Sex Differences Partially Moderate the Relationships Between Personal Values and the Preference for Cross-Sex Friendships (Heterosociality)
Abstract
Abstract. A key differentiation in studies on friendship research is between the same-sex and cross-sex friendships of women and men. Although most women and men prefer same-sex over cross-sex friends, most people do commonly have cross-sex friendships, and there are large interindividual differences in the proportions of cross-sex friends in individual friendship networks. Recent studies have suggested that same-sex and cross-sex friendships fulfill different goals for women and men. Therefore, individuals’ personal values (as representations of their enduring goals) may be associated with the types of friendships they prefer. The present study explores associations between personal values and people’s preferences for cross-sex friendships (heterosociality). A sample of 1,333 participants completed the assessment. Results showed that the associations were partially moderated by sex. For men, the value of tradition, whereas for women, the values of security, self-direction, and tradition were found to be significantly associated with the individual proportions of cross-sex friends. These findings contribute to understanding friendship selection and underline the importance of differentiating between same-sex and cross-sex friendships in women and men.
Friendship has repeatedly been shown to be an important type of social relationship. Friends contribute to success in coping with stress and anxiety (Winstead et al., 2016), individual well-being, general life satisfaction (Demir, 2015; Demir & Weitekamp, 2007; Lewis et al., 2015), and longevity (Giles et al., 2005) and may even have been key to survival during human evolution (Lewis et al., 2012). According to previous research, different types of friendships can be discerned: for instance, on the basis of closeness (e.g., general, close, or best friends), purpose (e.g., friends with benefits), or similarity (e.g., concerning sex, age, or personality). Of these, the distinction between same-sex and cross-sex friendships (i.e., similarity with regard to sex) may be pivotal (Altmann, 2020; Lewis et al., 2011; Reeder, 2003).
Studies have consistently shown that most people generally prefer to have same-sex over cross-sex friends (i.e., most people report having more same-sex than cross-sex friends; Baumgarte & Nelson, 2009; Reeder, 2003). However, the vast majority of people also report having at least one cross-sex friend, and on average, people report that around one-third of their friendship network consists of cross-sex friends (Altmann, 2020; Marsden, 1990). When looking at the distribution of individual preferences for cross-sex friendships (i.e., heterosociality), there appears to be a considerable amount of variance. People differ substantially in these preferences so that people can be found who exclusively have same-sex friendships (low heterosociality), people who exclusively have cross-sex friendships (high heterosociality), and everything in between (Altmann & Roth, 2020).
To explain this interindividual variance in friendship preferences using personality constructs, the Big Five and related constructs have predominantly been used in previous studies (Altmann & Roth, 2020; Brown & Lester, 1997; Laakasuo et al., 2016; Lönnqvist et al., 2014). Such studies have been able to show that these traits are associated with same-sex and cross-sex friendship choices and that these relationships differ between women and men, but also that these associations tend to be small and often inconsistent. Roccas and colleagues (2016) argued that trait constructs (e.g., the Big Five) might be less sufficient for explaining behaviors over which people have voluntary control, which may also apply to the forming and maintaining of friendships. Instead, in order to predict voluntary behavior, the authors suggested that it is more beneficial to look at motivation.
Functions of Same-Sex and Cross-Sex Friendships
In the evolutionary psychology perspective on friendship choices, it has been theorized that cross-sex friendships are formed and maintained if a certain motivation exists to do so, namely, if such friendships contribute to fulfilling needs and goals, such as mating and protection, thus enhancing the likelihood of reproduction and survival (Lewis et al., 2011; Tooby & DeVore, 1987). These goals and needs may differ dispositionally between women and men. For instance, it has been argued that men seek cross-sex friendships as potential mating opportunities, whereas women seek cross-sex friendships for protection from aggressive males who want to mate (Lewis et al., 2012). Indeed, Bleske-Rechek and Buss (2001) found that men judged sexual attraction and sexual desire as important in their cross-sex friendships, whereas women judged physical protection as more central for seeking and building cross-sex friendships. Thus, friendship choices may be better understood if viewed from the perspective of the goals they are aimed at fulfilling. To do so, previous studies have used qualities of friendships and individual values as representations of enduring goals, as detailed in the following.
Baumgarte and Nelson (2009) used a two-step process. First, they asked participants to list the qualities they valued in their male and female friends and then used these responses to construct a standardized questionnaire of 10 qualities (e.g., closeness, agreeable/easygoing, and common interests). Second, applying this questionnaire to a new sample, the authors found that both women and men who reported preferring same-sex friendships reported that these friendships were closer, more trusting, and included more shared interests compared with their cross-sex friendships. Men’s same-sex friendships were more agreeable/easygoing, whereas women’s same-sex friendships were more caring/supportive. The authors also found further sex differences, such as that male friends were preferred for doing activities, whereas female friends were preferred for holding conversations. These findings indicate that the qualities that are valued vary between same-sex and cross-sex friends as well as between women and men. However, it remains unknown whether these inductively created qualities exhaustively describe the realm of friendship qualities or even capture the most relevant ones. A theoretical framework on this issue is missing. Second, if Person A values a certain quality in Person B, it is difficult to guess which value(s) this quality corresponds to or potentially fulfills in Person A.
Chesney and colleagues (2016) explicitly focused on values they defined as “beliefs about desirable outcomes” (p. 691). In particular, they used the concept of social value orientation on avatar-mediated friendship. Participants each controlled a customized human avatar inside a virtual city in which they could meet, and if they chose to, they could befriend others. The authors found that participants with an individualistic social value orientation had larger numbers of friendship requests and mutually confirmed friendships compared with participants with prosocial or competitive value orientations. These effects were larger and more consistent than those of the Big Five. Chesney and colleagues concluded that values might be more important in friendship processes. However, first, the replicability of these findings in real-life interactions is unknown. Second, the authors did not differentiate between same-sex and cross-sex friendships or between women and men in their analyses. Third, social value orientation is a very narrow concept and has been assessed using only three categories (prosocial, competitive, and individualistic) that were based on a simplified resource allocation task. Again, a reference to a comprehensive framework of values is missing.
Personal Values: Conceptualization and Relationships to Friendship
A well-studied framework of a broad spectrum of human values was presented by Schwartz (e.g., 1992), who defined his concept of values on several occasions (e.g., Schwartz, 1992, 2012, 2015) as representations of broad motivational goals (i.e., desirable outcomes) that express what is generally important to people. He identified 10 personal values and clustered them into four value groups: openness to change includes the values of hedonism, stimulation, and self-direction; conservation includes tradition, conformity, and security; self-transcendence includes universalism and benevolence; self-enhancement includes power and achievement. These values and groups can be organized in a circular formation (e.g., Schwartz, 2012), with the values in each group potentially conflicting with the values in the juxtaposed group (Schwartz, 2015). Specifically, openness to change may conflict with conservation: Values emphasizing independence and willingness to change conflict with values emphasizing interdependence and preservation of the past. Similarly, self-enhancement may conflict with self-transcendence: Values emphasizing concern for the welfare and interests of others conflict with values emphasizing the pursuit of one’s own interests. In accordance with this concept, the values in each group are typically found to be negatively correlated with all the values from the respective juxtaposed group (Schwartz et al., 2001). Further evidence of the juxtaposition has been provided in studies relating personal values to the Big Five and to agency/communion (see Table E1 in the Electronic Supplementary Material, ESM 1).
Looking at the 10 values in relation to interpersonal behavior, Schwartz claimed that personal values serve as trans-situational guiding principles in people’s daily lives and are used as criteria for evaluating and judging not only actions and events but also people. Therefore, the values can be hypothesized to be associated with interpersonal behavior as well, such as the choices people make in forging their social networks. Indeed, Solomon and Knafo (2010) administered the Personal Value Questionnaire (Schwartz et al., 2001) to a sample of adolescents and found evidence for value similarity among friends compared with nonfriends, especially with respect to self-direction, conformity, and power. This finding supports the claim that values affect the formation of social relationships (Chesney et al., 2016). However, Solomon and Knafo (2010) focused on only value similarity and did not distinguish between the sexes. In general, the available research on personal values in relation to social relationships is scarce and has not been previously studied in relation to same-sex or cross-sex friendship choices.
Approach of the Present Study
The purpose of the present study is to explore the relationships between interindividual differences in personal values and interindividual differences in the preference for cross-sex friendships (i.e., heterosociality). From the literature described above, expectations about at least some of these relationships can be derived as follows (clustered by value group).
People with higher self-enhancement values (i.e., achievement and power) can be expected to prefer to seek the company and friendship of people in positions of influence and power who control the desirable resources (Lewis et al., 2011; Magee & Galinsky, 2008). In most of the current societal and economical structures, men typically occupy the positions of power (e.g., in companies, associations, or higher education; Bertrand, 2018). Therefore, people of both sexes with higher self-enhancement values can be expected to prefer friendships with men (i.e., cross-sex friendships for women and same-sex friendships for men).
By the same token and with respect to the juxtaposed value group, for people with higher self-transcendence values (i.e., benevolence and universalism), the opposite tendency can be expected. Women dominate in social and supportive positions; therefore, people of both sexes with higher self-transcendence values can be expected to prefer friendships with women (i.e., same-sex friendships for women and cross-sex friendships for men). However, research on this topic is lacking, so these expectations are based on only one possible interpretation of these personal values.
People with higher conservation values (i.e., conformity, tradition, and security) are described as feeling more strongly obliged to follow cultural customs and ideas (Schwartz, 2015). They can be expected to prefer the same-sex friendships that have been more common in society in the past and that are therefore the more traditional type (Bleske-Rechek et al., 2012). This would then lead to higher rates of same-sex friendships in both women and men. With regard to security, Lewis et al. (2011, 2015) claimed that women might seek men as cross-sex friends for protection. Similarly, Bleske-Rechek and Buss (2001) found that women judged physical protection as more central for seeking and building cross-sex friendships. Accordingly, women (but not men) with high-security values were expected to prefer to be friends with men and, therefore, to prefer cross-sex friendships.
People with higher openness to change values (i.e., hedonism, stimulation, and self-direction) can be expected to behave opposite to those with high conservation and tradition values (juxtaposed value groups). Considering that most people tend to seek same-sex friends (Baumgarte & Nelson, 2009; Reeder, 2003) and considering that most children tend to have same-sex friends exclusively (Smith & Inder, 1990), people with a higher orientation toward change, toward being stimulated by new experiences, and toward making decisions by themselves might prefer cross-sex friends as the more stimulating and change-invoking choice compared with the traditional choice of same-sex friendships.
To sum up, the following Hypotheses (H) can be posited:
The self-enhancement values of achievement and power are positively related to heterosociality in women and negatively related to heterosociality in men.
The self-transcendence values of benevolence and universalism are negatively related to heterosociality in women and positively related to heterosociality in men.
The conservation values of conformity and tradition are negatively related to heterosociality in both women and men.
The conservation value of security is positively related to heterosociality in women.
The openness to change values of hedonism, self-direction, and stimulation are positively related to heterosociality in both women and men.
Method
Sample and Procedure
The participants were recruited online and on-campus using print and digital flyers, email newsletters, email forwarding, and posts on social media websites. Data were collected using the online assessment tool Unipark by Tivian (https://www.unipark.com), which is an online assessment software developed for academic purposes. The initial sample consisted of 1,425 participants, of which 92 were excluded for providing incomplete data. The final sample consisted of 1,333 German participants (68.3% women) who completed the heterosociality measure and the personal values questionnaire. Their ages ranged from 18 to 77 years, with a mean of 28.7 (SD = 10.6). For their highest level of education, 12.1% reported that they attended school for 8–10 years, 57.8% reported that they attended school for 12–13 years, and 29.9% reported that they had a university degree; 53.9% were currently students, 41.9% were employed, and 3.8% were unemployed or retired. All assessments were conducted in accordance with local and national ethical guidelines.
Measures
Heterosociality
Prior to assessing heterosociality, participants were provided with a definition of friendship that was to be used across the present survey. Friends were defined as people they see regularly and share intimate or at least private information with. Participants were asked to exclude relatives, friends they knew only online, and sex partners. They were then instructed to write down the names (pseudonymized; e.g., abbreviations or nicknames) of as many friends as they could think of after applying the above definition. They were then instructed to count and report the numbers of women and men on the list. Individual heterosociality scores were calculated as described in previous research (Altmann & Roth, 2020; Kalmijn, 2002) by dividing the number of cross-sex friends by the total number of friends for each participant. Thus, scores could range from 0 to 1.
Personal Values
The German version (Schmidt et al., 2007) of the 40-item Portraits Value Questionnaire (PVQ; Schwartz et al., 2001) was used. It measures 10 personal values based on the theory of basic human values as conceptualized by Schwartz (1992) described above. Cronbach’s αs with 95% confidence intervals were .82 [.79, .84] for Achievement, .58 [.53, .63] for Benevolence, .61 [.56, .66] for Conformity, .70 [.66, .74] for Hedonism, .67 [.63, .71] for Power, .62 [.58, .67] for Security, .50 [.43, .55] for Self-direction, .67 [.63, .71] for Stimulation, .56 [.51, .61] for Tradition, and .73 [.69, .76] for Universalism. Low Cronbach’s α scores are typical for the PVQ. However, it has also been found to have good test-retest reliability despite its lack of internal consistency (Schwartz et al., 2010; see Discussion). Participants rated each item on a 5-point rating scale.
Results
To provide an overview of the data, Table 1 presents the ranges, means, and standard deviations of the study variables (correlations between the 10 personal values are presented in Table E1 in ESM 1). In all subsequent calculations (using the R package “lavaan” by Rosseel, 2012), mean-centered predictors and age and relationship status as control variables were used.
First, it was tested whether the associations between personal values and heterosociality were moderated by sex. To do so, two models were compared: an unrestricted model in which all regression coefficients were allowed to vary between the two sexes and a restricted model in which all regression coefficients were set to equality between the two sexes. Model comparison yielded a just nonsignificant difference, χ2(13) = 21.87, p = .057. This result does not indicate that sex generally moderates all the relationships between personal values and heterosociality. However, sex may moderate most of the relationships to some extent or some of the relationships more substantially.
To explore our expectation about sex differences in the relationships, the coefficients of the unrestricted model in Table 2 for women and men are reported. In the female sample, security and tradition were significant negative predictors, and self-direction was a significant positive predictor of heterosociality. In the male sample, only tradition was a significant negative predictor. Achievement and hedonism both showed effect sizes that were comparable to the one for the traditional value but were just nonsignificant with p-values of .066 and .064.
These findings indicate that women and men appeared to be similar in that higher scores on tradition were associated with lower scores on heterosociality, meaning higher proportions of same-sex friends. They were also similar in that the majority of the values were not associated with their friendship choices. The sexes differed in that, for women, higher scores on security and lower scores on self-direction were associated with lower scores on heterosociality, meaning higher proportions of same-sex friends, but those values were irrelevant for men. Two secondary findings were that heterosociality was positively predicted by, first, women’s relationship status but not men’s and, second, men’s age but not women’s. In general, all effect sizes have to be evaluated as small effects.
Discussion
The present study sought to explore whether personal values were associated with individual preferences for same-sex or cross-sex friendships (homo- or heterosociality) and whether these associations differed between women and men. Despite the theoretical potential of personal values to influence practically all aspects of individual behavior in everyday life (Chesney et al., 2016; Schwartz, 1992, 2015), only partial support for the hypotheses was found.
Interpretation of Main Findings
In H1 and H2, self-enhancement values (achievement and power) and self-transcendence values (benevolence and universalism) were expected to be differentially related to heterosociality in women and men. However, the associations were found to be nonsignificant. It is possible that achievement and power play stronger roles in seeking the friendship of men only in more competitive fields, such as in the financial business sector or in the military, compared with the predominantly social fields in which the present study was conducted and from which the participants were recruited (e.g., psychology, social work, education). Alternatively, people valuing achievement and power might also simply seek as many friends as possible (achievement: having many friends; power: being able to influence many people) independent of their sex.
H3 about the conservation values of conformity and tradition was supported. Both sexes showed a negative association with heterosociality, indicating that both women and men who placed high values on conformity and tradition had higher rates of same-sex friendships. These values involve restraining one’s actions and impulses when one might be likely to violate social expectations or norms and accepting and following the customs and behavioral norms of one’s culture (Schwartz, 2015). People high in these values share the goal of subordinating the self in favor of socially imposed expectations. Same-sex friendships are a comparably more common (Reeder, 2003) and also more traditional (Bleske-Rechek et al., 2012) type of friendship. People with a stronger orientation toward compliance with social norms might therefore tend to avoid the social challenges of cross-sex friendships common in society (O’Meara, 1989) and stick to the traditional choice.
With regard to H4 about the conservation value of security in women, a negative association was found for women. Of the two competing concepts included in H4, the finding for women supports the notion that security values are to be interpreted in the context of conservation values and not in the context of evolutionary psychology. The first suggests negative associations for all three conservation values, that is, conformity, tradition, and security alike (Schwartz, 1992, 2015). The second suggests a positive association between security and a preference for cross-sex friendships in women because women’s cross-sex friends could provide protection (Lewis et al., 2011), which would offer an evolutionary advantage.
Finally, partial support was found for the last H5 with regard to the openness to change values of hedonism, stimulation, and self-direction: Self-direction was positively associated with heterosociality in women. This indicates that women who value self-direction, which is defined by autonomy and independence in the thoughts and actions involved in choosing, creating, and exploring (Schwartz, 2015), have stronger preferences for cross-sex friendships. With regard to men, there were no significant associations with openness to change values. For women but not for men, going one’s own way was associated with reaching out to a larger number of cross-sex friends. Self-direction is part of the openness to change values, which are juxtaposed with the conservation values of security, tradition, and conformity. Considering that security and tradition were found to be negatively associated with heterosociality, the positive association with self-direction was plausible. However, the nonsignificant findings for men were surprising. It may be the case that for men, internal dispositions play a minor role in the formation of their friendships in general (Altmann & Roth, 2020).
In conclusion, values partially explained friendship choices with respect to preferences for same-sex and cross-sex friends and did so differentially for women and men. However, considering the exploratory character of the study and the small effect sizes, replications and extensions are needed to further substantiate these findings and to explore other potential associations.
Limitations
The following limitations should be mentioned. First, a convenience sample was used. Half of the sample was composed of students, which does reflect this group’s proportion in the general population. Second, one limitation may lie in the low Cronbach’s α scores found for the PVQ. The problem of low Cronbach’s α scores on the PVQ appears to be a typical finding (e.g., Schwartz et al., 2010) and also seems to apply to measures of moral cognition in general, as has been found and discussed by Kunnari et al. (2020). However, low Cronbach’s α scores are to be expected when broad and complex constructs (e.g., values) are assessed with only a few items. Such measures can still achieve sufficiently high test-retest reliability scores when the items assess the essential and stable aspects of the construct. Schwartz et al. (2010) were able to show that this applies to the PVQ.
Electronic Supplementary Material
The electronic supplementary material is available with the online version of the article at https://doi.org/10.1027/1614-0001/a000367
References
(2020). Distinctions in friendship research: Variations in the relations between friendship and the Big Five. Personality and Individual Differences, 154, Article 109727. https://doi.org/10.1016/j.paid.2019.109727
(2020). Individual differences in the preference for cross-sex friendship (heterosociality) in relation to personality. Personality and Individual Differences, 157, Article 109838. https://doi.org/10.1016/j.paid.2020.109838
(2009). Preference for same- versus cross-sex friendships. Journal of Applied Social Psychology, 39(4), 901–917. https://doi.org/10.1111/j.1559-1816.2009.00465.x
(2018). Coase lecture: The glass ceiling. Economica, 85(338), 205–231. https://doi.org/10.1111/ecca.12264
(2001). Opposite-sex friendship: Sex differences and similarities in initiation, selection, and dissolution. Personality and Social Psychology Bulletin, 27(10), 1310–1323. https://doi.org/10.1177/01461672012710007
(2012). Benefit or burden? Attraction in cross-sex friendship. Journal of Social and Personal Relationships, 29(5), 569–596. https://doi.org/10.1177/0265407512443611
(1997). The sex of one’s friends. Psychological Reports, 80(3, Pt 2), 1138. https://doi.org/10.2466/pr0.1997.80.3c.1138
(2016). How user personality and social value orientation influence avatar-mediated friendship. Information Technology & People, 29(4), 688–716. https://doi.org/10.1108/itp-10-2014-0242
(2015). Friendship and happiness: Across the life-span and cultures. Springer.
(2007). I am so happy cause today I found my friend: Friendship and personality as predictors of happiness. Journal of Happiness Studies, 8(2), 181–211. https://doi.org/10.1007/s10902-006-9012-7
(2005). Effect of social networks on 10 year survival in very old Australians: The Australian longitudinal study of aging. Journal of Epidemiology & Community Health, 59(7), 574–579. https://doi.org/10.1136/jech.2004.025429
(2002). Sex segregation of friendship networks: Individual and structural determinants of having cross-sex friends. European Sociological Review, 18(1), 101–117. https://doi.org/10.1093/esr/18.1.101
(2020). Challenges in process dissociation measures for moral cognition. Frontiers in Psychology, 11, Article 559934. https://doi.org/10.3389/fpsyg.2020.559934
(2016). The company you keep: Personality and friendship characteristics. Social Psychological and Personality Science, 8(1), 66–73. https://doi.org/10.1177/1948550616662126
(2015).
Friends and happiness: An evolutionary perspective on friendship . In M. DemirEd., Friendship and happiness: Across the life-span and cultures (pp. 37–57). Springer.(2012). Friends with benefits II: Mating activation in opposite-sex friendships as a function of sociosexual orientation and relationship status. Personality and Individual Differences, 53(5), 622–628. https://doi.org/10.1016/j.paid.2012.04.040
(2011). Friends with benefits: The evolved psychology of same- and opposite-sex friendship. Evolutionary Psychology, 9(4), 543–563. https://doi.org/10.1177/147470491100900407
(2014). The Five-Factor Model of personality and degree and transitivity of Facebook social networks. Journal of Research in Personality, 50, 98–101. https://doi.org/10.1016/j.jrp.2014.03.009
(2008). Social hierarchy: The self‐reinforcing nature of power and status. The Academy of Management Annals, 2(1), 351–398. https://doi.org/10.1080/19416520802211628
(1990).
Network diversity, substructures and opportunities for contact . In C. CalhounM. MeyerR. S. ScottEds., Structures of power and constraint (pp. 397–410). Cambridge University Press.(1989). Cross-sex friendship: Four basic challenges of an ignored relationship. Sex Roles, 21(7–8), 525–543. https://doi.org/10.1007/BF00289102
(2003). The effect of gender role orientation on same- and cross-sex friendship formation. Sex Roles, 49(3–4), 143–152. https://doi.org/10.1023/a:1024408913880
(2016). The Big Five personality factors and personal values. Personality and Social Psychology Bulletin, 28(6), 789–801. https://doi.org/10.1177/0146167202289008
(2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. http://www.jstatsoft.org/v48/i02/
(2007). Die Messung von Werten mit dem “Portraits Value Questionnaire”
[The measurement of values with the “Portrait Value Questionnaire”] . Zeitschrift für Sozialpsychologie, 38(4), 261–275. https://doi.org/10.1024/0044-3514.38.4.261(1992).
Universals in the content and structure of values: Theory and empirical tests in 20 countries . In M. ZannaEd., Advances in experimental social psychology (Vol. 25, pp. 1–65). Academic Press.(2012). An overview of the Schwartz Theory of basic values. Online Readings in Psychology and Culture, 2(1). https://doi.org/10.9707/2307-0919.1116
(2015).
Basic individual values: Sources and consequences . In D. SanderT. BroschEds., Handbook of value. Oxford University Press.(2010). Basic personal values, core political values, and voting: A longitudinal analysis. Political Psychology, 31(3), 421–452. https://doi.org/10.1111/j.1467-9221.2010.00764.x
(2001). Extending the cross-cultural validity of the theory of basic human values with a different method of measurement. Journal of Cross-Cultural Psychology, 32(5), 519–542. https://doi.org/10.1177/0022022101032005001
(1990). The relationship of classroom organisation to cross-age and cross-sex friendships. Educational Psychology, 10(2), 127–140. https://doi.org/10.1080/0144341900100202
(2010).
Value similarity in adolescent friendships . In J. C. TollerEd., Friendships: Types, cultural, psychological and social aspects (pp. 197–218). Nova Science.(1987).
The reconstruction of hominid behavioral evolution through strategic modeling . In W. G. KinzeyEd., The evolution of human behavior: Primate models (pp. 183–237). SUNY Press.(2016). Friendship, social interaction, and coping with stress. Communication Research, 19(2), 193–211. https://doi.org/10.1177/009365092019002004
