factors and situations within which contact with previously unfamiliar groups has had an impact on related attitudes (Dovidio, Glick, and Rudman 2005). Scholars have demonstrated that contact between different individuals can favorably impact intergroup relationships in regards to race (Amir 1998), age (Caspi 1984), and religion (Putnam and Campbell 2010, 526-529). For example, D.P. Green and J.S. Wong (2009) used an Outward Bound multiple week wilderness course as the setting for a randomized field experiment. They found that a month after completing the course, students in the racially diverse group expressed greater levels of racial tolerance and reduced racially aversive attitudes in comparison to the racially homogeneous control group (Green and Wong 2009).

The contact hypothesis has also been employed by researchers to explain attitudes towards LGBQ people. G.M. Herek and J.P. Capitanio (1996) found that heterosexuals who had contact with gay men or lesbian women reported more favorable attitudes towards same-sex attracted people. Additionally, heterosexuals with more intimate or multiple contacts with gay or lesbian people indicated higher rates of favorability than those with little or no contact (Herek and Capitanio 1996). G.G. Lewis’ (2011) meta-analysis of 27 nationally representative survey results revealed that respondents who knew someone gay, lesbian, or bisexual were more likely to support gay rights, a trend identified in every demographic, religious, and political group (Lewis 2011). Dimock, Doherty, and Kiley (2013) specifically sought to learn about the factors contributing to attitudinal change towards same-sex marriage. In a 2013 survey, it was determined that one-in-seven people have changed their mind to support gay marriage. When respondents who reported having changed their mind were asked why their views changed, the most common answer was contact with a person who is homosexual (32%; Dimock, Doherty, and Kiley 2013). J. Barth and J. Parry (2009) expanded on this line of research by testing whether knowing someone in a same-sex romantic relationship had greater impact on attitudes than simply knowing someone who identified as gay or lesbian. They concluded that knowing same-sex couples had a distinct impact on attitudes towards legal recognition of same-sex partnerships; however, knowing a lesbian or gay man, whether in couples or not, was a stronger predictor of support in regard to other policies (e.g., nondiscrimination laws, adoption of children by lesbians and gay men; Barth and Parry 2009).

The contact hypothesis has not been found to be effective in every heterogeneous interaction and situation. The research regarding religion and attitudes towards LGBQ relationships and families reveals the complex effects that multiple factors can have on attitudes. S.A. Skipworth, A. Garner, and B.J. Dettrey (2010) note that the effect of contact may be mitigated by strong predispositions. Their research on the relationship between knowing someone who is LGBQ and attitudes towards gay rights concluded that the contact effect varies based on the individual’s political, religious, and cultural predispositions. They found that contact with LGBQ people had a positive effect on all groups, but a limited effect among the most conservatively minded-groups (e.g., evangelical conservatives; Skipworth, Garner, and Dettrey 2010). B. Bramlett’s (2012) work demonstrated that even though contact with LGBQ people is associated with greater support for same-sex marriage, the extent may be limited or actually increase opposition for certain groups such as with White evangelical Protestants. Bramlett explains that evangelical Protestants endure in politics due to a separation from more secular society; the messages in their religious teachings may carry more weight than knowing someone who is LGBQ (Bramlett 2012). A. Garner’s (2013) research revealed that interpersonal contact with LGBQ people induces ambivalence or cross-pressure among citizens with conservative-leaning predispositions (e.g., Republicans, evangelical Christians). Evangelical Christians who had no contact with LGBQ people had a more uniform position against gay marriage. Evangelicals with contact continued to be opposed to gay marriage generally, but their responses were less uniform and more dispersed across the range of options (Garner 2013).

Another factor to consider regarding the relationship between contact and attitudes towards LGBQ relationships and families is the direction of causation. Some note that the favorable attitudes professed by those who have contact with LGBQ people may precede the contact. They theorize that LGBQ people are more likely to reveal their sexual identity to those who hold positive, accepting attitudes towards same-sex relations (Lewis 2007; Skipworth, Garner, and Dettrey 2010). Herek and Capitanio’s (1996) longitudinal research provided support for this supposition, finding that individuals with favorable attitudes toward gay men and lesbian women in Wave 1 were more likely than others to experience contact in Wave 2. Similarly, Lewis’ (2007) research concluded that there is evidence of both causal directions, since people whose characteristics make them more likely to be accepting of same-sex attracted individuals are more likely to know LGB people. However, Lewis also notes that knowing an LGB person has a noticeable impact even when controlling for other factors, including previous beliefs about same-sex attracted individuals (Lewis 2007).

According to the most recent national survey on same-sex relationships in the U.S., two-thirds of Americans report having a close friend or family member who is gay or lesbian (65%). Approximately seven in ten Millennials (71%) and almost six in ten White evangelical Protestants (58%) know someone who is LGBQ. Similar to previous trends, Millennials are the age cohort most likely to know someone who is LGBQ, whereas White evangelical Protestants are the religious tradition least likely (Jones, Cox, and Navarro-Rivera 2014).

Millennials, Evangelical Protestants, and the Contact Hypothesis

Millennials are less religious than older Americans. Almost one-fifth note that they were raised in a religion but are now unaffiliated with a particular faith. Millennials who do associate with a particular religious tradition, however, have similar levels of religious intensity to previous generations when they were young. For instance, 37 percent of Millennials consider themselves to be “strong” members of their faith. A comparable portion of Generation Xers and Baby Boomers asserted the same when they were younger (37% and 31% respectively; Pond, Smith, and Clement 2010).

Currently, 22 percent of Millennials identify evangelical Protestant churches as their religious affiliation (Pond, Smith, and Clement 2010). There is some evidence that young people are leaving the evangelical tradition, considering themselves religious but unaffiliated or not religious. C. E. Smidt’s (2013) work on American evangelicals notes that between 1964 and 2007, the evangelical Protestant tradition experienced a six percent decrease in those under the age of 35, the largest decrease among all of the affiliations examined (Smidt 2013, 81-82). Despite these changes, Smidt explains that a number of factors indicate that about one-quarter will continue to be evangelical in the future (e.g., evangelicals have larger families) (Smidt 2013, 225). Relatedly, a news release on the Public Religion Research Institute’s 2013 LGBT Issues and Trends Survey titled, “American Religious Landscape Transforming as Support for Same-Sex Marriage Dramatically Increases,” asserted that Millennials were leaving their childhood religions, specifying unfavorable church teachings about LGBT as the reason. It was also reported that a majority of Americans (51%) perceive evangelical Christians to be one of the three religious groups unfriendly to LGBT people (Public Religion Research Institute 2014). No specific information was provided about whether Millennials were leaving evangelical Protestant churches in particular.

As a sub-population of both age and religion, there is limited research providing specifics on this group’s attitudes about sexuality; however, the existing data reflects the contrasts of their age cohort and religious affiliation. The 2007 Pew Forum Religious Landscape survey found that 41 percent of Millennial evangelical Protestants agreed that “homosexuality should be accepted by society”. This is 15 percentage points larger than non-Millennials (26%) and 13 points larger than evangelical Protestants in general (28% in agreement; Smidt 2013, 206). Compared with older members of their faith, larger percentages of Millennials believe that society should accept homosexuality (Pond, Smith, and Clement 2010).

The 2014 Public Religion Research Institute survey on LGBT issues and trends included a brief narrative analysis of age and religion. The results exemplify the tension between age and religious factors, and LGBT attitudes. The authors note that although only 27 percent of White evangelical Protestants support same-sex marriage, Millennial White evangelical Protestants are twice as likely to favor it (43%) than older generations (Generation Xers 33%, Baby Boomers 22%, Silent Generation 19%; Jones, Cox, and Navarro-Rivera 2014). A majority of Millennial White evangelical Protestants (66%) believe gay and lesbian couples can be as good as parents as heterosexual couples, compared with a minority of the older generations who agree (32%). Relatedly, Millennial White evangelical Protestants are twice as likely as the Silent Generation to support allowing gay and lesbian couples to adopt (46% and 24% respectively; Jones, Cox, and Navarro-Rivera 2014).

To learn about how religious and political factors influence support for same-sex marriage, D.E. Sherkat, M. Powell-Williams, G. Maddox, and K. Mattias de Vries (2011) analyzed data from the General Social Survey. They find that religion plays a strong role, particularly within evangelical denominations. They explain that denominational ties and religious participation have a negative influence on support for LGBQ marriage rights. They comment that although some scholars have asserted that younger generations of conservative Christians are growing tolerant on sexuality issues, there continues to be a notable gap between the attitudes of younger conservative Christians and people with other or no faith commitments (Sherkat, Powell-Williams, Maddox, and Mattias de Vries 2011).

A consistent theme within the research literature is that evangelical Protestants are the religious tradition most opposed to same-sex relations, marriage, adoption, and parenting. Learning about the sexuality attitudes of the new generation of evangelical Protestant adults and how they are affected by contact with LGBQ people will provide insight into the Millennial evangelical Protestants’ support or opposition to same-sex-headed families. Conversely, the work illustrated that Millennials are the age cohort most likely to favor legal support for same-sex families. Learning about the sexuality attitudes of the youngest generation of evangelical Protestants and how their attitudes are affected by contact will also illuminate the future relationship of Millennials with the evangelical Protestant tradition.

Research Plan and Sampling

This project was part of a larger effort to learn about attitudes towards non-traditional families. The study used nonprobability sampling due to access to a sizeable group of participants attending a four-year, private, evangelical Christian college. Eighty-seven undergraduate students were recruited as volunteers from multiple sections of an introductory sociology course. Students enrolled in this general education course are more likely to be a cross-section of the student body than is it to be a concentration of sociology majors. Although a nonprobability sample of college students does not lead to results generalizable to the larger U.S. population, the conclusions will provide some understanding of the attitudes held by this specific demographic. The age range of the participants (93.0% age 18 – 22, 4% age 23 – 27; 2.3% did not respond) places them firmly in the Millennial generation. The survey instrument did not include questions regarding religious affiliation[3], rather recent demographic information collected in the college’s enrollment materials was used to understand the religious affiliation of the students. The vast majority of the college’s students indicated that they have an evangelical Protestant affiliation (74.4% evangelical Protestant, 23.8% mainline Protestant, 2.7% Roman Catholic, 0.2% other religions, 3.8% none/not stated). It was determined that these proportions would be the approximate composition of religious affiliation of students in general education courses such as the introduction to sociology class. Thus, the analysis was completed with the understanding that the findings would provide insight into the sexuality attitudes of Millennial Protestants (particularly evangelicals) with some college education.

Approximately 74 percent of the participants identified their race as White, 22 percent as non-White, and four percent did not respond. These percentages are extremely similar to the student body, which was 76 percent White, 22 percent non-White, and two percent without a response. It is also similar in racial composition to other private, four-year institutions (Aud, Fox, & Kewal Ramani 2010). Approximately 76 percent of the participants were female, 22 percent male, and two percent did not respond. This is about a 25 percent difference from the student body ratio which is 54 percent female and 46 percent male. It differs by about 18 percent from other private schools (Borzelleca 2012).

The survey questions were developed using items from the Baylor Religion Survey (The Association of Religion Data Archives 2005), the Pew Research Religion and Public Life Project (Pew Research Center 2014c), and the Values and Beliefs section of the GALLUP Poll (GALLUP). The word ‘homosexual’ was replaced with ‘same-sex’ to reflect the most recent appropriate term (Lesbian Gay Bisexual Transgender Campus Resource Center). This internet survey was administered through Qualtrics. Participants scheduled appointments with the research team and went through an informed consent process upon arrival. Once the consent was given, participants were directed to a computer in a single-person lab room where they could complete the survey, which took approximately 15 minutes to accomplish. A small number of the participants (approximately 2-3) did not complete all of the questions in the survey. Thus some of the tables reflect a sample size smaller than 87 in regards to certain questions.

Dependent Variables

The first two dependent variables were responses to an introductory question:

These items were measured on a four-point scale: 1 – Always Wrong, 2 – Almost Always Wrong, 3 – Only Wrong Sometimes, and 4 – Not Wrong at All. The other dependent variables were responses to:

These items were measured on a five-point scale: 1 – Strongly Disagree, 2 – Disagree, 3 – Neither Agree nor Disagree, 4 – Agree, and 5 – Strongly Agree.

Independent Variables

The questions for the independent variables measuring contact were located at the end of the survey with the demographic information.

These items were measured as 0 – No and 1 - Yes.

Analysis

Descriptive statistics were run to learn about the demographics and contact with LGBQ persons. Cross-tabulations were completed to provide insight into the frequency and proportion of the two independent variables in regards to the five dependent variables.

Because the dependent variables have a logical sequential order, ordered logit was initially selected as most appropriate for this analysis. The test of parallel lines, however, was significant for two variables (General Attitude toward Marriage p = .023, Same Sex Marriage Legal p = .012) calling into question the assumption of proportional odds. Thus, the dependent variables were treated as categorical, and multinomial logistic regressions were completed with the independent variables of LGBQ Family/Friend and LGBQ in Relationship as the main predictors. Race and sex were included as control variables.

The results revealed a pattern suggesting that contact with LGBQ people in a relationship had greater impact on attitudes than knowing someone who identifies at LGBQ. Since an exploratory analysis on the data revealed a non-normal distribution, the nonparametric Mann-Whitney test was used to learn about the differences between these two groups (LGBQ Family/Friends Only – 1, LGBQ in Relationship – 2).

Limitations

The nonprobability sampling method and small number of participants prevents the results from being scientifically generalizable to any larger population. As with most research in which college students are used as a convenience sample, the conclusions are used to learn about this particular group and provide insight for future research (Babbie 2005, 188-189).

Regarding religion, the participants were not asked any questions about religious affiliation, church attendance, or other measures of religiosity. However, since three quarters of the college student body recently identified as evangelical Protestant in their admissions applications, it can be assumed that the majority of research participants are or recently have been affiliated with the tradition. Thus, the conclusions made are referred to as being about evangelical Protestants although no measures of religious affiliation or participation were recorded.

Regarding sex, a disproportionately high number of participants identified as female (76%). It has been found that females may be slightly more supportive of LGBQ people and relationships than males (Jones, Cox, and Navarro-Rivera 2014; Olson, Cadge, and Harrison 2006). With this in mind, the sex variable is controlled in the multinomial regression analyses. Regarding race, the majority of the participants identified as White (74%). White Americans are more likely to support same-sex marriage than groups of color (Jones, Cox, and Navarro-Rivera 2014; Sherkat, Powell-Williams, Maddox, & Mattias de Vries 2011). Thus race is also controlled.

Regarding causal direction, it is recognized that, much like research on the relationship between LGBQ contact and attitudes towards same-sex relationships and families, this study cannot definitively establish that knowing someone who is LGBQ accounts for all of the change in an individual’s opinions and attitudes (Lewis 2007). Since researchers have found that LGBQ contact does have impact when controlling for previously established beliefs, the authors acknowledge that this is a limitation of the project.

Frequency of Contact

Almost half of the participants (48.2%) indicate having a close friend or family member who is LGBQ. One-quarter (24.7%) know someone in a same-sex relationship (Table 1). These percentages are lower than both Millennials (71%) and evangelical Protestants (58%) in the general public (Jones, Cox, and Navarro-Rivera 2014), yet demonstrate that contact is occurring.

Attitudes Regarding Whether Same-Sex Marriage is Wrong

Regarding general attitudes about same-sex marriage, attitudes of those with LGBQ family and friends are similar to those without with the exception that those without LGBQ family and friends are much more likely to consider it always wrong. This pattern is evident in the relevant percentages in Table 2 showing that 42.5 percent of those with LGBQ family and friends believe it is always wrong compared to 75 percent of those who do not have LGBQ family or friends.

Participants who know someone in a same-sex relationship have varied responses to general attitudes about same-sex marriage. A notable amount consider it always wrong (23.8%) or almost always wrong (19.0%), but almost half of these participants (47.6%) state that it is not wrong at all (Table 2). This finding is in stark contrast to those who do not know someone in a same-sex relationship, among whom 71.4 percent consider same-sex marriage always wrong and only 17.5 percent believe it is not wrong at all (Table 2). The results of the multinomial logit reveal that those who do not know someone in a same-sex relationship are over seven times more likely to consider same-sex marriage to be always wrong rather than not wrong at all in comparison to those who do know someone in a LGBQ relationship (odds ratio = 7.23, p = .023, Table 3).

Attitudes Regarding Whether Same-Sex Adoption is Wrong

Regarding general attitudes towards same-sex-headed families, participants were a bit more supportive of same-sex adoption than they were of same-sex marriage. Opinions about same-sex adoption were also more varied than those expressed about same-sex marriage. Almost one-third of participants with LGBQ family members or friends considered the adoption of a child by a gay couple wrong (17.5% always wrong + 12.5% almost always wrong), another third considered it only wrong sometimes (32.5%), and the final third considered it not wrong at all (37.5%, Table 2). Comparatively, a slightly larger proportion of people without LGBQ family members or friends considered it always wrong (31.8%) or almost always wrong (27.3%) compared to those who think it is only wrong sometimes or not wrong at all (13.6% and 27.3% respectively, Table 2).

A similar pattern was identified regarding LGBQ contact with many of the participants who know someone in a same-sex relationship indicating that the adoption of a child by a gay couple is not wrong at all (42.9%) which differs from the majority of participants who do not know someone in a same-sex relationship considering adoption of a child by a gay couple wrong (almost always wrong 25.4% + always wrong 28.6%, Table 2). The mean of 3.10 for those who know someone in a same-sex relationship compares to the mean of 2.46 for those who do not, showing that each group tends to fall on opposite sides of the issue (Table 2). This difference is not statistically significant in the multinomial logit results (Table 4).

Should Same-Sex Relationships be Legal?

The first of these findings regarding the legality of same-sex-headed families, reveals that contact seems to influence or predict attitudes. Participants with LGBQ family members or friends indicate a variety of opinions about whether romantic relationships between same-sex people should be legal, yet a large portion of this group choose to neither agree nor disagree with the statement “Same-sex relations between consenting adults should be legal,” (37.5%, Table 5). This contrasts with participants who do not have LGBQ family members or friends among whom the majority strongly disagreed (52.3%, Table 5).

A similar, yet more distinct theme surfaces among those who know someone in a same-sex relationship regarding legalizing same-sex relations. Of the participants who know someone in a same-sex relationship, the greatest portion chose to neither agree nor disagree with the statement “Same-sex relations between consenting adults should be legal,” (47.6%, Table 5). Additionally, the other participants indicating knowing someone in a same-sex relationship were more likely to agree (19.0%) or strongly agree (28.6%) than to strongly disagree with the statement (4.8%, Table 5). This pattern differs from those who do not know someone in a same-sex relationship among whom the greatest portion disagreed (42.9% strongly disagree, 19% disagree), and less than one-fifth agreed (3.2% strongly agree, 15.9% agree; Table 5).

It is important to note that no participants who know someone in a same-sex relationship disagreed with the statement, “Same-sex relations between consenting adults should be legal,” (0% disagree, Table 5). For this reason the five-point Likert scale ranging from strongly disagree to strongly agree was collapsed into a three-point scale combining all participants who disagreed (strongly disagree and disagree) into one point and combining all participants who agreed (strongly agree and agree) into another point while leaving those who neither agree nor disagree remaining as the middle point. This made it possible to analyze using multinomial logit. Table 6 details the results which reveal that people who do not know someone in a same sex relationship are almost 20 times more likely to disagree than to agree that same-sex relationships between consenting adults should be legal (OR = 19.48, p < .05). The statistically significant coefficient for the control variable of sex (OR = .176, p < .05) may suggest that male participants in this sample were slightly less likely to neither agree nor disagree rather than to agree in comparison to female participants (Table 6). This is an interesting finding that could be explored in future studies that have a more representative gender sample.

Should Same-Sex Marriage be Legal?

A slightly greater portion of participants who know someone who is LGBQ agree that same-sex marriage should be legal (strongly agree 17.5%, agree 30.0%), whereas the majority of participants who do not know someone who is LGBQ disagree that same-sex marriage should be legal (52.3% strongly disagree, 20.5% disagree, Table 5). There is a 1.16 point difference in means between these two groups with participants having LGBQ family members and friends at an average of 3.18 (slightly higher than neither agree nor disagree) and participants without LGBQ family members or friends at 2.02 (disagree, Table 5).

Similar to knowing someone who is LGBQ, participants who know someone in a same-sex relationship were much more likely to agree that same-sex marriage should be legal (23.8% strongly agree , 33.3% agree) whereas those who do not know someone in a same-sex relationship were more likely to disagree (44.4% strongly disagree, 19.0% disagree; Table 5). A 1.27 point difference in means (3.52 for those know someone in a same-sex relationship, 2.25 for those who do not) show that participants in these groups indicate differing opinions on same-sex marriage with each group landing on opposite side of neither agree nor disagree (Table 5). These differences in attitudes towards same-sex marriage between those who do not have contact with LGBQ people and those who do were not found to be statistically significant according to the results of the multinomial logit (Table 7).

Should Same-Sex Adoption be Legal?

Similar to general attitudes towards same-sex adoption, opinions about the legalization of same-sex adoption were varied. Participants who know someone who is LGBQ and those who know someone in a same-sex relationship were slightly more likely than their counterparts to agree that it should be legal, but no significant trends were identified (Table 5). Table 5 shows that the difference in the means ranged from 1.16 to 1.50 between those with and without LGBQ contact regarding the legalization of same-sex relations and same-sex marriage respectively. Yet the difference in means regarding the legalization of same-sex adoption was only .60 between those who do and do not know someone who is LGBQ and .87 between those who know someone in a same-sex relationship and those who do not (Table 5) illustrating the low level of contact-associated variability within this dependent variable. The lack of statistically significant predictors in the multinomial logit underscores these trends (Table 8).

Impact of Types of Contact

A review of the cross-tabulations seems to show that participants who know LGBQ people in relationships are more likely to support same-sex-headed families legally and in general compared to people who have LGBQ family members or friends (Tables 2 and 5). Regarding general attitudes towards same-sex marriage, the odds ratio for those who do not know someone in a same-sex relationship (7.23) is much greater (and statistically significant) compared to the odds ratio of those who do not have LGBQ identifying family or friends (1.11, not statistically significant) when comparing always wrong to not wrong at all (Table 3). Additionally, less than five percent of people who know someone in a LGBQ relationship disagreed with the statement, “Same-sex relations between consenting adults should be legal,” (4.8% strongly disagree, 0.0% disagree), whereas a quarter of participants who have LGBQ family and friends disagreed with the statement (12.5% strongly disagree, 12.5% disagree; Table 5).

For these reasons, the difference in the impact of types of contact was analyzed. Participants who had LGBQ contact were categorized into two groups: 1) those with LGBQ family and friends but who do not know someone in a same-sex relationship and 2) those who know LGBQ people in a relationship. The categorical nature of the dependent variables, their non-normal distribution, and the dichotomous categorical independent variable lead to the application of the Mann-Whitney test (Table 9). Attitudes towards same-sex marriage in general significantly differed between those who have family and friends that are LGBQ (Mdn = 1.00) and those who know LGBQ people in relationships (Mdn = 3.00, Table 9) with the latter having a greater impact. Attitudes towards legalizing same-sex relations also differed between the two groups (LGBQ family/friends Mdn = 3.00, LGBQ in relationship Mdn = 4.00; Table 9) demonstrating the greater prediction ability of contact with those in LGBQ relationships in this regard. There were no difference in medians for the three other dependent variables (Table 9).