Social identity support and friendship outcomes: A ... - SAGE Journals

13 downloads 42332 Views 122KB Size Report
Social identity support and friendship outcomes: A longitudinal study predicting who will be friends and best friends 4 years later. Carolyn Weisz & Lisa F. Wood.
07 weisz (ds)

18/5/05

3:24 pm

Page 416

Social identity support and friendship outcomes: A longitudinal study predicting who will be friends and best friends 4 years later Carolyn Weisz & Lisa F. Wood University of Puget Sound

ABSTRACT

This longitudinal study tested the hypothesis that social identity support would predict the long-term status of firstyear college friendships. Social identity support refers to perceived support for valued aspects of the self that are related to identification with social groups, categories, and roles. First-year undergraduates (N = 100) reported levels of closeness, contact, general support, and social identity support for new same-sex friendships. Logistic regression analyses indicated, as expected, that social identity support predicted whether or not a new friend would be a best friend 4 years later, after controlling for initial levels of closeness, contact, and general relationship-specific social support. Social identity support, however, did not predict status as a mere friend at follow-up. The results regarding best friend status suggest that social identity support may be a unique form of social support that contributes to the development of close relationships. KEY WORDS:

best friends • friendship • friendship maintenance • relationship longevity • social identity • social support

This study was supported by a grant from the University of Puget Sound Enrichment Committee. We gratefully acknowledge the help of our research assistants Erin Abrahams, Jennifer Binsfield, Christy Curtis, Tory Gildred, Brett Howard, Elizabeth Humphrey, Nathan Jarrett, Laura Linnell, Daniel Morelli, Amity Neumeister, Toni Schlener, and Subrina Uppal. All correspondence concerning this article should be addressed to Carolyn Weisz, Department of Psychology, University of Puget Sound, 1500 North Warner CMB #1046, Tacoma, WA 98416, USA [e-mail: [email protected]]. Stanley O. Gaines was the Action Editor on this article. Journal of Social and Personal Relationships Copyright © 2005 SAGE Publications (www.sagepublications.com), Vol. 22(3): 416–432. DOI: 10.1177/0265407505052444

07 weisz (ds)

18/5/05

3:24 pm

Page 417

Weisz & Wood: Identity and friendships

417

Same-sex friendships provide a central vehicle for personal contact, companionship, help, and emotional support across the lifespan (Adams, 1986; Bukowski, Hoza, & Boivin, 1994; Hartup & Stevens, 1997; Heller & Lakey, 1985; Roberto & Kimboko, 1989; Rook, 1987). Social science researchers have demonstrated the importance of friends for overall psychological adjustment (Syrotuick & D’Arcy, 1984) and for coping with major life stressors and transitions (Coyne & DeLongis, 1986; Cutrona & Russell, 1990; Rawlins, 1994; Sarason & Sarason, 1985). Studies have also shown that children’s success in forming and maintaining friendships is positively related to academic achievement and later success in relationships, and negatively related to mental health problems, including aggression (Doll, 1996; Ginsberg, Gottman, & Parker, 1986; Newcomb & Bagwell, 1995). Thus, it is evident that close, same-sex friendships are significant elements of individuals’ social support networks and are important for psychosocial functioning. Given the critical role of friendships in people’s lives, understanding factors that impact the stability and quality of friendships has practical as well as theoretical value. Our research examines the impact of a particular type of social support within friendships, specifically, the role of social identity support in predicting the fate of new, same-sex, college friendships after 4 years. This research is important because the concept of social identity support fills a gap in the literature on social support, and draws a much-needed connection among literatures on social identity and close relationships. Moreover, longitudinal research on friendships is rare in the literature, and yet advantageous because it allows an examination of changes in relationships over time and avoids the shortcomings associated with retrospective responses. Social identity support We define social identity support (or, simply, identity support) as the degree to which individuals believe that another person understands, accepts, or provides instrumental support for valued social identities (Weisz & Wood, 2000). Social identities are aspects of self-definition related to social groups, categories, and roles (Deaux, 1992, 1993). Within psychology, social identity theorists typically make distinctions between personal identities, which reflect perceptions of one’s unique attributes (e.g., responsible, creative), and social identities, which reflect identification as a member of particular social categories (e.g., gender, ethnicity) or interacting groups (e.g., fraternity) (Brewer & Gardner, 1996; Hogg, 2003; Simon, 1997; Tajfel & Turner, 1979). Within sociology, social identity theorists focus primarily on identities that reflect relationships between individuals in different but complementary roles (e.g., teacher–student) (Stets & Burke, 2000). Given the overall importance of self-categorization, interacting groups, and rolebased relationships in perceptions of self, we have adopted a rather broad definition of social identities, one that includes aspects of people’s selfconcepts related to identification with social categories, groups and roles. Although identity support within friendships can take many forms, we

07 weisz (ds)

418

18/5/05

3:24 pm

Page 418

Journal of Social and Personal Relationships 22(3)

focus primarily on support in the form of accurate understanding of identities, instrumental help in fulfilling identity-related goals, and general perceptions of support for particular identities. Accurate understanding is important because social identities can have different meanings and associations for different individuals (Deaux, 1993). For example, being an artist might be associated with creative expression for one person, and mastery of technical skills or professional recognition for another. When individuals accurately understand each other’s social identities, they may be better able to provide instrumental help and may be perceived as generally more supportive of one another’s identities. Thus, we assume that different forms of social identity support are typically interrelated. We also assume that the importance of particular social identities can vary for different individuals, and that people especially seek or appreciate support for identities high in importance. Research by Ethier and Deaux (1994) demonstrated a relationship between identity importance and identity-bolstering behavior by showing that the importance of college students’ Hispanic identity predicted the degree to which they became involved with Hispanic activities on campus. In the present study, we operationalized identity support in terms of perceived support for identities that are rated as important to a particular individual. We did this acknowledging that importance may be just one of various dimensions of social identities (e.g., evaluation, social embeddedness, ease of exit) that can influence processes linking social identities to behavioral, psychological, and social outcomes (Ashmore, Deaux, & McLaughin-Volpe, 2004; Bergami & Bagozzi, 2000; Branscombe, Ellemers, Spears, & Bertjan, 1999; Burke & Reitzes, 1991; Ellemers, Kortekaas, & Ouwerkerk, 1999; Rosenberg, 1979; Serpe & Stryker, 1987; Stryker & Serpe, 1982; Thoits, 1991, 1992). Comparing general social support and identity support Our research adds to the broader literature connecting social support with psychosocial and physiological outcomes. Previous studies have linked the amount and quality of support available within people’s social networks to coping and health-related variables (e.g., Barrera, 1986; House, Landis, & Umberson, 1988; Peirce, Frone, Russsell, Cooper, & Mudar, 2000; Wortman & Dunkel-Schetter, 1987). Other research has demonstrated an association between support within specific relationships and variables including loneliness (Pierce, Sarason, & Sarason, 1991), depression (Pierce, Sarason, Sarason, Solky-Butzel, & Nagle, 1997), and relationship quality (Brock, Sarason, Sarason, & Pierce, 1996). None of these studies, however, has specifically examined social identity support as a unique form of social support linked to well-being or relationships. We believe that social identity support and general support are likely to be related in that both may entail open communication, understanding, positive affect, and instrumental forms of help. Both may also reflect what Sarason, Shearin, Pierce, and Sarason (1987) describe as a general feeling that one ‘is accepted, loved, and involved in relationships in which communication is open’ (p. 830). However, identity support differs conceptually from

07 weisz (ds)

18/5/05

3:24 pm

Page 419

Weisz & Wood: Identity and friendships

419

more general social support because it reflects more specific forms of understanding and help concerning identity-related self-perceptions, obligations, and goals (Weisz & Wood, 2000). The study reported here tested whether identity support within friendships predicted friendship outcomes independent of the level of general, relationship-specific support people thought they received from friends. Friendship outcomes Friendships throughout the lifespan may vary in their intensity and level of commitment, with closer friendships reflecting higher levels of trust, mutuality, stability, and intimacy (Berndt, 1982; Bukowski et al., 1994; Smollar & Youniss, 1982). In studies comparing an individual’s experience with a friend as opposed to a best friend, researchers have noted that an individual’s closest friendships are perceived as more intimate and more reciprocal with regard to the quality and depth of communication about personal topics (Roberto & Kimboko, 1989) and more reciprocal with regard to commitment and interest level (Hartup & Stevens, 1997). Studies with children indicate that best friends are more readily trusted, in part because there is lower conflict and because closest friends are more likely to transcend relationship conflicts (Berndt, 1982; Bukowski et al., 1994) or to use conflict productively (Hartup & Stevens, 1997). Because the literature suggests that best friendships are qualitatively and quantitatively distinct from other friendships, we examined best-friend status separately from status as a friend in our assessment of relationship outcomes. Identity support as a predictor of friendship outcomes Research suggests that individuals are generally motivated to maintain and protect aspects of self related to their identification with social groups, categories, and roles (e.g., Abrams & Hogg, 1990; Burke & Stets, 1999; Deaux, Reid, Mizrahi, & Ethier, 1995; McCall & Simmons, 1978; Tajfel & Turner, 1979). For example, in longitudinal research on social identities and recovery from substance abuse, Weisz (1996) found that individuals reporting higher levels of conflict between valued social identities (e.g., spouse, employee) and substance use at the beginning of treatment, were more likely to be abstinent 3 months later than individuals reporting less conflict. This finding suggests that people change their behaviors and social environments in the service of social identities. In a similar vein, we propose that individuals take action to maintain relationships that support important social identities. Several specific theories linking aspects of self to interpersonal behavior and outcomes provide the foundation for our hypothesis that friendships providing social identity support will develop and last. For example, theories positing people’s need for self-verification (Burke & Stets, 1999; Swann, 1987; Swann, Hixon, & De La Ronde, 1992) or self-consistency (Hogg & Abrams, 1990) might predict that individuals are attracted to or feel committed to others who convey an accurate understanding of their identities or who enable successful role-related behaviors. Similar theories that

07 weisz (ds)

420

18/5/05

3:24 pm

Page 420

Journal of Social and Personal Relationships 22(3)

focus on the desire to reduce self-discrepancies (e.g., Burke, 1991; Higgins, 1989) might predict that individuals will withdraw from relationships with nonsupportive friends who interfere with their ability to meet identityrelated standards and goals. Pragmatic concerns might also cause people to invest in friendships that provide instrumental support for identity-related goals (e.g., help studying, partnership for exercising, transportation). Other theories posit that people derive self-esteem based on the status of their in-groups, and that as a result people show a variety of in-group biases (Abrams & Hogg, 1988; Brewer, 1991; Cialdini, Borden, Walker, Freeman, & Sloan, 1976; Luhtanen & Crocker, 1992; Tajfel, Billig, Bundy, & Flament, 1971; Tajfel & Turner, 1979). As an extension of these theories, we propose that people may boost self-esteem by seeking out friends who affirm their identities as members of groups (e.g., ‘You’re a real Christian.’) or who affirm the status of those groups (e.g., ‘It’s so cool that you play sax for the Stanford Band!’), whether or not these friends share their social identities. In sum, we suggest that social identity support may contribute to the success of friendships as a result of various motivational processes involving self-verification, self-consistency, self-esteem, and pragmatic concerns. Although distinguishing among these processes is beyond the scope of this project, we argue that because of some or all of these processes, social identity support should increase longevity and commitment within friendships. Background and goals of the present study This article reports findings from a 4-year longitudinal study of college students’ same-sex friendships. Data from the first 2 years of the study indicated that identity support was positively correlated with ratings of relationship closeness measured at the same time point, even after controlling for general relationship-specific social support as measured by Pierce’s (1994) Quality of Relationships Inventory (QRI) support scale (Weisz & Wood, 2000). Initial ratings of identity support also predicted levels of closeness with a friend 1 year later, after controlling for initial levels of closeness and QRI support. In the current analysis, we examined the relationship between initial ratings of identity support and the status of friendships 4 years later. All analyses used one friendship per participant. We predicted that social identity support would be positively related to the likelihood of being ‘friends’ and ‘best friends’ at follow-up. We also hypothesized that identity support would predict these friendship outcomes after controlling for general support as measured by the QRI (Pierce, 1994). Closeness and contact were included as covariates in these analyses to control for the fact that relationships differing in identity support might also differ in other relationship characteristics associated with friendship outcomes. Additional exploratory analyses were conducted to better understand the link between identity support and friendship outcomes. Specifically, we tested whether gender interacted with identity support in predicting friendship outcomes and examined the mediating role of contact at follow-up. We

07 weisz (ds)

18/5/05

3:24 pm

Page 421

Weisz & Wood: Identity and friendships

421

also investigated relationships between support for specific identities (e.g., student, artist) and friendship outcomes. Method Participants Random samples of male and female college students in their first year at a private liberal arts university were contacted by telephone for an initial screening. Students aged 18–20 years were invited to participate, and over 90% of students invited completed the initial assessment (50 men and 50 women) for which they were paid $5. Most participants (n = 86) completed a follow-up survey 1 year after the initial assessment for $10, and results from that assessment are reported elsewhere (Weisz & Wood, 2000). Findings reported here come from 80 participants (40 men and 40 women) who completed a telephone interview 4 years after the first assessment. These participants were primarily White (68, 6 Asian American, 1 African American, 5 other or mixed). All were single, without children, and living on campus at the beginning of the study. At the 4-year follow-up, most were unmarried (88%) and were no longer undergraduate students (86%). Procedure During the spring of their first year at college, participants completed a set of instruments in a laboratory setting. Instruments relevant to the current hypotheses measured demographics, social network variables, social identities, social identity support within specific friendships, and general social support within specific friendships. Participants also gave consent to be contacted by the investigators at a later date to provide follow-up information. Four years later, we attempted to contact participants by telephone or e-mail to schedule a brief telephone interview. In the telephone interview, an interviewer blind to the earlier responses of participants asked questions about the current status of the friendships following a script described later. Materials Demographic questionnaire. Participants provided information about their age, gender, ethnicity, family, and living situation. Social network questionnaire. Using a matrix format, this instrument asked participants to list individuals they currently considered to be their friends. For each friend, participants indicated how long they had known the person, whether the friend went to the same college, perceived closeness (1 = not at all close, 5 = extremely close), amount of contact in the past 6 months including phone, mail, e-mail, and in-person contact (1 = no contact, 7 = daily), and gender. Social identity questionnaire. A social identity questionnaire adapted from Weisz (1996) asked two questions about each of 16 identity labels (e.g., student). Participants indicated whether they currently thought of themselves as a ______? (no, somewhat, or yes) and rated how important being a ______ was to them at the current time (1 = not at all, 5 = very important). The set of

07 weisz (ds)

422

18/5/05

3:24 pm

Page 422

Journal of Social and Personal Relationships 22(3)

identities, although not exhaustive, included identities likely to be important to students, and those related to major life domains including school, work, hobbies, and social relationships. The identities were: (1) dating partner, romantic partner or spouse; (2) parent; (3) other family member (daughter, brother, aunt, etc.); (4) friend; (5) student; (6) employee/worker; (7) person with homemaking skills; (8) person skilled in practical things; (9) creative person (artist, musician, writer, dancer, etc.); (10) athlete/person who exercises; (11) man or woman (gender identity); (12) leader of an organization or group; (13) member of a sorority or fraternity; (14) member of an ethnic or cultural group; (15) member of a religious organization or group; and (16) member of another type of group, club, organization, or category (specify). The identities appeared in the same order for all participants. For conceptual reasons, we excluded all data related to being a homemaker or person skilled at practical things (we questioned whether these items could be considered social identities). We also excluded data related to the parent identity because no participants reported having children at the initial assessment and none indicated that the identity described them. Social identity support questionnaire. Participants identified their three closest, new, nonromantic, same-sex friends at their university from their social network questionnaire. They completed a support questionnaire for each friend in an order determined at random by the experimenter. For each of the identities listed earlier, participants rated how much the friend supported (How much does this person support you in being the kind of ______ you want to be?), understood (To what extent does this person understand what it means to you to fulfill each of these roles or identities?), and provided instrumental support for (To what degree does this person actually provide you with concrete help related to being the kind of ______ that you want to be?) the identity (1 = not at all, 5 = an extreme amount). Participants could also indicate that a question was not applicable, which occurred primarily when participants did not endorse a particular identity label as descriptive (e.g., parent). To create a single identity support variable for each friendship, we first determined which identities were rated as important by each participant. We defined an important identity as one that a participant rated both as descriptive (a yes response) and as important or very important (at least a 4 on the 5-point scale). For each of the three identity support items (general, understanding, and instrumental support), we took the mean score across a participant’s important identities. In a sample using one randomly selected friendship per participant, these three indexes were highly related (Cronbach’s alpha = .91), so we created a single identity support index for each friendship by taking the average of the indexes for the three support items. General social support. For each friend, participants also completed the 7-item QRI support subscale (Pierce, 1994), which measures general relationshipspecific support using a 4-point scale (e.g., To what extent could you count on this person for help with a problem? 1 = not at all, 4 = very much). A general support index was calculated for each friendship by taking the mean of the QRI support items (Cronbach’s alpha = .84).

07 weisz (ds)

18/5/05

3:24 pm

Page 423

Weisz & Wood: Identity and friendships

423

Telephone follow-up script. The interviewer reminded participants of the names of the three individuals whose support they had rated at Time 1 (i.e., their three closest, new, same-sex friends at the initial assessment). For each of these individuals, the interviewer asked two dichotomous questions: Do you currently consider this person to be a friend? (yes or no), and Do you currently consider this person to be among your best friends? (yes or no). The interviewer also asked participants how much contact they had on average with the person over the last 3 months, and described contact as including in-person, telephone, or e-mail contact, but not attending events with someone without socializing with that person. The researcher described the following scale and coded participants’ responses: daily (7), a few times a week (6), once a week (5), every other week (4), once a month (3), less often (2), not at all (1). Participants also provided demographic information.

Results Overview of analyses Data for a particular friendship were considered missing if participants indicated that they had known the individual more than a year at the initial assessment (23 friendships), if the friend was no longer a student at the university at the 1-year follow-up assessment (13 friendships), or if follow-up data were not provided (6 friendships). Because each participant’s responses concerning three different friendships constituted nonindependent observations, we conducted all analyses using data from one randomly selected friendship per participant. Descriptive statistics for study measures appear in Table 1. To test our primary hypotheses, separate logistic regression models were used to predict status as a friend and status as a best friend at follow-up. Biserial correlations indicated that social identity support was correlated with closeness, r(78) = .57, p < .001, and contact, r(78) = .24, p < .05, at the initial assessment. Thus, in all logistic regression analyses, we included ratings of closeness and contact at Time 1 as covariates. Additional analyses exploring the relationship

TABLE 1 Means, standard deviations, and frequencies for study variables using one randomly selected friendship per participant (n = 80) Measure Initial assessment Closeness Contact General support index (QRI) Identity support index Follow-up Contact Currently a friend Currently a best friend

M

SD

3.41 6.76 3.11 3.47

1.14 .56 .60 .84

2.49

1.68

%) Frequency (%

60 (75) 24 (30)

07 weisz (ds)

18/5/05

3:24 pm

Page 424

424

Journal of Social and Personal Relationships 22(3)

between support for specific identities and friendship outcomes were conducted using independent-groups t-tests. Alpha was set at .05 for all tests. Predicting status as a friend and as a best friend To predict best-friend status (0 = not a best friend, 1 = best friend), we first entered closeness and contact as one block in a logistic regression model, and then entered the identity support index as a separate block. As predicted, the identity support index was positively related to being a best friend at follow-up after controlling for initial closeness and contact, Wald chi-square = 6.53, p = .01, accounting for a change in Nagelkerke R2 of .12. A similar analysis used to predict status as a friend (0 = not a friend, 1 = friend) found that the identity support index was not a statistically significant predictor of later friendship status, Wald chi-square = 3.41, p = .07, change in Nagelkerke R2 = .06. Table 2 presents odds ratios, 95% confidence intervals, and Nagelkerke estimates of R2 for these analyses. The indexes of identity support and general support were positively correlated at the initial assessment, r(78) = .68, p < .001. To test whether identity support was an independent predictor of outcomes after controlling for general support, we entered initial ratings of closeness, contact, and general relationship-specific support as one block in the model, and then added social identity support as a second block. After controlling for closeness, contact, and general support, identity support was a statistically significant predictor of best-friend status, Wald chi-square = 5.50, p = .02, accounting for a change in Nagelkerke R2 of .10 (see Table 3). A similar analysis with status as a friend as the criterion variable indicated that identity support was not a statistically significant predictor of whether people were still friends at follow-up, Wald chi-square = 1.45, p = .23, change in Nagelkerke R2 = .03. For exploratory purposes, we repeated the four analyses described earlier including gender and identity support in the first block and only a gender-byidentity support interaction term in the second block. The interaction term was not significant in any analysis, p-values > .10, suggesting that the relationships between identity support and friendship outcomes did not depend on gender.

TABLE 2 Odds ratios and 95% confidence intervals for predicting status as a friend and best friend (1 = friend/best friend, 0 = not a friend/not a best friend) (n = 80) Status as a friend Step/Predictor Step 1 Closeness Contact Step 2 Closeness Contact Identity support

Status as a best friend

Odds ratio

Confidence Nagelkerke Odds Confidence Nagelkerke ratio interval R2 interval R2

1.61 0.59

(0.98–2.66) (0.21–1.66)

1.20 0.48 2.16

(0.67–2.15) (0.17–1.38) (0.95–4.91)

.069

.124 1.95* (1.13–3.34) 0.81 (0.28–2.35)

.131

*p < .05 for Wald chi-square test.

.243 1.33 (0.71–2.49) 0.59 (0.18–2.00) 3.41* (1.33–8.76)

07 weisz (ds)

18/5/05

3:24 pm

Page 425

Weisz & Wood: Identity and friendships

425

TABLE 3 Odds ratios and 95% confidence intervals for predicting status as a friend and best friend (1 = friend/best friend, 0 = not a friend/not a best friend) controlling for general QRI Support (n = 80) Status as a friend Step/Predictor Step 1 Closeness Contact QRI support Step 2 Closeness Contact QRI support Identity support

Status as a best friend

Odds ratio

Confidence Nagelkerke Odds Confidence Nagelkerke interval R2 ratio interval R2

1.03 0.57 3.44

(0.52–2.04) (0.21–1.51) (0.93–12.75)

0.99 0.49 2.37 1.73

(0.48–1.91) (0.18–1.36) (0.58–9.78) (0.71–4.22)

.131

.146 1.50 0.81 2.11

(0.75–2.99) (0.26–2.52) (0.61–7.37)

1.35 0.59 0.96 3.45*

(0.63–2.90) (0.17–2.02) (0.21–4.36) (1.23–9.73)

.156

.243

*p < .05 for Wald chi-square test.

We also considered that the amount of contact at follow-up, which was positively correlated with the identity support index, r(78) = .31, p = .007, might mediate the relationship between identity support and later best-friend status. To examine this, we conducted a logistic regression analysis in which we entered contact at follow-up along with initial closeness, contact, and general support as one block, and the identity support index as a separate block. The predictive role of identity support remained significant, suggesting that level of contact at follow-up did not account for the relationship between identity support and whether or not participants reported someone to be a best friend, Wald chi-square = 4.60, p = .03, change in Nagelkerke R2 = .08. Support for specific identities as a predictor of friendship outcomes An advantage of calculating the mean rather than the sum of support scores across important identities is that the identity support index was not confounded with a participant’s total number of important identities, r(78) = .05, p = .67. One limitation of using an index aggregated across identities, however, is that it is not clear whether support for particular identities was primarily responsible for observed effects. To explore this possibility, we used independent groups t-tests with best-friend status at follow-up as the independent variable and initial support for particular important identities as dependent variables. We calculated a separate identity support score for each of a participant’s important identities by averaging responses to the three identity support items, again using one randomly selected friendship per participant. Cronbach’s alphas ranged from .57 to .94 (ns = 7–78 depending on how many participants rated each identity as important). We conducted the test for 11 identities that at least 10 participants indicated were descriptive and important (fewer than 10 participants considered identities related to ethnicity and leadership to be important). Table 4 shows differences between mean identity support scores for best and not best friendships.

07 weisz (ds)

18/5/05

3:24 pm

426

Page 426

Journal of Social and Personal Relationships 22(3)

TABLE 4 Differences between best and not best friendships in initial levels of support for specific valued identities Best friend Identity

n

Romantic partner Family member Friend Student Employee/worker Creative person/artist Athlete/exercise Gender Leader Sorority/fraternity Ethnic/cultural group Religious group Other identity

6 19 24 21 3 7 12 18 4 5 3 4 9

M (SD) 4.00 3.47 4.35 4.03 3.33 4.33 3.81 3.97 3.67 4.67 3.78 3.92 3.67

(1.07) (.90) (.66) (.63) (.88) (.96) (1.20) (1.03) (1.12) (.47) (.38) (1.32) (1.29)

Not a best friend n

M (SD)

t

18 37 54 50 10 16 23 35 3 5 5 13 11

2.98 (1.30) 2.71 (1.02) 3.66 (.97) 3.66 (.90) 2.40 (.98) 3.21 (1.21) 3.23 (1.12) 3.27 (1.19) 3.11 (1.02) 4.13 (.80) 2.40 (1.01) 2.09 (1.07) 3.18 (1.34)

1.73 2.74** 3.16** 1.72 1.47 2.17* 1.40 2.12* – 1.28 – 2.85* 0.82

Note. Sample sizes reflect best-friend status as well as the number of participants rating each identity as descriptive and important.

*p < .05; **p < .01.

In all cases, identity support was higher for relationships that were later identified as best rather than not best friendships, and differences were statistically significant for five identities: artist, family member, friend, gender, and religion, p values < .05.

Discussion Findings from this study provide some support for the hypothesis that social identity support predicts long-term outcomes of friendships. Our primary finding was that college students who initially reported higher rather than lower levels of identity support from a new friend were more likely to describe that person as a best friend 4 years later, even after controlling for initial ratings of closeness, contact, and general relationshipspecific social support. However, after controlling for these factors, identity support did not predict whether or not a person was described as merely a friend at the 4-year follow-up. The ability of identity support to predict best-friend status, but not status as a friend, can be explained in several ways. One interpretation is that identity support promotes relationship characteristics such as intimacy, trust, or commitment that are more integral to best friendships than to other friendships (Roberto & Kimboko, 1989). Alternatively, social norms may dictate that best friends support each other’s personal and social goals,

07 weisz (ds)

18/5/05

3:24 pm

Page 427

Weisz & Wood: Identity and friendships

427

and thus may determine how individuals identify or treat their best friends. People’s definitions or expectations of mere friends may be more liberal, requiring only positive contact or liking. Also, fortuitous factors such as where people live after college might have a large influence on who maintains status as a friend, but might have less impact on communication and intimacy within best friendships. Indeed, in our study, identity support predicted best-friend status independent of the amount of contact individuals reported at follow-up. Although we did not attempt to distinguish among various processes that might mediate the observed relationship between identity support and later best-friend status, we offer some speculations about the psychological and interpersonal processes involved. For example, positive affect or commitment within friendships may depend on the role of identity support in boosting identity-based self-esteem (Abrams & Hogg, 1990; Luhtanen & Crocker, 1992), providing self-verification (Burke & Stets, 1999; Swann, 1987), or helping people achieve identity-related goals. Moreover, when individuals believe their friends support their valued identities, they may be more likely to reciprocate support, express commitment, or tolerate disagreement within relationships. Individual differences that determine who people become friends with and how friendships develop could also explain our findings. For example, individuals adept at understanding multiple facets of others and at adapting to changes within relationships may be especially good at providing identity support and also especially likely to form close bonds with others. The processes involved may also depend on the type of social identity. Although some social identities, such as student, can operate interchangeably as role-based, group-based, and category-based identities, others more consistently reflect particular types of identification (Brewer & Gardner, 1996; Sedikides & Brewer, 2001) and may relate in unique ways to identity support. For example, identification with interacting groups may enable or encourage different forms of identity support than identification with larger collectives. It is interesting that the positive relationship between identity support and best-friend status appeared consistently across the identities we measured, and for identities reflecting social category memberships (e.g., gender, religion) and social roles (e.g., friend, romantic partner). This pattern suggests that the processes involved are not limited exclusively to either group-based or role-based identities. Findings regarding support for individual identities should be interpreted extremely tentatively, however, because some analyses had very small n values and we did not attempt to control for initial differences in closeness or contact. Limitations and strengths of the present study Our findings should be interpreted with caution due to several methodological limitations. First, the validity of our measures may be limited because we assessed identity importance and friendship outcomes with single items and asked about a restricted set of identities in a fixed order.

07 weisz (ds)

428

18/5/05

3:24 pm

Page 428

Journal of Social and Personal Relationships 22(3)

Second, it is possible that our identity support instrument simply measured a general positive attitude toward a particular friend. The fact that identity support predicted best-friend status independent of closeness, contact, and general support, however, argues against this explanation. Third, we did not rule out other possible confounds such as similarity and the degree to which friends self-verified characteristics of participants not related to social identities. Regarding similarity in the form of shared identities, we found that for some category affiliations that friends shared (e.g., participants and their new friends were the same gender), the degree of support for specific, corresponding identities (e.g., gender) was related to later best-friend status. This suggests that variations in identity support can predict friendship outcomes even when individuals share the same category membership. Although we assume that support for nonshared identities can also predict relationship outcomes, future research must examine the role of identity support independent of both general similarity and similarity in identities. Our study may also have been limited by the nature of friendships that form at a private, liberal arts university. For example, identity support may have different consequences in environments where friends derive less identity support from the larger social context (i.e., less than that available for students at a small, private university). The role of identity support may also differ in more ethnically diverse populations (Hamm, 2000) or in settings where the possibility of discrimination based on gender, race, or other social characteristics is more salient. Despite these problems, the strengths of this research outweigh its limitations. Using a longitudinal design allowed us to avoid retrospective response biases and to study the predictive role of identity support in the context of shifts in people’s environments, occupations, and social networks. Few existing studies use longitudinal designs to examine the consequences of identity-related variables (e.g., Ethier & Deaux, 1994; Thoits, 1995), and those that do have not focused on the notion of support for social identities within dyadic relationships. Of particular importance was our ability to demonstrate the predictive role of identity support independent of general support. This finding calls attention to a form of interpersonal support that has received little attention in the social support literature and that may influence the quality and development of friendships. Research on identity support is also of theoretical importance because it brings together largely independent literatures on social identity and close relationships. Conclusion and future directions Our findings demonstrate that college students who perceive new friends as supporting their important social identities are more likely to consider those individuals to be best friends (but not merely friends) 4 years later. The predictive role of identity support was reliable even after controlling for initial levels of closeness, contact, and general relationship-specific support within new friendships, and even after controlling for contact at the

07 weisz (ds)

18/5/05

3:24 pm

Page 429

Weisz & Wood: Identity and friendships

429

follow-up assessment. These findings suggest that social identity support may be a unique form of social support that has consequences for the development of close relationships. Our study paves the way for further research addressing issues such as the nature and origins of identity support, the function of identity support during times of developmental transition in social identities, and the role of identity support in other types of relationships (e.g., parent–child, romantic partners). Future research would benefit from methods that allow for the measurement of identity support at the behavioral level (e.g., behavior checklists, diary methods) and that use data from both members of friendship dyads to shed light on the role of reciprocity of identity support over time. Such research would have implications for improving individual well-being and interpersonal relationships, and for understanding the importance of friendships in identity development and change.

REFERENCES Abrams, D., & Hogg, M. A. (1988). Comments on the motivational status of self-esteem in social identity and intergroup discrimination. European Journal of Social Psychology, 18, 317–334. Abrams, D., & Hogg, M. A. (1990). Social identity theory: Constructive and critical advances. London: Harvester-Wheatsheaf. Adams, R. (1986). Secondary friendship networks and psychological well-being among elderly women. Activities, Adaptation and Aging, 35, 59–72. Ashmore, R. D., Deaux, K., & McLaughlin-Volpe, T. (2004). An organizing framework for collective identity: Articulation and significance of multidimensionality. Psychological Bulletin, 130, 80–114. Barrera, M. (1986). Distinctions between social support concepts, measures, and models. American Journal of Community Psychology, 14, 413–445. Bergami, M., & Bagozzi, R. P. (2000). Self-categorization, affective commitment and group self-esteem as distinct aspects of social identity in the organization. British Journal of Social Psychology, 39, 555–577. Berndt, T. J. (1982). Fairness and friendship. In K. H. Rubin & H. S. Ross (Eds.), Peer relationships and social skills in childhood (pp. 253–278). New York: Springer. Branscombe, N. R., Ellemers, N., Spears, R., & Bertjan, D. (1999). The context and content of social identity threats. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity: Context, commitment, content (pp. 35–58). Oxford: Blackwell. Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, 17, 475–482. Brewer, M. B., & Gardner, W. (1996). Who is this ‘we’? Levels of collective identity and selfrepresentation. Journal of Personality and Social Psychology, 71, 83–93. Brock, D. M., Sarason, I. G., Sarason, B. R., & Pierce, G. R. (1996). Simultaneous assessment of perceived global and relationship-specific support. Journal of Personal and Social Relationships, 13, 143–152. Bukowski, W. M., Hoza, B., & Boivin, M. (1994). Measuring friendship quality during preand early adolescence: The development and psychometric properties of the friendship qualities scale. Journal of Social and Personal Relationships, 11, 471–48. Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56, 836–849. Burke, P. J., & Reitzes, D. C. (1991). An identity theory approach to commitment. Social Psychology Quarterly, 54, 239–251.

07 weisz (ds)

430

18/5/05

3:24 pm

Page 430

Journal of Social and Personal Relationships 22(3)

Burke, P. J., & Stets, J. E. (1999). Trust and commitment through self-verification. Social Psychology Quarterly, 62, 347–366. Cialdini, R., Borden, R. A., Walker, T. M., Freeman, S., & Sloan, L. (1976). Basking in reflected glory: Three (football) field studies. Journal of Personality and Social Psychology, 56, 124–131. Coyne, J. C., & DeLongis, A. (1986). Going beyond social support: The role of social relationships in adaptation. Journal of Consulting and Clinical Psychology, 54, 454–460. Cutrona, C. E., & Russell, D. W. (1990). Type of social support and specific stress: Towards a theory of optimal matching. In B. R. Sarason, I. G. Sarason, & G. R. Pierce (Eds.), Social support: An interactional view (pp. 321–366). New York: Wiley. Deaux, K. (1992). Focusing on the self: Challenges to self-definition and their consequences for mental health. In D. N. Ruble, P. R. Costanzo, & M. E. Oliveri (Eds.), The social psychology of mental health (pp. 301–327). New York: Guilford Press. Deaux, K. (1993). Reconstructing social identity. Personality and Social Psychology Bulletin, 19, 4–12. Deaux, K., Reid, A., Mizrahi, K., & Ethier, K. A. (1995). Parameters of social identity. Journal of Personality and Social Psychology, 68, 280–291. Doll, B. (1996) Children without friends: Implications for practice and policy. School Psychology Review, 25, 165–183. Ellemers, N., Kortekaas, P., & Ouwerkerk, J. W. (1999). Self-categorization, commitment to the group and group self-esteem as related but distinct aspects of social identity. European Journal of Social Psychology, 29, 371–389. Ethier, K. A., & Deaux, K. (1994). Negotiating social identity when contexts change: Maintaining identification and responding to threat. Journal of Personality and Social Psychology, 67, 243–251. Ginsberg, D., Gottman, J., & Parker, J. (1986). The importance of friendship. In J. M. Gottman & J. G. Parker (Eds.), Conversations of friends (pp. 1–50). Paris: Cambridge University Press. Hamm, J. V. (2000). Do birds of a feather flock together? The variable bases for African American, Asian American, and European American adolescents selection of similar friends. Developmental Psychology, 36, 209–219. Hartup, W. W., & Stevens, N. (1997). Friendships and adaptation in the life course. Psychological Bulletin, 121, 355–370. Heller, K., & Lakey, B. (1985). Perceived support and social interaction among friends and confidants. In I. G. Sarason & B. R. Sarason (Eds.), Social support, theory research and applications (pp. 287–300). Dordrecht: Martinus Nijhoff. Higgins, E. T. (1989). Self-discrepancy theory: What patterns of self-beliefs cause people to suffer? Advances in Experimental Social Psychology, 22, 93–136. Hogg, M. A. (2003). Social identity. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 462–479). New York: Guilford Press. Hogg, M. A., & Abrams, D. (1990). Social motivation, self-esteem, and social identity. In D. Abrams & M. A. Hogg (Eds.), Social identity theory: Constructive and critical advances (pp. 28–47). New York: Springer. House, J. S., Landis, K. R., & Umberson, D. (1988). Social relationships and health. Science, 241, 540–544. Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318. McCall, G. J., & Simmons, J. L. (1978). Identities and interactions: An examination of human associations in everyday life. New York: The Free Press. Newcomb, A. F., & Bagwell, C. L. (1995). Children’s friendship relations: A meta-analytic review. Psychological Bulletin, 117, 306–347. Peirce, R. S., Frone, M. R., Russell, M., Cooper, M. L., & Mudar, P. (2000). A longitudinal model of social contact, social support, depression, and alcohol use. Health Psychology, 19, 28–38.

07 weisz (ds)

18/5/05

3:24 pm

Page 431

Weisz & Wood: Identity and friendships

431

Pierce, G. R. (1994). The Quality of Relationships Inventory: Assessing the interpersonal context of social support. In B. R. Burleson, T. L. Albrecht, & I. G. Sarason (Eds.), Communication of social support: Messages, interactions, relationships, and community (pp. 247–264). Thousand Oaks, CA: Sage. Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1991). General and relationship-based perceptions of social support: Are two constructs better than one? Journal of Personality and Social Psychology, 61, 1028–1039. Pierce, G. R., Sarason, I. G., Sarason, B. R., Solky-Butzel, J. A., & Nagle, L. C. (1997). Assessing the quality of personal relationships. Journal of Personal and Social Relationships, 14, 339–356. Rawlins, W. K. (1994). Being there and growing apart: Sustaining friendships during adulthood. In D. J. Canary & L. Stafford (Eds.), Communication and relational maintenance (pp. 275–294). New York: Academic Press. Roberto, K. A., & Kimboko, P. J. (1989). Friendships in later life: Definitions and maintenance patterns. International Journal of Aging and Human Development, 28, 9–35. Rook, K. S. (1987). Reciprocity of social exchange and social status among older women. Journal of Personality and Social Psychology, 51, 770–778. Rosenberg, M. (1979). Conceiving the self. New York: Basic Books. Sarason, I. G., & Sarason, B. R. (1985). Life change, social support, coping, and health. In R. M. Kaplan & M. H. Criqui (Eds.), Behavioral epidemiology and disease prevention (pp. 219–236). New York: Plenum Press. Sarason, B. R., Shearin, E. N., Pierce, G. R., & Sarason, I. G. (1987). Interrelationships of social support measures: Theoretical and practical implications. Journal of Personality and Social Psychology, 52, 813–832. Sedikides, C., & Brewer, M. B. (2001). Individual self, relational self, and collective self: Partners, opponents, or strangers? In C. Sedikides & M. B. Brewer (Eds.), Individual self, relational self (pp. 1–4). Philadelphia, PA: Psychology Press. Serpe, R. T., & Stryker, S. (1987). The construction of self and reconstruction of social relationships. In E. Lawler & B. Markovsky (Eds.), Advances in group processes (pp. 41–66). Greenwich, CT: JAI Press. Simon, B. (1997). Self and group in modern society: Ten theses on the individual self and the collective self. In R. Spears, P. J. Oaks, N. Ellemers, & S. A. Haslam (Eds.), The social psychology of stereotyping and group life (pp. 318–335). Cambridge, MA: Blackwell. Smollar, J., & Youniss, J. (1982). Social development through friendship. In K. H. Rubin & H. S. Ross (Eds.), Peer relationships and social skills in childhood (pp. 279–298). New York: Springer. Stets, J. E., & Burke, P. J. (2000). Identity theory and social identity theory. Social Psychology Quarterly, 63, 224–237. Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior: A theory and research example. In W. Ickes & E. S. Knowles (Eds.), Personality, roles, and social behavior (pp. 199–218). New York: Springer. Swann, W. B. (1987). Identity negotiation: Where two roads meet. Journal of Personality and Social Psychology, 53, 1038–1051. Swann, W. B., Hixon, J. G., & De La Ronde, C. (1992). Embracing the bitter ‘Truth’: Negative self-concepts and marital commitment. Psychological Science, 3, 118–121. Syrotuick, J., & D’Arcy, C. (1984). Social support and mental health: Direct, protective, and compensatory effects. Social Science & Medicine, 18, 229–236. Tajfel, H. M., Billig, G., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behavior. European Journal of Social Psychology, 1, 149–178. Tajfel, H. M., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole. Thoits, P. (1991). On merging identity theory and stress research. Social Psychology Quarterly, 54, 101–112.

07 weisz (ds)

432

18/5/05

3:24 pm

Page 432

Journal of Social and Personal Relationships 22(3)

Thoits, P. (1992). Identity structures and psychological well-being: Gender and marital status comparisons. Social Psychology Quarterly, 55, 236–256. Thoits, P. (1995). Identity-relevant events and psychological symptoms: A cautionary tale. Journal of Health and Social Behavior, 36, 72–82. Weisz, C. (1996). Social identities and response to treatment for alcohol and cocaine abuse. Addictive Behaviors, 21, 445–458. Weisz, C., & Wood, L. F. (2000). Social identities and friendships: A longitudinal study of support for social identities. Journal of Social Behavior and Personality, 15, 441–458. Wortman, C. B., & Dunkel-Schetter, C. (1987). Conceptual and methodological issues in the study of social support. In A. Baum & J. E. Singer (Eds.), Handbook of psychology and health: Vol. 5. Stress (pp. 63–108). Hillsdale, NJ: Erlbaum.