Adolescents’ social relationships have been largely interrupted the past few years due to the COVID-19 pandemic. Though research in this area is still emerging, based on your observati

Journal of Youth and Adoleseence, Vol. 26, No. 5, 1997

Adolescent Intimacy Revisited

Shmuel Shuhnan,1 Brett Laurseis,2 Zwi Kahnan,3 and Siga! Karpovsky4 Received August 2, 1995; accepted October 26, 1996

Two studies examined intimacy in adolescent friendships. In the first, 7th-, 9th-, and llth-grade students completed a questionnaire assessing perceived friend- ship intimacy. Age and sex differences were identified in emotional closeness, self-disclosure, emphasis on individuality, control, and conformity. Across ages, emphasis on individuality increased, whereas control and conformity declined. There were no age differences in emotional closeness and self-disclosure. Fe- males reported more emotional closeness and self-disclosure than males. In the second study, individual differences in friendship intimacy were examined in a sample of 9th-grade adolescents. A joint problem solving task identified interdependent and disengaged friends. Perceived intimacy among interdepend- ent and disengaged friends was contrasted with that in a control group of sub- jects without friends. Adolescents with friends reported more closeness than those without friends. Interdependent friends reported greater levels of respect for individuality than disengaged friends. The results underscore the salience of intimacy for peer relationships during the adolescent years and suggest that intimacy may be an important construct distinguishing between different types of close friendships.

INTRODUCTION

Intimate close friendships, found across the life span, first appear dur- ing early adolescence. Developmental studies of friendship intimacy em-

1 Associate Professor, Department of Psychology, Bar-Ilan University, Ramat-Gan 52900, Israel. Received Ph.D. from Bar-Han University. Research interests include close relationships in adolescence in normal and pathological samples. To whom correspondence should be addressed.

2Associate Professor, Department of Psychology, Florida Atlantic University. Received Ph.D. from the University of Minnesota. Research interests include adolescent friendships and conflict in close relationships.

3Graduate student, Tel-Aviv University, Israel. 4Graduate student, Tel-Aviv University, Israel.

S97

0047-239!/97/IOOaOS97SI2^QO @ 1997 Plenum Publishing Corporation

phasize the increasing significance of self-disclosure, closeness, and mutual assistance during the adolescent years (e.g., Jones and Dembo, 1989; Shara- bany et al., 1981). According to Selman (1990), the ability to balance close- ness and individuality heralds a mature form of friendship intimacy that typically does not emerge until adolescence. Building on this empirical and theoretical foundation, two studies of adolescent friendship intimacy are described. The first examines age and sex differences in intimacy across adolescence. The second explores differences between types of adolescent friendships in expressions of intimacy. Together, the studies provide sup- port for a model of intimacy that emphasizes dyadic closeness and indi- viduality. Furthermore, they suggest that intimacy may be critical in differentiating friendship types.

Several features of intimacy characterize adolescent friendships: Ado- lescents emphasize mutual trust, loyalty, and exclusivity as central to friend- ship (Berndt, 1983; Sharabany et al, 1981; Youniss and Smollar, 1985). Friends know one another's feelings and preferences, and support each other emotionally and materially. They discuss secrets, exchange ideas, and share feelings in a secure and accepting environment (Bigelow, 1977; Fur- man and Bierman, 1983; Oden et al., 1984). Across adolescence, these char- acteristics of intimacy emerge as increasingly important constructs around which close friendships are organized (Hartup, 1993; Jones and Dembo, 1989).

Conceptually, adolescent friendship intimacy has converged on two themes: closeness and individuality. Closeness describes mutual empathy, love, and felt security (Sullivan, 1953). This closeness provides the impetus for self-disclosure, prompting discussions of personal matters such as sexu- ality, family problems, and money. Thus, closeness captures the interper- sonal processes whereby friends share important feelings and information (Reis and Shaver, 1988). The need for a close friend is especially strong during early adolescence. Studies have shown how this need for affiliation can be manifested in an increase of willingness to be similar to the other and in conformity to peer pressure (Berndt, 1979; Brown, et al., 1986). In addition, through rewarding exchanges, friends strongly influence the thoughts, feelings, and behavior of one another (Kelly et al., 1983; Laarsen, 1993). In sum, the penchant for closeness between friends may be accom- panied by conformity to or control of the mend.

Individuality, in contrast, describes the development of separate and distinct identities. Erickson (1963) elaborated on the interplay between in- timacy and identity. He claimed that a successful intimacy requires feelings of power and control allowing "fusion without fear of ego toss" (p. 264). In such a relationship intimate partners are confident to express their own views. In addition, partners are respected by each other. Though the need

Shuintan et al.598

for close friends has been described as a hallmark of the developing ado- lescent (Sullivan, 1953), individuation also emerges gradually during ado- lescence, as children strive to distinguish themselves from both parents and peers (Grotevant and Cooper, 1985). Through individuation, adolescents express personal styles and create unique selves to be shared in intimate relationships.

In a recent discussion of the nature of friendships across adolescence, Selman (1990) suggested that the maturity of a relationship is evident in the interplay between intimacy and autonomy. At the lowest levels of in- timacy, partners have a sense of shared experience where the feelings and actions of one are imitated by the other. Though such partners have a strong sense of closeness, the capacity to negotiate individual preferences is lacking. At the next higher level of intimacy, friends appear to share activities and preferences; still, one partner tends to impose on or control the other. At the highest level, friends negotiate and integrate their needs, carefully balancing closeness and individuality.

Taken together, a new conceptualization of adolescent intimacy is sug- gested. Intimacy entails closeness, affection, disclosure, and commitment between friends. In addition, partners may tend to control or conform to the other in order to enhance the sense of closeness. Yet intimacy also concerns balancing needs and respecting individuality as well as different views. The manner in which this tension is resolved reflects the maturity of a relationship. Thus, balanced relationships and respect for the friend in adolescent close friendships should increase with age, whereas the ten- dency to control or to establish enmeshed relationships should decrease.

Evidence to date suggests that there are sex differences in friendship intimacy. Adolescent females are reportedly closer and more inclined to self-disclosure than males (Camerana et al., 1990; Jones and Dernbo, 1989; Sharabany et al., 1981). Males tend to express themselves through sepa- rateness, characterizing friendships in terms of shared activities, whereas females perceive relatedness, emphasizing mutual closeness and reciprocity in friendships (Smollar and Youniss, 1982). However, to the best of our knowledge, former studies have not explicitly compared how adolescent males and females balance closeness and individuality within their close friendships. It would also be reasonable to assume that girls, for whom it is important to secure connectedness, will be more inclined to give up in- dividuality. Boys, in contrast, will tend to control their peers and insist on their individuality.

Few studies have examined individual differences in adolescent rela- tionship intimacy. Research suggests that variations in intimacy produce three types of romantic relationships: merger, pseudointimacy, and genuine intimacy (Orlofsky, 1976; Orlofsky et al., 1973). Merger describes a rela-

Adolescent Intimacy 598

tionship that lacks balance and free expression. Pseudointimate relation- ships offer balanced roles and room to explore individuality, but little com- mitment to the relationship. Genuine intimacy is characterized by depth of roles and mutual commitment to the relationship. Similar distinctions are hypothesized for adolescent friendships. Research adapting a relationship typology from general systems theory (Minuchin, 1974; Reiss, 1981; Wynne, 1958, 1970) identified two types of friendships: interdependent and disen- gaged (Shuiman, 1993). Interdependent friends were found to be connected by an emotional bond—it was important for them to cooperate. However, this closeness neither involved total dependence nor precluded separateness in thinking and acting. In the disengaged type, close friends probably were unable to integrate differing opinions and therefore ended up working separately and insisting on their individuality. We suggest that the distinc- tion between the two types of friendship might further be demonstrated by their concepts of intimacy. For adolescents belonging to the interde- pendent type of friendship, this concept might reveal a better balance be- tween closeness and individuality, whereas for disengaged adolescents this concept might reflect less respect for their close friend's individuality.

The present investigation is based on a model of adolescent friendship intimacy that considers both normative developmental depictions of friend- ship as well as individual differences in the quality of close relationships Shulman et al., 1994). From this perspective, friendship intimacy entails af- fection, disclosure, and commitment, as well as a tendency to impose con- trol or to conform to the other in order to secure the friendship. In addition, intimacy also entails aspects of individuality like respect for the friend and ability to cope with differing views. The maturity of the rela- tionship reflects a balance of closeness and individuality. Across adoles- cence, more mature relationships should develop as children are better able to integrate the competing demands of friendship. Individual differences are expected, however, as some adolescents lag behind others in intimacy, especially respect for individuality and expressions of closeness.

To elaborate this model of adolescent friendship intimacy, two studies of closeness and individuality are described. The first explores attributes and developmental patterns of adolescent friendship intimacy. Three ques- tions concerning normative features of adolescent friendship are addressed: (1) What attributes characterize intimate adolescent friendships? (2) Does friendship intimacy change across adolescence? (3) Are there gender dif- ferences in perceptions of intimacy? Regarding the second and third ques- tions raised by Bh:wipJOYO-2605electJOYO-2605_ART_006JOYO-2605_ART_006thhis study, it is expected that such attributes of individuality as Balanced Relatedness and Respect for Friend will increase with age, whereas such attributes of closeness such as Control and Conformity will decrease with age. It is also expected that females will report higher levels

Shulman et al.600

601

of attributes of intimacy reflecting emphasis on closeness, whereas males will report higher levels of individuality and control of friend. The second study identifies qualitative differences between adolescent friendships in perceptions of intimacy. Two questions concerning variations in types of adolescent friendships are addressed: (1) Do perceptions of intimacy dis- tinguish adolescents with close friends from those without close friends? (2) Does intimacy vary across different types of adolescent friendships?

These studies were designed to advance our understanding of adoles- cent friendship, specifying the normative role of intimacy in the relationship and distinguishing between different types of friendship OH the basis of in- timacy. Based on previous findings, it was hypothesizedthat mature char- acteristics of intimacy (i.e., a balance of closeness and individuality) were expected to increase with age, corresponding with a decrease in less mature characteristics of intimacy (i.e., control and conformity). Gender differences were also anticipated, with greater intimacy, conformity, closeness, and self- disclosure among females than males. In contrast, males were expected to emphasize individuality more than females. Individual differences should emerge in types of friendship, with interdependent friends reporting more closeness, individuality, and self-disclosure than disengaged friends. In con- trast, disengaged friends were expected to report more control and con- formity than interdependent friends.

STUDY 1

Method

Subjects

A total of 288 adolescents in the 7th (M = 12.6 years), 9th (M = 14.5 years), and llth (M = 16.6 years) grades participated in the research. The sample included 57 males and 59 females in the 7th grade, 43 males and 45 females in the 9th grade, and 46 males and 38 females in the llth grade. Subjects resided in the Tel Aviv metropolitan area; most were from middle and iower-to-middle class families. The final sample represented 93% of all adolescents originally Invited to participate.

Instruments

Two scales (see Appendix) were constructed to assess intimacy and self-disclosure. Scale items were adapted from established instruments (originally designed for college students and young adults), which assess

Adolescent Intimacy

characteristics of friend, romantic,and marital relationships (Jourard, 1964; Schaefer and Edgerton, 1979; Sharabany et aL, 1981). Subjects rated inti- macy items describing a same-sex friendship on a 4-point scale ranging from low (1) to high (4): "To what extent do the following statements charac- terize your relationship with your dose friend?" The final questionnaire consisted of 5 intimacy subscales with 8 items in each: (1) Emotional Close- ness included shared affect, availability, and instrumental assistance; (2) Balanced Relatedness described tolerance for differing opinions and ideas; (3) Respect for Friend assessed mutual respect for individuation, compe- tence, and uniqueness; (4) Conformity assayed similarity in appearance and ideas, and the importance of conforming on these issues; and (5) Control measured preference for unilateral decisioa making. Scores for each subscale ranged from 8 to 32.

Subjects also completed a questionnaire describing self-disclosure,5 rat- ing items on a 4-point scale ranging from rarefy (1) to almost always (4): "To what extent do you share with your close friend about the following issues?" The fiaai questionnaire consisted of 3 self-disclosure subscales with 8 items in each: (1) Disclosure About Family described parental attributes and home atmosphere; (2) Disclosure About Friends assessed perceptions of and exchanges in close peer relationships; and (3) Disclosure About Physical Development included concerns of appearance and maturation. Scores for each subscale ranged from 8 to 32.

The final questionnaires were derived from pilot measures adminis- tered to 134 adolescents (64 males and 70 females) in Grades 9 and 10 (M = 15.6 years). This pilot sample was not part of the final pool. The pilot intimacy questionnaire included 60 items, with 12 items for each of the 5 domains. The pilot of the self-disclosure questionnaire included 36 items, with 12 items for each of the 3 domaitis. Two separate confirmatory factor analyses supported the hypothesized 5 factor structure of intimacy aad 3 factor structure of self-disclosure, explaining 62 and 58% of the vari- ance, respectively. The feat intimacy and self-disclosure subscales included the 8 items with the highest factor loadings (factor loadings given in pa- rentheses): Emotional Closeness (.56 to .84); Balanced Relatedness (.44 to .76); Respect for Friend (.47 to .73); Control (.68 to .81); Conformity (.41

5Conceptuatty, self-disclosure reflects a measure of closeness between friends and as such is similar to the subscale of Emotional Ctossnem. However, in former studies, self-disclosure was considered as the essence of intimacy. Jourard (1964) considered self-disclosure as the central attribute of intimacy. Reis «d Shaver (1988) defined intimacy as a process in which a person expresses important self-relevant feelings and information to another. Youniss and Smotiar (19S5) have shown the emergent and increasing importance of self-dtsctosiire in adolescent friendships. For this reason a separate questionnaire for measuring self-disclosure was constructed.

fit Shulman et al.

603

to .64); Disclosure About Family (.69 to ,81); Disclosure About Friends (.51 to .84); and Disclosure About Physical Development, (.40 to .79).

Procedure

Upon receipt of parent consent, questionnaires were administered in class to groups of 20 adolescents. Internal reliability coefficients (alphas) were computed for the sample as a whole as well as separately by age and gender (see Appendix). In all cases, subscale alphas were adequate, ranging from .69 to .95 (M = .88). All questionnaires were administered in Hebrew.

Results

Separate sets of analyses were conducted on the intimacy and self-dis- closure scales. First, Pearson correlations (ps < .01), conducted for the sam- ple as a whole and for each grade, determined interrelations among subscales. Second, multivariate analyses of variance (MANOVA) explored grade and sex differences in the five intimacy subscales and the three self- disclosure subscales. Analyses of variance (ANOVA) and Scheffe contrasts (ps < .05) followed statistically significant MANOVAs to specify differ- ences.

Pearson correlations of intimacy subscales revealed similar results for the sample as a whole and for each grade in associations among Emotional Closeness, Balanced Relatedness, and Respect for Friend (rs — .35 to .56). In addition, Control was negatively associated with Emotional Closeness, Balanced Relatedsess, and Respect for Friend (rs = -.15 to -.50). Corre- lations between Conformity and other intimacy subscales were statistically significant only for 7th graders (rs = .27 to .50). Pearson correlations of self-disclosure subscales revealed similar results for the sample as a whole and for each grade in associations among Family Disclosure, Friendship Disclosure, and Physical Development Disclosure (rs — .52 to .72).

Intimacy

A MANOVA was conducted with 2 (sex) x 3 (grade) levels of between subject and 5 (intimacy subscales) levels of repeated measure within subject independent variables. Intimacy was the dependent variable. A significant main effect emerged for intimacy, F(4,279) = 212.39, p < .001. In addition, there were two-way interactions between grade and intimacy, F(8, 558) = 3.74, p < .05, and sex and intimacy, F(4, 279) = 8.64, p < .01. Follow-up

Adolescent Intimacy

ANOVAs and Scheflte contrasts or simple main effects specified grade and sex differences on each intimacy subscale, as well as differences between intimacy subscales within each grade and sex.

The first set of follow-up comparisons entailed five separate ANOVAs with grade as the independent variable. Each intimacy subscale was con- sidered, in turn, as the dependent variable. F values, means, and standard deviations are presented in Table I. Significant grade differences emerged for Balanced Reiatedness, Respect for Friend, Control, and Conformity. Scheffe follow-up contrasts elaborated these grade differences. Balanced Relatedness was lower among 7th graders than 9th and llth graders. Re- spect for Friend was lower among 9th graders than 7th and llth graders. Control was highest for 7th graders and lowest for 9th graders, with llth graders at an intermediate level. Finally, Conformity was higher among 7th graders than 9th and llth graders.

The second set of follow-up comparisons entailed five separate ANO- VAs with sex as the independent variable. Each intimacy subscale was con- sidered, in turn, as the dependent variable. Significant sex differences emerged on four intimacy subscales: Emotional Closeness, F(l, 286) = 23.45, p < .001; Balanced Relatedness, F(l, 286) = 8.35, p < .01; Control, F(l, 286) = 8.65, p < .01; and Conformity, F(l, 286) = 7.01, p < .01. Females (M = 25.29, SD = 4.00, and M = 25.73, SD = 4.52, respectively) reported more Emotional Closeness and Balanced Relatedness than males (M = 23.59, SD = 4.42, and M = 24.48, SD = 5.48, respectively). In ad- dition, males (M = 12.07, SD = 4.96, and M = 18.66, SD = 5.14, respec-

*For each subscale, potential scores ranged from 8 to 32, with higher scores indicating greater intimacy. Standard deviations are given in parentheses. Within rows, means with different subscripts indicate significant grade differences in Scheffe follow-up contrasts (p < .OS).

*p < .05. cp < .01.

604 Shulman et al.

Table I. Friendship Intimacy Across Adolescence"

Total Sample 7th Grade

Emotional Closeness

Balanced Relatedness

Respect for Friend

Control

Conformity

(n =288)

24.44 (4.22) 25.10 (5.06) 20.82 (5.38) 11.35 (4.58) 18.13 (4.66)

(n =116)

24.12 (534) 23.86" (5.71) 21.21" (5.51) 12.30a

(5.23) 19.02" (5.30)

9th Grade (n = 88)

24.48 (4.67) 25.82* (3.96) 19.86b

(4.92) 10.29b

(3.34) 17.73b

(3.87)

llth Grade (n = 84)

24.75 (4.8S) 26.03b

(4.45) 21.31" (5.57) 11.06ab

(4.24) 17.33b

(3.93)

F Value

0.74

5.08b

4.32*

7.33C

8.51C

605Adolescent Intimacy

lively) reported more Control and Conformity than females (M = 17.59, SD = 4.06, and M – 10.63, SD = 4.05, respectively).

The third set of follow-up ANOVAs compared intimacy subscales within each grade and sex. Each revealed a significant main effect for in- timacy (ps < .05). Subjects in each grade and gender group ranked differ- ently the five intimacy subscales. Simple main effects specified differences between intimacy subscales. As can be seen in Table I, subjects reported the highest levels of intimacy for Emotional Closeness and Balanced Re- latedness, followed by Respect for Friend, then Conformity, and finally Control.

Self-Disclosure

A MANOVA was conducted with 2 (sex) x 3 (grade) levels of between subject and 3 (self-disclosure subscales) levels of repeated measure within subject independent variables. Self-disclosure was the dependent variable. A significant main effect emerged for self-disclosure, F(2, 281) = 4.38, p < .05. In addition, there was a two-way interaction between sex and self- disclosure, F(l, 281) = 8.26, p < .01. Follow-up ANOVAs and Scheffe contrasts or simple main effects specified sex differences on each intimacy subscale, as well as differences between intimacy subscales within each sex.

The first set of follow-up contrasts entailed three separate ANOVAs with sex as the independent variable. Each self-disclosure subscale was con- sidered, in turn, as the dependent variable. A significant main effect of sex emerged for Family Disclosure, F(l, 286) = 11.98, p < .01, Friendship Dis- closure, F(l, 286) = 4.75, p < .05, and Physical Development Disclosure, F(l, 286) = 17.01, p < .001. Females reported higher levels of Family Dis- closure (M = 20.03, SD = 7.70), Friendship Disclosure (M = 19.12, SD = 6.65), and Physical Development Disclosure (M = 21.45, SD = 7.33) than males (M – 17.15, SD = 6.85; M = 18.03, SD = 6.33; and M – 18.30, SD = 7.30, respectively).

Two additional sets of follow-up ANOVAs compared self-disclosure subscales within each sex. A significant main effect for self-disclosure was found only among females F(2, 281) = 5.38, p < .05. Simple main effects specified differences between self-disclosure subscales. Females reported the highest level for Physical Development Self-Disclosure followed by Family and Friendship Self-Disclosure. Males ranked the three Self-Dis- closure subscales in a similar order.

Discussion

The factor structure supported five hypothesized dimensions of inti- macy that included emotional closeness, balanced relatedness, respect for friend, conformity, and control, as well as three dimensions of self-disclo- sure (family, friends, and physical development). Adolescents tended to em- phasize intimate aspects of emotional closeness, on the one hand, and tolerance for differering opinions and individuality, on the other hand. Con- formity and control were not as characteristic of friendship intimacy. These findings suggest a comprehensive friendship iatimacy construct that differs somewhat from descriptions that exclusively emphasize emotional closeness and self-disclosure (Reis and Shaver, 1988). Such findings are consistent with a systems theory perspective on friendship, wherein the needs of the Individual are negotiated alongside the needs of the relationship (Shulman, 1993). SteiBglass (1978), for instance, argues that it is the balance between vectors of closeness and vectors of individuality that keeps system "parti- cles" (i.e., participants) together.

Results showed that balanced relatedness in friendships increase with age, whereas control and conformity decline. These trends conform to Sei- man's (1990) developmental sequence of friendship intimacy: Shared ex- perience is first achieved by acting the same as a close friend or by directing their behavior; later, shared intimate experience involves communication, integ^ting the needs of both parties. A closer inspection of the age dif- ferences reveals that age differences in perception of intimacy were mainly found between 7th graders and between 9th and llth graders. We speculate that it is probably during the transition to adolescence their children change tfaek perceptions of close relationships. After being more involved with peers, older adolescents learn more how to balance their own needs as well as the needs of others. In the younger group, friends are more inclined to use measures like control. Age differences on the Respect for Friend attribute showed a curvilinear tread; It was low for the 9th graders and high for the 7th and llth graders. This finding is contrary to our expecta tions and does not fit the other findings that suggested balanced relation- ships increase and control and conformity decrease with age. Further studies could clarify this issue. There were no age differences in emotional closeness and self-disclosure, supporting assertions of consistency in these friendship characteristics during adolescence (Jones and Dembo, 1990; Sharafaany et a!., 1981). Closeness and self-disclosure become increasingly salient to Mends across childhood, and their stability during adolescence reflects the position of importance already accorded to these relationship characteristics.

Shulman et al.606

Although the relative importance of friendship attributes were similar for males and females, hypothesized sex differences emerged in overall lev- els of closeness, self-disclosure, control, and conformity. Previous studies have also reported that females feel closer to and are more open with friends than males (Camarena et a/., 1990; Jones and Dembo, 1990; Sfaara- bany et a!., 1981). Greater mutual disclosure might prompt females to re- alize thaat friends have distinctive private domains that require respect to maintain the relationship; this was reflected in the higher level of tolerance that females expressed regarding differing opinions. In contrast, male friendships, which consist more of joint activities than shared ideas and emotions (Smollar and Youniss, 1982), tended to resort to control and con- formity to maintain the relationship. Still, it is important to note that both males and females in the present study rated closeness as the most impor- tant feature of their relationships.

STUDY 2

Building on the normative description of adolescent friendships pro- vided by Study 1, an exploratory investigation of individual differences in friendship intimacy was conducted in Study 2. Previous research (see Sfaul- man, 1993) suggests variations in adolescent friendship styles. The present study was designed to replicate and extend these findings. To this end, a laboratory task identified different types of adolescent friendships (e.g., in- terdependent and disengaged). These groups of friends (and a control group of nonfriends) are compared on the attributes of intimacy identified in Study 1. It is hypothesized that adolescents who belong to interdepend- ent friendships will display a balance between intimacy attributes reflecting closeness and individuality. In addition, disengaged friends and adolescents without friends are expected to report either a lower level of closeness to Mend or a relationship that emphasizes aspects of control or conformity.

Method

Subjects

Participants included 13? adolescents (M = 14.3 years) ia the 9th grade. Subjects were drawn from a pool of 228 volunteers with parental consent The final sample included 98 adolescents (50 mates and 48 fe- males) with at least one reciprocated friendship and 39 adolescents (12 males and 27 females) without a reciprocated friendship. Subjects resided in a small city in central Israel; most belonged to middle-class families.

Adolescent Intimacy 607

A two-step procedure identified adolescents with reciprocated friends. First, subjects named their closest friends in an open-ended questionnaire Second, subjects were interviewed about closeness in these friendships, rat- ing each on a 7-point sc