Introduction

The present research is designed to investigate the connection between friendship value ratings and cross-category friendship patterns among adult women and men in the United States. This research is comparative by design and considers whether individuals who have, versus do not have, cross-category friendships differ in what they value as important in a close friendship. In addition, it is designed to determine whether friendship values are similarly related to different types of cross-category friendships or whether there are unique considerations for cross-orientation and cross-race friendships. Therefore, parallel analyses were conducted to examine cross-orientation and cross-race friendships across gender and identity status (minority and majority). Participants completed a friendship profile questionnaire by reporting basic demographic information about themselves and their close friends. Participants’ importance ratings of six different friendship values were utilized to interrogate existing friendship patterns. Three general friendship values (trust and honesty, respect friend as person, there when needed) and three cross-identity salient friendship values (similar lives & experiences, similar values, nonjudgmental) were considered.

This research uses feminist intersectional theory (Collins 2000; Hooks 1984; McCall 2005) to understand patterns of cross-category friendship across identity of participants. Feminist intersectional theory attends to not only the intersections among gender, sexual orientation, and race for the present research, but also allows an analysis of how dimensions of identity (minority and majority) are related to inequality and / or oppression. Considering inequality in the context of friendship ultimately permits an understanding of how identity differences shape social and personal relationships. Because notions of identity are culturally constructed, the review of past research and the rationale for the present study is situated in the U.S. research literature where all samples of research cited are U.S. based unless otherwise noted. Although the specific research conceptualization of gender, sexual orientation, and race used for the study is U.S. based, larger notions of how identity negotiation in friendships may be impacted across minority / majority dimensions of identity could have implications for understanding friendship in other cultural contexts.

Cross-Category Friendships

Homophily – the notion that people connect with others who are similar to themselves – is a basic principle of social relationships (McPherson et al. 2001). Research has demonstrated that friendships reflect homophily where individuals choose friends who are similar to themselves on the basis of gender, age, race, social economic status, and sexual orientation (Duck 1991; Galupo 2009; Rose 1985; Ueno 2010). In general, friendships have been regarded as distinct from other types of social relationships precisely because they are constructed around mutuality (McWilliams and Howard 1993) and equality (Suttles 1970). Research on same-category friendships emphasizes that these friendships are particularly important because they represent a social relationship in which individuals experience relative equality. For example, in his research on friendships between gay men, Nardi (1999) describes how these friendships can provide opportunities for gay men to experience equity not easily achieved in other friendships where social identity is being bridged and where sexual orientation or gender require constant negotiation.

In contrast, cross-category friendships exist between individuals who are positioned differently across social identities. Because cross-category friendships defy the expected friendship pattern of homophily, these friendships are often conceptualized as novel and as developing despite significant barriers or obstacles (Galupo 2009; O’Meara 1989). Research characterizing cross-category friendships has focused on the negotiation of difference and inequality as a primary focus for friends. In fact, exposure to different perspectives is often cited as a primary benefit of cross-category friendships (Canary et al. 1997; Galupo and St. John 2001; Hall and Rose 1996; Werking 1997). For example, women interviewed about their cross-orientation friendships discussed ways in which friendships allowed for critical evaluation of stereotypes, moving beyond dichotomous thinking, and the ability to consider alternate perspectives (Galupo and St. John 2001). Likewise, Hall and Rose (1996) emphasize that cross-race friendships benefit friends by allowing a context for developing greater appreciation for cultural diversity and can lead toward more effective political organizing.

Although cross-category friendships are a site for exploring differences between friends, individuals with majority (e.g. heterosexual, White) and minority (e.g. sexual minority, racial minority) identities may approach and experience friendships differently. For example, friendship in general is particularly salient for individuals whose identity puts them at odds with social norms (Weeks 1995). Social minorities, then, may rely on friendship to meet needs not otherwise met by larger society. And although friendship patterns in general reflect homophily, there is evidence to suggest that sexual and racial minorities are more likely to form cross-category friendships (Galupo 2009) than their majority counterparts. Cross-category friendships may offer minority individuals a unique location for finding acceptance from the majority that is not otherwise found in other family and friendship relationships. However, because of the heightened social importance of such friendships, minority individuals may enter into these cross-category friendships with different friendship expectations than majority individuals. For example, past research on cross-orientation friendships has suggested that these friendships may exist, in some cases, despite the judgmental attitudes and non-acceptance directed toward sexual minority issues and identities on the part of the heterosexual friend (Galupo 2007a; Galupo et al. 2004; Price 1999). Some sexual minorities view the routine judgment and invalidation of their identity as a necessary sacrifice for maintenance of friendships with heterosexual friends (Galupo et al. 2004). Thus, the study of cross-category friendships can serve as an important research model as it presents a unique location for considering how individuals are able to bridge difference and negotiate inequities within their personal relationships.

Cross-Category Friendship Patterns and Feminist Intersectional Theory

The present research utilizes feminist intersectional theory in order to understand cross-category friendship patterns in an adult U. S. sample. Feminist intersectional theory emphasizes the importance of examining relationships among social identities as intersecting categories of oppression / inequality (Collins 2000; Hooks 1984; McCall 2005). This theoretical framework initially focused on race, class, and gender and originated from the critique of both gender- and race-based research for failing to acknowledge individuals living at the intersections of the two.

More recent conceptualizations are inclusive of sexual orientation, specifically addressing the role of homophobia and heterosexism in the lives of women and racial minorities (Anzaldúa 1990; King 1990; Trujillo 1991). This literature can inform an intersectional approach to researching friendship experience across gender, sexual orientation, and race. Framing sexual orientation from an intersectional lens brings into focus dimensions of inequality and power that surround cultural meanings of gender and sexual orientation. Viewing sexual orientation within intersectional theory shifts the focus from the “unnatural” and “abnormal” conceptualizations of sexual minority experience traditionally highlighted by psycho-medical perspectives, while making heterosexual identity subjective and open as category of inquiry. Understanding friendships across sexual minority / heterosexual experience, then, allows a comparative dimension which invites exploration of inequalities across sexual orientation potentially revealing how heteronormative attitudes shape friendship dynamics and choice. Simultaneously, this framework allows a parallel exploration of friendship experiences across racial minority / White dimensions of identity. Ultimately feminist intersectional theory provides a conceptual framework from which to consider the intersecting influences of sexism, homophobia, and racism on friendship dynamics (Galupo 2006).

Feminist intersectional theory with its focus on structural inequalities provides a particularly useful framework for understanding cross-category friendships as they are, by definition, structured across dimensions of social inequality. O’Meara (1989) discusses the inherent inequities present in the structural context of cross-gender friendships as presenting a major challenge to their development and maintenance. The same could be said for cross-race and cross-orientation friendships. In addition, feminist intersectional theory attends to the simultaneous and intersecting influences across categories. Beyond a few notable exceptions past research on cross-category friendships has not allowed for a comparative analysis of friendship patterns across multiple dimensions as the majority of this research has focused on one social identity in isolation. There has been a longer history of research focused on understanding cross-gender and cross-race friendships, with more recent attention on cross-orientation friendships. Although understanding friendships across any one of these identities is important in its own right, interpreting results regarding cross-category friendship patterns when studied in isolation should be done cautiously. For example, if studying cross-orientation friendships it is possible that emerging patterns due to a general difference in social identity could be erroneously attributed to differences specific to sexual orientation (and vice versa). An intersectional analysis comparing cross-orientation and cross-race friendships would allow a better understanding of the effect of identity differences in general, and sexual orientation differences, in particular.

McCall (2005) outlined several options for intersectional researchers. Some of the initial research on cross-category friendships utilized an intracategorical complexity approach, which focuses on particular social groups at neglected points of an intersection (McCall 2005). This is typified by research, for example, that provides an in-depth description of friendship experiences between White and Black lesbians, as does Hall and Rose (1996) or between gay and straight men (Price 1999). This type of research has been successful in revealing nuances in friendship dynamics that exist when simultaneously negotiating sexism, homophobia, and / or racism, and in bringing to the forefront the lived experience for individuals who enjoy friendships at the nexus of these intersections. Intracategorical complexity approaches, however, have been criticized in general for managing complexity by focusing on the intersection of single dimensions of multiple categories (McCall 2005). For friendship research this is exemplified by research that seeks to understand the intersection between sexual orientation and sex, but focuses exclusively on men’s (e.g. Nardi 1992; 1999; Tillman-Healy 2001) or women’s experiences (e.g. Galupo et al. 2004; O’Boyle and Thomas 1996).

McCall (2005) also describes intercategorical complexity approaches which provisionally adopt categories in order to document relationships among multiple and potentially conflicting dimensions of identity and experience. When applied to friendship research, the intersecting patterns or relationships among social groups become the center of analysis. Categorization, then, is used provisionally as a way to understand intersecting relationships and these relationship patterns become central to the analysis rather than as a backdrop for explaining friendship experiences. Intercategorical complexity approaches, then, incorporate multiple groups and are comparative by nature.

Taking an intercategorical complexity approach, Galupo (2009) demonstrated that sexual orientation, gender, and race do interact in complex ways to shape friendship patterns. In a study that compared cross-orientation, cross-gender, and cross-race friendships between sexual minority and heterosexual adults Galupo (2009) described larger friendship patterns that could be explained by social inequalities. Although same-category friendships are still the overall basic friendship pattern, cross-gender friendships are more common than cross-race friendships, which are in turn more common than cross-orientation friendships. In addition, cross-category friendship patterns are impacted by identity where, for example, heterosexual participants report significantly fewer cross-orientation friendships than sexual minority participants. Participants who were both sexual and racial minorities experienced more of their friendships in a cross-race context. These patterns suggest that the more marginalized an individual is with respect to larger culture, the less likely she / he will find friendships among individuals of similar social identities. That is, the overall trend toward friendship homophily is disrupted at the intersections of minority identity. In interpreting these findings Galupo (2009) asserts that these results cannot be solely explained by opportunity or demography. Rather, sociopolitical attitudes shape these friendship “choices.”

Galupo’s (2009) research provides a blueprint for mapping adult friendship patterns across gender, sexual orientation, and race. Future research is needed to elucidate how these friendship patterns are related to larger sociopolitical attitudes and friendship expectations. Although gender differences have been found with regard to friendship expectations of same-gender friends (Hall 2011), friendship expectations have not been explored in relation to cross-category friendships. Studying cross-category friendship patterns specifically allows an understanding of when friendship homophily is being violated and what attitudes allow this violation.

Present Study and Hypotheses

The present research examines whether individuals who have, versus do not have, cross-category friendships differ in what they value in a close friendship. Using an intersectional (intercategorical) approach this research is methodologically unique in that it allows for four distinct levels of comparisons, allowing for: 1) statistical comparisons between individuals who have, versus do not have cross-category friendships; 2) statistical comparisons across majority (male, heterosexual, White) and minority (female, sexual minority, racial minority) dimensions of identity; 3) a consideration of patterns of responses across general and cross-identity salient friendship values; and 4) a consideration of patterns of responses across cross-orientation and cross-race friendships.

This research investigates the relation between cross-category friendship patterns and six different friendship values. These six friendships values were selected based on a broad review of the friendship literature and are considered general characteristics of friendships (e.g. Fehr 1996; Rawlins 1992.) Three general friendship values (trust & honesty, respect friend as a person, there when needed) were selected because they are regarded as basic friendship characteristics with no conceptual or theoretical connection to cross-category friendships specifically. In addition, these characteristics are not particularly salient to differences across sexual or racial identity.

Three cross-identity salient friendship values (similar lives & experiences, similar values, non-judgmental) were selected for this research because they, too, are basic friendship characteristics. However, because cross-category friendships violate notions of homophily and similarity and because cross-category friendships center on the negotiation of differences between friends (Canary et al. 1997; Galupo and St. John 2001; Hall and Rose 1996; Price 1999; Rose 1996; Werking 1997) these friendship values may be uniquely related to the development of cross-category friendships. Specifically, these cross-identity salient friendship values may be emphasized less for individuals who have cross-category friendships. In addition, these friendship values may be differently considered in cross-category friendship by individuals who identify with the majority versus the minority. The framing of general and cross-identity salient friendships values is done theoretically, as different predictions are expected across these two groups of friendship values. However, it is important to note that each of the six friendship values was analyzed separately. That is, each of the six friendship values was treated as a separate dependent variable.

The following specific hypotheses were explored and past research findings supporting these predictions are cited, where relevant: 1) Consistent with past research (Hall 2011) it was predicted that women would have higher friendship value ratings than men for both general and cross-identity salient friendships values; 2) Because cross-orientation friendships in general are structured around difference (Galupo and St. John 2001; Hall and Rose 1996; Muraco 2012; Price 1999), it was predicted that those with cross-orientation friendships would rate similar lives & experiences and similar values as less important than those without cross-orientation friendships; 3) Because cultural debates surrounding sexual orientation are uniquely framed in a religious and moral context (Herek 1987, 1991) it is expected that non-judgmental would be related to cross-orientation friendships. Specifically, it was expected that sexual minority participants with cross-orientation friendships would rate non-judgmental as lower than those without such friendships. This is consistent with research that has suggested that some sexual minorities view the routine judgment and invalidation of their identity as a necessary sacrifice for the maintenance of friendships with heterosexual friends (Galupo et al. 2004). A de-emphasis, then, on non-judgmental values may be related to the development of cross-orientation friendships; and 4) Because cross-race friendships (like cross-orientation friendships) are structured around difference (Hall and Rose 1996; Rawlins 2009), it was predicated that those with cross-race friendships would rate similar lives & experiences and similar values as less important than those without cross-race friendships.

Method

Participants

Participants were 1415 women (n = 1033) and men (n = 382). Participants ranged in age from 18–80 (M = 44.78, SD = 14.32). With regard to sexual orientation, participants self-identified as heterosexual (n = 1010) and sexual minority (n = 405). There was considerable diversity in the sample where 1018 (71.9 %) participants identified as White and 397 (28.0 %) identified as racial minorities, specifically: 225 (15.9 %) African American / Black; 58 (4.1 %) Asian / Asian American; 58 (4.1 %) Other / Bi-Racial; 48 (3.4 %) Hispanic / Non-White; and eight (.6 %) Native American. With regard to educational background, 1022 (72.23 %) had either attended some college or were college graduates; 201 (14.20 %) had either attended some graduate school or had already earned professional or graduate degrees; 188 (13.29 %) had either attended some high school or were high school graduates; and four participants (0.28 %) did not report. With regard to social economic status, participants described themselves in the following way: Middle Class (n = 792; 56.0 %); Upper Middle Class (n = 309; 21.8 %); Working Class (n = 289; 20.4 %); Upper Class (n = 19; 1.3 %); and six individuals (.5 %) did not report. Participants represented all regions of the United States, residing in all 50 states and Washington D.C.

Table 1 provides demographic information disaggregated by gender. Preliminary analysis confirmed that there were no significant gender differences in demographic information for sexual orientation, racial identity, education, and social class. However, women in the sample were significantly older than men in the sample. Although age was not used as a grouping variable for the present research, age was controlled for in all analyses reported here.

Table 1 Sample demographics across gender
Full size table

Recruitment Procedure

Participants were recruited into the study utilizing two different methods. In both cases, participation was completely voluntary and was offered with no incentives. On-line participants (n = 802; 56.7 %) were initially recruited from two general websites for posting psychological studies. Paper and pencil surveys (n = 613; 43.3 %) were administered from 12 volunteer recruiters in different states across the country (five colleges, seven businesses). From each of the 12 sites, snowball methodology was used where interested participants were given additional surveys to circulate among their contacts. On-line participants were also encouraged to share the survey link with other individuals. A preliminary analysis revealed no significant difference in the responses across data collection methods (paper-and-pencil and on-line survey) with regard to participant demographics and on total number of friends. All subsequent analyses, then, used the combined sample.

Measures

Participants completed a friendship questionnaire developed by the first author reporting basic demographic information (gender, race, age, education, sexual orientation) about themselves and up to eight of their close friends. Participants were asked to include who they consider to be “close friends” based on their own personal definition.

In developing a friendship profile, the number of close friendships and the number of same- and cross-category friendships (sexual orientation and race) were determined. For this study, friendship profiles were coded on the basis of whether the participant had at least one cross-category friendship per category. For example, participants with no cross-orientation friendships were coded as 0 and those individuals with one or more cross-orientation friendships were coded as 1. Among those with at least one cross-orientation friendship, a preliminary one-way MANOVA found no differences between individuals with different numbers of such friendships for each of the six friendship values. Likewise, among those with at least one cross-race friendship, a one-way MANOVA found no differences for each of the six friendship values.

Participants were also asked to provide importance ratings on a 5-point likert scale (1 = not important / 5 = extremely important) for 6 different friendship values. Three were general friendship values (trust & honesty, there when needed, respect as friend) while three were cross-identity salient friendship values (similar lives & experiences, similar values, nonjudgmental). Participants rated each of the 6 friendship values independently. Even though different predictions are made for general and identity salient friendship values, each of the six friendship values were analyzed as separate dependent variables.

Only participants with complete friendship profiles (including demographic information about themselves and their friends) were included in this sample. Because of the different recruitment strategies and sites we do not have a reliable estimate of how many participants were excluded for incomplete friendship profile data. In several cases participants were missing one or two of the friendship value scores. Because friendship value scores were calculated and analyzed independently from one another, we were able to retain those participants in the sample and use their data where available. The level of missing data was very low. Out of the 1415 total participants, missing data points for each of the friendship values were as follows: trust / honesty = 11; respect friend as person = 12; there when needed = 12; similar lives & experiences = 14; similar values = 11; nonjudgmental = 11).

Results

Preliminary Analyses

Consistent with a feminist intersectional approach which focuses on experiences across dimensions of inequality, participants were grouped based on majority / minority identities for both sexual orientation (i.e. heterosexual / sexual minority) and race (White / racial minority). Using this categorization, then, all individuals who were not heterosexually identified were grouped together as sexual minorities and all individuals who were not White were grouped together as racial minorities. A preliminary MANOVA found no significant interaction or main effects of sexual orientation and gender for any of the six friendship values among sexual minorities. This supports the grouping of lesbian / gay and bisexual participants together as sexual minorities for the purpose of exploring differences between sexual minority and heterosexual participants. Likewise, a preliminary additional MANOVA found no significant interaction or main effects across race and gender for any of the six friendship values among racial minorities. This supports the grouping together participants who identify as African American, Asian American, Hispanic, Native American and Other as racial minorities.

Friendship Values & Cross-Category Friendships

Table 2 illustrates descriptive statistics and correlations among the six friendship values. The relation between friendship value ratings and cross-category friendship patterns was investigated separately for cross-orientation and cross-race friendships. Two separate 3-way MANOVAs were conducted to investigate whether friendship value ratings differed for individuals with, versus without, at least one cross-category friendship.

Table 2 Descriptive statistics and correlations among general and cross-identity salient friendship values
Full size table

Cross-Orientation Friendships

Table 3 provides mean scores for each of the six friendship values across gender, sexual orientation, and cross-orientation friendships. A 2 × 2 × 2 factorial investigated the potential intersecting effects of gender, sexual orientation, and cross-orientation friendships on friendship value ratings. In cases of interaction effects, Bonferroni post hoc t-tests were used to explore simple main effects.

Table 3 Friendship value ratings across gender, sexual orientation and cross-orientation friendships (COF)
Full size table

General Friendship Values

Importance ratings for each of the three general friendship values (trust / honesty, respect friend as person, there when needed) did not differ across sexual orientation (heterosexual, sexual minority), or based on whether participants had, or did not have, cross-orientation friendships. There were no significant 2 or 3 way interactions across gender, sexual orientation, and cross-orientation friendships. Additionally, there were no significant main effects for sexual orientation or cross-orientation friendships. However, for all three of the general friendship values, there was a significant main effect of gender. Women (M = 4.92, SD = 0.35) rated trust / honesty as more important when compared to men, (M = 4.82, SD = 0.58), F(1, 1373) = 8.54, p = .004, partial η2 = .006. Women (M = 4.88, SD = 0.38) rated respect friend as person as more important when compared to men, (M = 4.74, SD = 0.67), F(1, 1373) = 6.18, p = .013, partial η2 = .004. Additionally, women (M = 4.64 SD = 0.66) rated there when needed as more important when compared to men, (M = 4.41, SD = 0.91), F(1, 1373) = 6.83, p = .009, partial η2 = .005.

Cross-Identity Salient Friendship Values

For similar lives & experiences there were no significant 2 or 3 way interaction effects between sexual orientation, gender, and having at least one cross-orientation friendship. Additionally, there was no significant main effect of gender. There was a significant main effect of sexual orientation where sexual minorities (M = 3.06, SD = 1.11) rated similar lives & experiences as more important than heterosexuals, (M = 2.93, SD = 1.16), F(1, 1373) = 11.78, p = .001, partial η2 = .009. There was also a significant main effect of cross-orientation friendships where individuals with cross-orientation friendships (M = 2.86, SD = 1.14) rated similar lives & experiences as less important than those individuals without such friendships, (M = 3.02, SD = 1.15), F(1, 1373) = 10.88, p = .001, partial η2 = .008.

For similar values there were no significant interaction effects among sexual orientation, gender, and having at least one cross-orientation friendship. However, there were significant main effects for sexual orientation and cross-orientation friendships. Sexual minorities (M = 3.86, SD = 1.00) rated similar values as more important than heterosexuals, (M = 3.81, SD = 1.06), F(1, 1373) = 4.32, p = .045, partial η2 = .003. In addition, individuals with cross-orientation friendships (M = 3.75, SD = 1.03) rated similar values as less important without such friendships, (M = 3.85, SD = 1.05), F(1, 1373) = 3.11, p = .05, partial η2 = .003.

For non-judgmental ratings, there was no significant 3 way interaction among sexual orientation, gender, and having at least one cross-orientation friendship. There was a significant interaction between sexual orientation and having at least one cross-orientation friendship, F(1, 1372) = 4.71, p = .05, partial η2 = .004. For heterosexual participants, there was no significant difference in non-judgmental ratings when comparing those with (M = 4.21, SD = 1.00) and without (M = 4.07, SD = 1.07) a cross-orientation friendship, t(978) = 1.51, p = .13, d = .13. For sexual minority participants, individuals with at least one cross-orientation friendship (M = 4.40, SD = .92) rated non-judgmental as less important than those without a cross-orientation friendship (M = 4.65, SD = .76), t(402) = 2.23, p = .02, d = .30. Additionally, there was a main effect of gender where women (M = 4.28, SD = 0.97) rated non-judgmental as more important when compared to men, (M = 3.94, SD = 1.14), F(1, 1373) = 16.14, p = .001, partial η2 = .012.

Cross-Race Friendship Patterns

Table 4 provides mean scores for each of the six values across gender, race, and cross-race friendships. A 2 × 2 × 2 factorial was conducted to investigate the effects of gender, race, and cross-race friendships on friendship value ratings. In cases of interaction effects, Bonferroni post hoc t-tests were used to explore simple main effects.

Table 4 Friendship value ratings across gender, race and cross-race friendships (CRF)
Full size table

General Friendship Values

Importance ratings for each of the three general friendship values (trust / honesty, respect friend as person, there when needed) did not differ across race (White, racial minority), or based on whether participant had, or did not have, cross-race friendships. There were no significant 2 or 3 way interactions across gender, race, and cross-race friendships. Additionally, there were no significant main effects for race or cross-race friendships. Consistent with the findings reported above, a main effect of gender was found for all three general friendship values (means and statistics above).

Cross-Identity Salient Friendship Values

As predicted, importance ratings for the three identity salient friendship values (similar lives & experiences, similar values, non-judgmental) did reveal some differences. For similar lives & experiences there were no interaction effects between gender, race, and having at least one cross-race friendship. There were no main effects for gender or race. However, there was a main effect of cross-race friendships where individuals with at least one cross-race friendship rated similar lives & experiences as less important (M = 2.85, SD = 1.11) than those with no cross-race friendships (M = 3.04, SD = 1.16), F(1, 1392) = 4.45, p = .035, partial η2 = .003.

For similar values, there was no significant 3-way interaction. There was a significant interaction between race and having at least one cross-race friendship, F(1, 1388) = 4.63, p = .03, partial η2 = .003. Post-hoc t-tests indicated that for racial minorities, ratings for similar values did not differ between individuals with and without cross-race friendships, t(387) = 1.23, p = .21, d = .06. For Whites, individuals with at least one cross-race friendship (M = 3.77, SD = 1.02) rated similar values as less important than those without a cross-race friendship (M = 3.89, SD = 0.97), t(1013) =4.17, p < .03, d = .12.

For nonjudgmental there were no interaction effects between gender, race, and having at least one cross-race friendship. There was no main effect for cross-race friendships. There was a main effect of race where racial minorities rated nonjudgmental as less important (M = 4.09, SD = 1.09) than Whites (M = 4.23, SD = 1.00), F(1, 1392) = 4.06, p = .044, partial η2 = .003. Replicating findings above, a main effect for gender was found.

Discussion

The present research is the first to consider the relation between friendship value importance ratings and adult U.S. friendship patterns across gender, sexual orientation, and race. In interpreting the findings of this research it is important to note that the present focus is on close, rather than casual, friendship networks. Close friendships are characterized by more frequent interaction and affection, are less dependent upon proximity, provide more benefits, and are enacted more exclusively than are casual friendships (Hays 1989; Rose and Serafica 1986). It is likely, then, that the criteria for developing close friendships would be more stringent than the criteria for casual friendships and that close friendships patterns may exhibit stronger trends toward similarity.

Cross-Category Friendships and Similarity

In the context of the larger body of friendship literature, the study of cross-category friendships is, in part, a study of friendships that develop despite the general tendency toward homophily. A main finding of the present research is that individuals who report having cross-orientation and cross-race friendships place less importance on the friendship value of similar lives & experiences than those with no such friendships. Likewise, individuals with cross-orientation friendships and Whites with cross-race friendships rated similar values as less important than those with no cross-race friendships. These findings are consistent with past research that suggests that a primary focus of cross-category friendships is on the negotiation of difference (Canary et al. 1997; Galupo and St. John 2001; Hall and Rose 1996; Price 1999; Rawlins 2009; Werking 1997). The implication, then, is that close cross-category friendships do not seem to develop by ignoring or downplaying differences between friends but rather they are more likely when friends do not place as much emphasis on similarity as being important within the friendship.

Cross-Category Friendships and Feminist Intersectional Theory

Intersectional theory, specifically intercategorical complexity approaches, suggest that meaning and analysis can be interpreted from the pattern of results across multiple dimensions of analysis. The present research established clear patterns of results where: 1) individuals with, and without, cross-category friendships rated friendship values differently; 2) the relationship between friendship value ratings and cross-category friendship patterns differed, in some cases, across minority and majority identity; 3) general and cross-identity salient friendship values were differently related to cross-category friendship patterns; and 4) friendship values and cross-category friendship patterns differed, at times, across cross-orientation and cross-race friendships.

Taken together, these research findings detail the relation between friendship value ratings and overall cross-category friendship patterns and present pockets of evidence that may serve to help further direct future research. In particular these findings suggest that some intersections of identity may be interrogated for different purposes. For example, for researchers interested in how cross-orientation and cross-race friendships are similarly negotiated, a focus on similar lives & experiences and similar values may provide a research model to explore. For those interested in how cross-orientation friendships are uniquely experienced, a fruitful direction might be further exploring how judgment is conceptualized in connection to sexual minority friends. Likewise, for those interested in how cross-category friendships function similarly to friendship in general, this research has established three friendship values (trust & honesty, respect friend as a person and there as needed) that did not differ across cross-category friendship patterns.

Cross-Category Friendships and Inequality

Although recognizing intersecting points of identities is fundamental to an intersectional perspective so, too, is the notion of inequality and / or oppression. The present research established that importance ratings for key friendship values are related to cross-category friendship patterns. That cross-orientation and cross-race friendship patterns were differently related to some friendship values provides an avenue for understanding how heterosexism, sexism, and racism uniquely shape cross-category friendship patterns. In addition, an important methodological approach for this research allowed consideration of the relationship between friendship value ratings and cross-category friendships across majority / minority dimensions of identity. Interpretation of the current findings, then, is best understood with the lens of inequality and how that shapes lived experience.

A present finding that illustrates the importance of this approach follows. Sexual minority individuals rated nonjudgmental as less important than those without such friendships. The same finding did not hold true for heterosexual participants. Although this finding may seem counterintuitive, it becomes clearer if interpreted within an understanding of the relative inequality between cross-orientation friends. Social attitudes toward sexual minorities are uniquely framed within cultural debates based on religious and moral judgment (Herek 1987, 1991). In addition, sexual minority individuals often experience social isolation or rejection associated with homophobia in larger culture, including loss of friendships (Weinstock 2000). Friendships, then, become emphasized in the social lives of sexual minorities who often experience their friends as “family” (Nardi 1992; Weinstock 2000) or as “families of choice” (Weston 1991). Cross-orientation friends provide a unique opportunity for sexual minorities to gain acceptance from heterosexuals not readily offered in their within community friendships or from their families of origin (Galupo and St. John 2001). However, past research has also shown that when cross-orientation friendships do develop they are often maintained at the expense of sexual minorities’ identities where identity is either rendered invisible, is expected to be concealed, or is somehow otherwise managed, (Galupo 2007a; Price 1999). In fact, some sexual minority individuals view the routine judgment and invalidation directed toward their identity as a necessary sacrifice for the maintenance of friendships with heterosexual friends (Galupo et al. 2004). It follows, then, that sexual minority individuals with cross-sexual orientation friendships rate nonjudgmental as less important than those who do not have such friendships. The development of cross-orientation friendships, then, may be related to sexual minorities lowering their expectations of heterosexual friends with regard to being judgmental. Alternatively, sexual minorities without cross-orientation friendships may emphasize nonjudgmental as a friendship value more which may then serve as a barrier or obstacle to cross-orientation friendship development. Although the present finding is correlational and these interpretations are consistent with an inequality / intersectional framework, future research using a more direct means of assessing perception is necessary to fully account for this connection.

Although cross-category friendships are differently experienced for majority and minority individuals, it is important to move beyond using inequality solely to explain minority friendship experience. Although it is seemingly easier to decipher how heterosexism, sexism, and racism operate to shape friendship patterns for minority individuals, these larger sociopolitical attitudes impact majority individuals’ friendship patterns as well (Galupo 2006). An example from the present research that highlights the distinction between majority and minority experience emerged relative to cross-race friendship patterns. For White participants (but not for racial minority participants) cross-race friendships are associated with lower importance ratings for similar values. This may suggest that White participants view racial minorities as having different values from themselves and cross-race friendships became more likely when they placed less importance on sharing similar values within their friendship. Although this interpretation is speculative it is consistent with O’Connor (1992) suggestion that cultural ideas of “otherness” are inherent within friendship structures and that these play out in the way similarity is conceptualized within friendships. Future research is needed in order to directly assess perceptions of similarity between individuals in cross-category friendships.

The fact that friendship value ratings may be differently related to cross-category friendship patterns for majority and minority individuals points to the need for future research that takes a more direct approach in understanding how perceptions of others impact cross-category friendship patterns. These findings also suggest that future research should equally attend to both minority and majority experience.

Limitations of the Present Study and Directions for Future Research

One limitation of the present research is that the sampling strategy included snowball recruitment and this may have disproportionately represented certain friendship groups and characteristics, ultimately biasing the results toward similarity. Additionally, individuals reported both their own identity and that of their friends, a method that may actually bias responses toward homophily where participants overestimate similarities between themselves and their friends (Ueno 2010). For the current research we allowed participants to define close friendships for themselves which may have led to variability in the way they approached the questions. Friendship nomination studies, typical in childhood friendship research or based on investigating friendships within a defined setting (such as school or work) allow confirmation, not only of identity, but of friendship – where friendships are defined when both individuals nominate one another. This type of method may be ideal in many ways because of the way friendships are characterized by reciprocity and mutuality (McWilliams and Howard 1993). This presents a useful direction for future research on cross-category friendships.

The following statistical issues should also be considered in the wider interpretation of the present findings. Although the present results include statistically significant cross-category friendship patterns across gender, race, and sexual orientation it is important to note that the effect sizes are relatively small. Additionally, the mean scores for the general friendship values were higher (4.58–4.89 on a 5-point scale) than those for the cross-identity salient friendship values (2.96–4.2). It is possible that the ceiling effect in the general friendship values could have contributed to the lack of differences between the groups.

The simple dichotomous grouping of all independent variables represents an additional limitation. The grouping of participants by identity across gender (female and male), sexual orientation (sexual minority and heterosexual), and race (racial minority and White) is consistent with intercategorical complexity approaches to feminist intersectional methodology, which allows for provisional adoption of categories in order to understand larger patterns of inequality (McCall 2005). In addition, these simplified categories were necessary in order to allow for enough statistical power to make comparisons and meaningful findings, did in fact, emerge given these groupings. This simplified grouping of identity, however, is not without its limitations. Past research has found distinct friendship patterns among sexual minorities (Galupo 2007b) and racial / ethnic minorities (Way and Chen 2000; Way et al. 2001) and it is clear that future research should allow for more graded comparisons among these groups in order to allow for a fuller picture of how friendship experiences are influenced at the intersections of social identities. Likewise cross-category friendships for the present research were measured categorically as either having, or not having friendships across either sexual orientation or race. Future research should address within group differences among individuals who have cross-category friendships.

Despite the limitations, the present research makes a unique contribution to the friendship literature as it is the first to explore the relation between friendship value ratings and cross-category friendship patterns. Using a comparative methodology, this research extends past research on cross-category friendship patterns across both sexual orientation and race. Specifically it suggests that cross-category friendships do not seem to develop by ignoring or downplaying differences between friends but rather they are more likely when friends do not place as much importance on similarity as a criterion for friendship. Although only six friendship values were utilized for the present study, these findings suggest that continued exploration of the relation between friendship values and cross-category friendship patterns may be an important direction for future research.