gender role orientations and friendship: some attenuation, but gender differences abound

16
Sex Roles, Vol. 24, Nos. 9/10, 1991 Gender Role Orientations and Friendship: Some Attenuation, But Gender Differences Abound I Paul H. Wright s University of North Dakota Mary Beth Scanlon Ramsey Clinic Hastings, Minnesota In an exploration of the degree to which gender role orientations would at- tenuate differences on a set of variables specified in a friendship model 105 women and 101 men described themselves and their best friend of each gender on the Bern Sex Role Inventory. They also described their relationships with each friend using the Acquaintance Description Form F. Women's friend- ships with female friends were especially strong and rewarding as compared to their friendships with men, and as compared to men's friendships with either women or men. There was a modest attenuating effect due to sub- jects" GROs. Women, but not men, were sensitive to the perceived GROs o f friends, with androgynous friends providing the strongest and broadest array of rewards. Contrary to the common conception of women's friend- ships as expressive and men's as instrumental the results suggest that wom- en's friendships are both expressive and instrumental. Studies by researchers from a variety of disciplines using differing concep- tual foci, differing research settings, differing methodologies, and differing analytic procedures indicate consistently that women's friendships differ from those of men in several specifiable ways. Current and comprehensive reviews of the sizable and growing literature on this problem are available (Sherrod, 1989: Winstead, 1988; Wright, 1989). ~We gratefully acknowledge the assistance of Terry Schieffer in the data collection. 2To whom reprint requests should be addressed at Department of Psychology, University of North Dakota, Grand Forks, ND 58202. 551 0360-0025/91/0500-0551506.50/0 © 1991 Plenum Publishing Corporation

Upload: paul-h-wright

Post on 12-Aug-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Gender role orientations and friendship: Some attenuation, but gender differences abound

Sex Roles, Vol. 24, Nos. 9/10, 1991

Gender Role Orientations and Friendship: Some

Attenuation, But Gender Differences Abound I

Paul H. Wright s University of North Dakota

Mary Beth Scanlon Ramsey Clinic Hastings, Minnesota

In an exploration o f the degree to which gender role orientations would at- tenuate differences on a set o f variables specified in a friendship model 105 women and 101 men described themselves and their best friend o f each gender on the Bern Sex Role Inventory. They also described their relationships with each fr iend using the Acquaintance Description Form F. Women's friend- ships with female friends were especially strong and rewarding as compared to their friendships with men, and as compared to men's friendships with either women or men. There was a modest attenuating effect due to sub- jects" GROs. Women, but not men, were sensitive to the perceived GROs o f friends, with androgynous friends providing the strongest and broadest array o f rewards. Contrary to the common conception o f women's friend- ships as expressive and men's as instrumental the results suggest that wom- en's friendships are both expressive and instrumental.

Studies by researchers f rom a variety of disciplines using differing concep- tual foci, differing research settings, differing methodologies, and differing analytic procedures indicate consistently that women's friendships differ from those of men in several specifiable ways. Current and comprehensive reviews of the sizable and growing literature on this problem are available (Sherrod, 1989: Winstead, 1988; Wright, 1989).

~We gratefully acknowledge the assistance of Terry Schieffer in the data collection. 2To whom reprint requests should be addressed at Department of Psychology, University of North Dakota, Grand Forks, ND 58202.

551 0360-0025/91/0500-0551506.50/0 © 1991 Plenum Publishing Corporation

Page 2: Gender role orientations and friendship: Some attenuation, but gender differences abound

552 Wright and Scanion

Common usage now characterizes women's same-gender friendships as "expressive," "communal," or "face-to-face" and men's same-gender friend- ship as "instrumental," "agentic," or "side-by-side." These broad characteri- zations are virtually identical in connotation, and mask a number of specific elements in a model pattern of differences. Specifically, women's friendships are more likely than those of men to involve personalism, confiding, and emotional supportiveness (Booth, 1972; Booth & Hess, 1974; Davison & Pack- ard, 1981; Hill & Stull; 1981; Weiss & Lowenthal, 1975; Winstead & Darle- ga, 1984; Wright, 1985; Wright & Crawford, 1981; Yoon, 1978). Women's friendships are more likely to involve "mere talk" and intimate self-disclosure (Aries & Johnson, 1983; Caldwell & Peplau, 1982; Hacker, 1981; Johnson & Aries, 1983; Walker & Wright, 1976). Women's friendships are likely to be affectively richer, more complex, and more "wholistic"; men's friendships are more likely to be centered around structured activities, and men are like- ly to have less wholistic "special purpose" friendships (Block, 1980; Weiss & Lowenthal, 1975; Wright, 1982). It is noteworthy that, with one excep- tion, this modal pattern identifies characteristics that women's friendships exhibit more frequently or in greater degrees than those of men. The excep- tion is that men's friendships are more likely to involve structured tasks and activities.

Researchers have devoted considerably less attention to cross-gender than to same-gender friendships. Nevertheless, a modal pattern of differ- ences has emerged that, on the surface, appears simpler than that for same- gender friendships. Averaging across noncollege samples (Bell, 1981a; Block, 1980; Booth & Hess, 1974; Rubin, 1985), roughly 40% of the men and 30% of the women report having one or more cross-gender friends. These per- centages are somewhat higher for college samples (e.g., Rose, 1985). With respect to the quality of the friendships, men tend to be more open, more self-disclosing, and more intimate with their women than with their men friends; women tend to be less open, less self-disclosing, and less intimate with their men than with their women friends (Block, 1980; Rose, 1985; Rubin, 1985).

One might be tempted at this point to conclude simply that men con- sider their cross-gender friendships as expressive whereas women consider their cross-gender friendships as, if anything, instrumental. However, studies of comparative personal relationships (Wright, 1985; Wright & Bergloff, 1984) suggest a somewhat different conclusion. Women indicated that both female friends and heterosexual partners provided such interpersonal rewards as self-affirmation and emotional support to a significantly greater degree than did male friends. Men indicated that heterosexual partners provided interpersonal rewards to a greater degree than did either female or male friends, although they saw female friends as providing somewhat more sup-

Page 3: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 553

port and security than male friends. These findings are consistent with others, suggesting that men most often turn for close, highly personalized interac- tion to neither male nor female friends, but to romantic partners or female kin. Women, on the other hand, are likely to turn to female friends (Block, 1980; Rubin, 1985; Tognoli, 1980). Perhaps the safest conclusion at the present time is that cross-gender friendships, while not unheard of, are sel- dom extremely close, especially for women. Therefore, from the standpoint of sharing confidences and communal interaction, cross-gender friendships are not important. However, on those relatively infrequent occasions when a man does claim a close cross-gender friendship, he is likely to consider it more supportive and emotionally rewarding than his friendship with another man.

The robustness and apparent pervasiveness of such gender differences led Bell (1981a, p. 55) to comment as follows: "When we look at friendship in society, we can see many variations. But there is no social factor that is more important than sex in leading to friendship variations." Among sever- al reasons for considering Bell's conclusion premature and possibly overdrawn (see Wright, 1988) is the status of gender as an "organismic" or "subject" variable. As Underwood (1957) emphasized long ago, such variables are vul- nerable to uncertainties of interpretation due to their inevitable correlation with other factors. In short, it is difficult to determine whether the focal vari- able, e.g., gender, or one of its known or unknown correlates is more in- fluential or basic (see especially Deaux, 1984). Indeed, research has uncovered several variables that attentuate and sometimes override gender differences in friendship. Bell (1981b) himself concluded that a nonventional personal orientation may nullify such gender differences. McAdams, Healy, and Kraus (1984) found that high intimacy motivation and high power motivation were related, respectively, to a communal and an agentic emphasis in friendship regardless of the gender of the subject, effectively overriding gender differences.

Probably the most obvious and thoroughly explored attenuating vari- able is gender role orientation as defined and measured by Bern (1974). In a study by Lombardo and Lavine (1981), androgynous men reported self- disclosing to their friends at higher levels than did traditional men, and self- disclosing at equally high levels to their female and male friends. Androgynous women reported self-disclosing to friends of both genders while traditional women self-disclosed more to their female friends. In a subsequent study (Lavine & Lombardo, 1984), androgynous men reported disclosing as much to best male friends as traditional or androgynous women disclose to their best friends, and much more than traditional men disclose to their best friends. Studies of intimacy in friendship suggest an attenuating effect due to femininity, but not necessarily androgyny. Women and men scoring high

Page 4: Gender role orientations and friendship: Some attenuation, but gender differences abound

554 Wright and Scanlon

on femininity report high levels of intimacy in their friendships regardless of whether they score high or low on masculinity. In other words, both an- drogynous and cross-sex typed (feminine) men scored higher on intimacy than masculine men: androgynous and sex-typed women scored higher on intimacy than masculine women (Fischer & Narus, 1981; Williams, 1985).

Two points seem noteworthy about studies of sex role orientations and gender differences in friendship. First, the studies focus upon single, rela- tively circumscribed aspects of friendship rather than upon friendship as a whole, i.e., as a distinctive relationship involving an array of relevant varia- bles. Second, they are limited to expressive rather than instrumental charac- teristics, i.e., variables that are more characteristic of women's than men's friendships. In other words, the emphasis is upon whether differing gender role orientations make either women's or men's friendships more or less ex- pressive, not upon whether they make them more or less instrumental.

In the present study, we addressed the question of which, if any, of the variables in a multivariate model of friendship would be related to the subject's gender role orientation (GRO). Following Bem (1974), (3ROs were conceptualized in terms of personal traits positively associated with femi- ninity and masculinity. Subjects of either sex who identified strongly with traits positively associated with femininity but only weakly with those posi- tively associated with masculinity were considered feminine. Subjects who did the opposite were considered masculine. Subjects who identified strong- ly with both were considered androgynous, and those who identified them- selves weakly with both were considered undifferentiated. We also examinied the impact of the subject's attribution to her or his friend (target person) of the same feminine and masculine traits, and categorized target persons according to the GRO typology. Although it was not strictly accurate do so, for convenience we called these categories the perceived GROs of the target persons.

We were particularly interested in gender differences in same- and cross- gender friendships, and how these differences are related to the subject's own GRO and the perceived GRO of her or his friend. We organized our study around the following set of questions: (1) On which, if any, of the friend- ship variables will women and men differ in their responses to same- and cross-gender friends? (2) On which, if any, of the friendship variables will subjects' responses to their same- and cross-gender friends vary with the sub- ject's GRO? (3) On which, if any, of the friendship variables will subjects' responses to their same- and cross-gender friends vary with the perceived GRO of the friend? (4) In what ways, if any, will the subjects' GROs attenuate any differences found in answer to Question 17 Specifically, on those variables show- ing an overall gender difference, will the responses of masculine and /or androgynous women to their female and male friends be similar to those of

Page 5: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 555

the average male, and will the responses of feminine and /or androgynous men to their female and male friends be similar to those of the average female?

M E T H O D

Procedure and Characteristics of the Sample

The investigators and cooperating contact persons solicited the partici- pation of individuals from work settings, business organizations, and serv- ice groups in the Minneapolis, Minnesota, metropolitan area. Each volunteer received a packet of forms including a demographic information sheet, three copies of the Bern Sex Role Inventory (BSRI, 1974), and two copies of an instrument assessing various facets of friendship called the Acquaintance Description Form F (ADF-F: Wright, 1985). The subject's instructions were (1) to fill out the demographic information sheet; (2) to complete the BSRI as it applied to her or himself; (3) to complete the BSRI with respect to the way (s)he perceived her or his best same-gender and best cross-gender friend, excluding relatives, spouses or romantic partners; and (4) to use the ADF-F to describe her or his relationship with each of the same two friends (s)he had described with the BSRI. Subjects completed the forms at their leisure and returned them to the contact person.

All 108 women and 103 men who volunteered to participate in the study returned data packets. Three of the packets from women and two from men did not include responses to a cross-gender friend. Responses from these in- complete data sets were not used in any of the analyses, leaving the actual sample size at 105 women and 101 men.

Subjects in both the women's and men's groups ranged in age from the early 20s to the late 50s with respective mean ages of 37 and 38 years. Sixty- four percent of the women and 61°70 of the men were married. Of the re- maining women 18070 were single, 15°70 were divorced, and 3% were sepa- rated. Of the remaining men, 26o/0 were single, 12O7o were divorced, and 1% were separated.

With respect to educational levels attained, 37% of the women had completed high school, 6O7o a vocational-technical program, 16°/0 an associ- ate in arts degree, 33°7o a bachelor's degree, and 8°70 a master's degree. Thirty percent of the men had completed high school, 1% a vocational-technical program, 20% an associate in arts degree, 32% a bachelor's degree, 11% a master's degree, and 7°/0 a professional degree (e.g., taw). The subjects worked in a wide variety of occupations and professions. None of them were unemployed or retired, and none worked primarily in the home.

Page 6: Gender role orientations and friendship: Some attenuation, but gender differences abound

556 Wright and Scanlon

Lengths of friendships varied from less than 5 to more than 21 years, with 80% of the women reporting same-gender and 86% reporting cross- gender friendships of over 5 years' duration. Among the men, 79% reported same-gender and 50% reported cross-gender friendships of over 5 years' du- ration.

In sum, the sample represents a broad cross section of nonstudent adults describing both same- and cross-gender friendships that have, for the most part, stood the test of time.

Instruments

The BSRI and the Assignment o f GROs. Subjects using the BSRI indi- cate on a 7-point Likert scale the degree to which each of 20 masculine, 20 feminine, and 20 neutral adjectives applies to themselves. In the present study, we asked each subject to also indicate in the same manner the degree to which the adjectives applied to her or his same- and cross-gender friend (target per- son). The long-form scoring procedure was used. That is, each subject's responses to the feminine items and to the masculine items were summed and divided by 20 to provide separate overall scores. Such scores are statisti- cally independent of one another and show high stability over time. Basic psy- chometric information is available from Bern (1974).

In assigning GROs to subjects and perceived GROs to their respective target persons (TPs), we followed the common practice of using four categories based on median masculinity and femininity scores. Median scores for subjects responding to themselves were 4.8 for femininity and 4.85 for masculinity. These are strikingly similar to the respective median scores of 4.9 and 4.95 reported by Bern (1981). Subjects scoring above the median on masculinity and below the median on femininity were classed as masculine; those showing the opposite pattern were classed as feminine. Subjects scoring above the median on both were classed as androgynous, and those scoring below on both were classed as undifferentiated. The perceived SROs of TPs were assigned in the same way, with median splits based on the femininity and masculinity scores attributed to them. Because each subject responded to both a same- and a cross-gender friend, it was necessary to calculate two median scores for perceived femininity and two for perceived masculinity in order to maintain independence of observations. In each case, the median varied only minimally from one another and from those based on subjects responding to themselves. Therefore, assignments of TPs' perceived GROs were based on the same median values as for assignments of subjects' GROs.

The ADF-F. The ADF-F is a self-report technique for measuring 13 variables specified in a relationship model (Wright, 1984, 1985). The subject

Page 7: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 557

indicates on a 0-6 point scale the frequency or probability of occurrence of each of a series of statements about a specified acquaintance or TP. Each variable is measured by responses to 5 items for a maximum score of 30. The scales include the following: (1) two measures of relationship strength called Voluntary Interdependence (unconstrained or nonobligatory interac- tion) and Person-qua-Person (a personalized interest in and concern for one's TP); (2) five classes of interpersonal rewards or "values" found in the rela- tionship, i.e., Ego Support Value, Self-affirmation Value, Stimula- tion Value, Utility Value, and Security Value; (3) four scales for differen- tiating among different kinds of relationships, i.e., Exclusiveness, Per- manence, Salience of Emotional Expression, and Degree of Social Regu- lation; (4) an index of tension or strain in the relationship called Maintenance Difficulty; and (5) a measure of the subject's tendency to respond to her or his TP in a globally favorable or unfavorable manner, called General Favora- bility. Studies of reliability, validity, and factor structure are reported in Wright (1985) and Lea (1989).

RESULTS

Preliminary Analyses and Overall Gender Differences

The number of female subjects falling into each GRO category was as follows: feminine, 39; masculine, 18; androgynous, 31; undifferentiated, 17. The distribution of GROs of female TPs as perceived by female subjects was: feminine, 35; masculine, 15; androgynous, 36; undifferentiated, 19. The dis- tribution of GROs of male TPs as perceived by female subjects was femi- nine, 12; masculine, 40; androgynous, 31; undifferentiated, 22. The number of male subjects in each GRO category was feminine, 19; masculine, 33; androgynous, 22; undifferentiated, 27. The distribution of GROs of fe- male TPs as perceived by male subjects was feminine, 46; masculine, 17; an- drogynous, 20; undifferentiated, 18. The distribution of GROs of male TPs as perceived by male subjects was feminine, 7; masculine, 54; androgynous, 11; undifferentiated, 29.

We checked first on the possibility that subjects might tend to regard their friends as similar to themselves in GRO. After calculating expected fre- quencies of similar and dissimilar choices on the basis of the number of sub- jects and the number of TPs in each category, we computed chi-squared statistics for each of the groupings, i.e., females responding to female friends, females responding to male friends, etc. None of these analyses showed an appreciable tendency for subjects to attribute their own GROs to their respec- tive TPs. The largest chi-squared, 13.11 (df = 7; p = .069), was for male subjects responding to female friends.

Page 8: Gender role orientations and friendship: Some attenuation, but gender differences abound

558 Wright and Scanlon

Next, we assessed overall differences between women's and men's responses to their same- and cross-gender friends on each of the ADF-F vari- ables. We conducted a series of 2 (gender of subjects) × 2 (gender of TP) analyses of variance (ANOVAs) with repeated measures on the gender of TP factor. In order to maintain a protection level of .05 for the entire set of analyses, the alpha level was set at .004 for each overall F ratio. Where indicated, these analyses were followed by the Newman-Kuels test for in- dividual contrasts. The results are summarized in Table I. The seemingly anomolous pattern of significant and nonsignificant differences for Social Regulation is due to the fact that each of the Newman-Keuls adjustments was based on two error terms, one for within-subject and one for between- subject comparisons.

Ten of the 13 ADF-F variables showed significant overall differences with respect to either main effects, interaction effects, or both. Individual contrasts revealed that, for 9 of these variables, female and male subjects differed significantly in their responses to female TPs. Only in the case of Ego Support Value did female and male subjects differ significantly in their responses to male TPs, with females showing the higher ESV mean.

Concerning within-gender effects, comparisons involving female sub- jects' responses to female vs. male TPs showed significant differences on 6 ADF-F variables. Specifically, female subjects rated female TPs as higher than male TPs on the two measures of relationship strength (VID and PQP), on three of five interpersonal rewards (UV, SAV, and SECV), and on General Favorability. They rated their relationships with female TPs as significantly lower on Social Regulation. Comparisons involving male subjects' responses to female vs. male TPs showed a significant difference on only one variable, with males rating male TPs higher than female TPs on Voluntary Inter- dependence.

Gender Role Orientations of Subjects and Perceived Gender Role Orientations o f Target Persons

Subjects and TPs were arranged into four different 4 × 4 groups. Each group was based on the gender of the subject and the gender of the TP, i.e., female subjects responding to female TPs, female subjects responding to male TPs, male subjects responding to female TPs, and male subjects responding to male TPs. In each case, entries in the 16 cells were determined by the GRO of the subject and the perceived GRO of her or his TP. The plan was to perform 4 (GRO of subject) × 4 (Perceived GRO of TP) ANOVAs on each ADF-F variable for female subjects responding to female TPOs, female sub- jects responding to male TPs, etc. However, in every case, the distribution

Page 9: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 559

% o

0

"0

0 r~

r~

<

O I

~D

X

g~

~L <

~ ~ ~ = ~ '

~ . . . . . . . . . .

~ M d d d d M d d ~ d ~ d

~ 0

~g

~g X>g

~ ' ~ -~

©

.F g

o

V

.~oo ~ v v v

Page 10: Gender role orientations and friendship: Some attenuation, but gender differences abound

560 Wright and Scanlon

of entries was extremely uneven, with several very low-frequency cells. This precluded valid sets of 4 × 4 ANOVAs. Despite this limitation, we performed such ANOVAs for each of the four groupings to examine main effects, and to get a workable estimate of possible interactions between the GRO of the subject and the perceived GRO of the TP.

In no case was there any indication of a significant interaction effect. Therefore, we will deal separately with effects due to the GROs of the sub- jects and those due to perceived GROs of the TPs.

Attenuating Effects of Subjects" GROs. In spite of some variation in means associated with the GRO of the subject, in only one case did this vari- ation result in a significant overall Fra t io . This effect occurred for male sub- jects responding to female TPs with respect to self-affirmation value. The SAV means for feminine and androgynous males responding to female TPs were, respectively, 22.63 and 22.68, for the masculine and undifferentiated males, 20.88 and 18.87 (F = 4.90; p < .004).

SAV excepted, the lack of differences associated with subjects' GROs would seem to preclude the possibility of an appreciable attenuating effect. There is, however, another question that may be asked about such attenuat- ing effects, i.e., to what degree do the responses of feminine, masculine, an- drogynous, or undifferentiated females or males differ f rom the "typical" responses of members of the opposite gender? To answer this question, we applied the Dunnett test for comparing each of an array of means against a common standard. For each of the 9 variables showing a gender differ- ence in responses to female TPs, we took the overall mean for males respond- ing to female TPs as the standard against which to compare the means of feminine, masculine, androgynous, and undifferentiated females also responding to female TPs. Similarly, for the one variable (ESV) showing a gender difference in responses to male TPs, we took the overall male mean as the standard against which to compare the means of female subjects in the four GRO categories. These means and outcomes of the various com- parisons are presented in Table IIA. Next, we took the overall mean for fe- male subjects responding to female TPs as the standard against which to compare the corresponding means of male subjects in the four GRO categories. Finally, for the ESV variable, we compared the overall mean of female subject s responding to male TPs with those of male subjects in the four GRO categories. These means and outcomes of the comparisons are presented in Table liB.

Because we were concerned with possible attenuating effects, our in- terest in Tables I IA and IIB was in ADF-F means that were most similar to, i.e., did not differ significantly from, their respective standards. Thus, in Table IIA, eight of the nine means of undifferentiated female subjects responding to female TPs did not differ significantly f rom the overall mean

Page 11: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 561

Table IIA. Comparisons of Selected ADF-F Means of Female Subjects Having Differing Gender Role Orientations with Overall Male Means

Means by GRO of female subjects

Male Feminine Masculine Androgynous Undifferentiated Variable (n = 101) (n = 39) (n = 18) (n = 31) (n = 17) /~

Female TP VID 15.24 19.95 ~ 18.78 a 20.13 b 17.76 7.82 POP 21.89 26.25 b 24.61 a 25.26 c 24.35 9.58 SV 17.72 20.08 18.27 21.19 c 19.12 5.14 UV 20.35 24.72 b 21.50 24.45 b 20.88 9.45 SAV 21.04 24.74 b 23.11 24.97 b 21.82 8.80 ESV 21.60 24.69 c 22.28 24.29 ~ 22.29 6.02 SECV 22.69 26.41 b 23.56 26.00 b 25.12 a 8.08 PERM 12.70 16.77 b 17.22 ~ 16.13 a 15.00 8.27 GF 23.79 26.33 a 23.67 26.42 d 24.00 5.96

Male TP ESV 20.66 24.13 ~ 22.17 23.06 d 21.18 5.86

~All F ratios significant at p < .004. ~,°,dDiffers from the male mean according to the Dunnett test or .05.

at, respectively, p < .01, .02,

Table liB. Comparisons of Selected ADF-F Means of Male Subjects Having Differing Gender Role Orientation with Overall Female Means

Means by GRO of male subjects

Female Feminine Masculine Androgynous Undifferentiated Variable (n = 105) (n = 19) (n = 33) (n = 22) (n = 27) F ~

Female TP VID 19.45 17.11 15.39 d 15.09 c 13.85 b 8.17 PQP 25.37 23.53 22.09 21.36 b 20.93 b 9.99 SV 19.94 17.26 18.12 19.09 16.44 b 4.89 UV 23.47 20.26 a 20.97 a 20.95 a 19.14 b 6.49 SAV 24.06 22.63 20.88 c 22.68 18.78 b 10.56

ESV 23.77 22.11 21.58 a 22.54 20.51 b 4.88 SECV 25.59 24.00 22.96 d 21.27 b 22.59 c 7.57 PERM 16.37 13.95 12.58 c 11.54 b 12.93 ~ 8.37 GF 25.52 22.58 b 24.91 23.72 23.33 a 4.33

Male TP ESV 23.00 19.42 b 20.26 c 22.32 20.93 a 5.61

aAll F ratios significant at p < .004. b°'aDiffers from the female mean according to the Dunnett test or .05.

at, respectively, p < .01, .02,

o f m a l e s u b j e c t s r e s p o n d i n g t o f e m a l e T P s , n o r d i d s ix o f t h e n i n e m e a n s

f o r m a s c u l i n e f e m a l e s u b j e c t s . I n c o n t r a s t , all o f t h e m e a n s f o r a n d r o g y n o u s

f e m a l e s u b j e c t s , a n d e i g h t o f t h e m f o r f e m i n i n e f e m a l e s u b j e c t s , w e r e s ig-

n i f i c a n t l y h i g h e r t h a n t h e o v e r a l l m e a n f o r m a l e s u b j e c t s . S i m i l a r l y , t h e E S V

m e a n s f o r u n d i f f e r e n t i a t e d a n d m a s c u l i n e f e m a l e s u b j e c t s d e s c r i b i n g m a l e

Page 12: Gender role orientations and friendship: Some attenuation, but gender differences abound

562 Wright and Scanion

TPs did not differ significantly from the overall mean for male subjects describing male TPs, whereas those for androgynous and feminine female subjects did.

In Table liB, only two of the nine means for feminine male subjects describing female TPs differed significantly from the corresponding overall means for female subjects. For androgynous male subjects, five of the nine means were significantly lower than the female standard. For masculine male subjects, six of the means were lower than the female standard, and for un- differentiated male subjects all of them were. For male subjects responding to male TPs on ESV, the mean for androgynous subjects was not signifi- cantly different from the overall female mean, whereas those for feminine, masculine, and undifferentiated male subjects were significantly lower.

Perceived GROs of Target Persons. Concerning perceived GROs of TPs, significant differences were found on 6 variables for female subjects respond- ing to female TPs and on 3 variables for female subjects responding to male TPs. For the male subjects, there were no significant differences due to the GRO of either female or male TPs. Means for female subjects showing significant differences are presented in Table III, with the corresponding means for male subjects included for purposes of comparison.

Female subjectsresponded differentially to female TPs perceived to have differing GROs on each of the 5 interpersonal rewards, and on General Favorability. In every case, the mean for TPs perceived to be androgynous was highest and that for TPs perceived to be undifferentiated was lowest, with means for TPs perceived to be feminine and masculine falling between. However, in the case of Self-affirmation Value, means for both undifferen- tiated and masculine TPs were significantly lower than that for androgynous TPs, and in the case of General Favorability, the mean for feminine as well as that for androgynous TPs was significantly higher than the mean for un- differentiated TPs.

Female subjects responded differentially to male TPs perceived to have differing GROs on 3 of the 5 interpersonal rewards. Again, in every case, the mean for TPs perceived to be androgynous was highest and significantly different from the means for TPs perceived to be undifferentiated. The Self- affirmation Value mean for TPs perceived to be feminine was also signifi- cantly higher than that for those perceived to be undifferentiated, and the Ego Support Value mean was nearly so. However, the mean for Stimulation Value for feminine TPs was significantly lower than that for androgynous TPs, and identical with that for undifferentiated TPs.

D I S C U S S I O N

Our first orienting question addressed the issue of which ADF-F varia- bles would show overall gender differences in same- and cross-gender frien-

Page 13: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 563

Table III. Selected ADF-F Means by Gender Role Orientation of Target Persons a

GRO of TP

Feminine Masculine Androgynous Undifferentiated Variable (n = 35) (n = 15) (n = 36) (n = 19) F

Female subjects-Female TP SV 19.03ob 20.53ob 21.83° 17.58b 4.77 b UV 22.94~b 23.80ob 25.58° 20.16b 6.06 b SAV 24.00~b 22.27b 26.11° 21.68b 7.03 b ESV 23.77ob 22.80°b 25.50. 21.26b 5.62 b SECV 26.48.b 24.20°b 26.58° 23.16b 4.75 b GF 26.14~ 25.20.b 26.64~ 22.53b 6.24 b

Female subjects-Male TP (n = 12) (n = 40) (n = 31) (n = 22)

SV 17.00~ 19.88oe 21,42° 17.00~ 5.37 b SAV 23.67, 21.12oh 24.45° 18.72b 8.13 b ESV 24.00,~ 22.38o~ 24.96° 20.87b 5.42 b

Male subjects-Female TP (n = 46) (n = 17) (n = 20) (n = 19)

SV 17,06 19.06 18.40 17.39 1.38 UV 20.17 20.23 20.80 20.39 .10 SAV 20,54 21.00 22.65 20.56 1.21 ESV 21.47 22.82 21.95 20.39 1.28 SECV 22.74 23.17 21.80 23.11 .42 GF 23.91 23.82 24.55 22.61 .98

Male subjects-Male TP (n = 7) (n = 54) (n = 11) (n = 29)

SV 18.29 17.87 20.72 18.58 1.77 SAV 22.14 20.00 23.27 21,48 2.44 ESV 21.57 19.70 22.54 21,51 2.32

~Means in a given row not having the same subscript differ significantly (p < .01) according to the Newman-Keuls test.

bp < .004.

ships. This question can be answered parsimoniously by simply noting that the relationships of women with their women friends were especially strong and especially rewarding as compared to their friendships with men, and as compared to men's friendships with either women or men. As a result, women and men differed markedly in their friendships with women, but minimally in their friendships with men. Nor were there appreciable differences in wom- en's friendships with men vs. men's friendships with women. There were two exceptions to this overall pattern: women considered their men friends more encouraging and emotionally supportive than men considered their men friends, and women considered their men friends more interesting and stimulating than men considered their women friends.

The subjects' GROs had an attenuating effect on gender differences in their friendships that must be considered modest for two reasons. First, the attenuating effect applies almost exclusively to differences in women's and men's responses to women friends because, for men friends, there was only one such difference to attenuate. Second, our initial comparisons among

Page 14: Gender role orientations and friendship: Some attenuation, but gender differences abound

564 Wright and Scanlon

ADF-F scores for subjects with differing GROs revealed a significant differ- ence only for male subjects responding to female friends on the SAV varia- ble. Attenuating effects were in evidence only when the relevant ADF-F means of feminine, masculine, androgynous, and undifferentiated women were com- pared with the overall means of men, and similarly, when the means of femi- nine, masculine, androgynous, and undifferentiated men were compared with the overall means of women.

In general, the responses of undifferentiated and masculine women were similar to, i.e., not significantly different from, those of the average male subject, whereas the responses of feminine and androgynous women were not. For men, the pattern was less clear. In general, the means for feminine men were similar to those of the average woman whereas the means for undifferen- tiated men were not. For masculine and androgynous men, some of the means were similar to those of women and some were not. These findings are not inconsistent with past research on the attenuating effects of GROs. However, a better understanding of how strong and pervasive those effects are, and of how they operate, awaits the outcome of more extensive and more clearly focussed research.

A much more definite pattern is found when we consider the perceived GRO of the friend: women were sensitive to differing GROs in their friends; men were not. Specifically, women regarded female friends whom they per- ceived as androgynous as highest and those whom they perceived as un- differentiated as lowest on all the interpersonal rewards measured by the ADF-F. They regarded masculine and feminine friends as falling between these two extremes, and for the most part, not significantly different f rom either. Women were also sensitive to the perceived GROs of their male friends, but to a much lesser extent. They regarded androgynous male friends as more interesting and stimulating, self-affirming, and ego supportive than un- differentiated male friends. They considered their feminine male friends to be as self-affirming and almost as ego supportive as androgynous male friends, but as lacking in stimulation value. They considered masculine men intermediate on these characteristics.

CONCLUSION

A primary purpose of the present study was to examine possible ef- fects of GROs in attenuating the well-documented gender differences in friendship. What the study has accomplished instead is a possible sharpening of those differences and a broadened conception of the nature of women's same-gender friendships. The findings clearly suggest that women's friend- ships with other women play an especially significant role in their l i v e s -

Page 15: Gender role orientations and friendship: Some attenuation, but gender differences abound

Gender and Friendship 565

more so than their friendships with men, and more so than men's friend- ships with either women or men. This observation may not b e - i n fact, is not-part icular ly new. However, we are accustomed to thinking of wom- en's friendships as expressive and men's friendships as instrumental. By im- plication, this distinction means that women's friendships are expressive but not instrumental, whereas men's friendships are instrumental but not expres- sive. In light of the present results, this conception does not do justice to the breadth of women's friendships.

Women were highly responsive to their women friends with respect to such instrumental rewards as Stimulation Value and Utility Value as well as the more expressive rewards of Ego Support Value, Self-affirmation Value, and Security Value. Moreover, women were especially responsive, with respect to such rewards, to friends whom they perceived to be androgynous. A plau- sible reason for this is that the greater behavioral and personal flexibility associated with androgyny enables the friend to provide a broader range of interpersonal rewards. Thus, our findings do nothing to challenge the con- ception that men's friendships tend to be instrumental but not expressive. However, they strongly suggest that women's friendships tend to be both ex- pressive and instrumental. This possibility should be relatively easy to fol- low up in studies of women's and men's friendships that observe or otherwise catalog in detailed and extensive ways what actually happens when friends get together.

REFERENCES

Aries, E. J. , & Johnson, F. L. (1983). Close friendships in adulthood: Conversational content between same-sex friends. Sex Roles, 9, 1183-1196.

Bell, R. R. (1981a). Worlds of friendship. Beverly Hills, CA: Sage. Bell, R. R. (1981b). Friendships of women and men. Psychology of Women Quarterly, 5, 402-417. Bern, S. L, (1974). The measurement of psychological androgyny. Journal o f Consulting and

Clinical Psychology, 47, 152-162. Bern, S. L. (1981). A manual for the Bern Sex Role Inventory. Palo Alto, CA: Consulting

Psychologists Press. Block, J. D. (1980). Friendship: How to give it, how to get it. New York: Macmillan. Booth, A. (1972). Sex and social participation. American Sociological Review, 37, 183-192. Booth, A., & Hess, E. (1974). Cross-sex friendship. Journal of Marriage and the Family, 36, 38-47. Caldwell, M. A. , & Peplau, L. A. (1982). Sex differences in same-sex friendships. Sex Roles,

8, 721-732. Davidson, S., & Packard, T. (1981). The therapeutic value of friendship between women. Psy-

chology of Women Quarterly, 5, 495-510. Deaux, K. (1984). From individual differences to social categories: Analysis of a decade's research

on gender. American Psychologist, 39, 105-116. Fischer, J. L., & Narus, L. R. (1981). Sex roles and intimacy in same and other sex relation-

ships. Psychology of Women Quarterly, 5, 444-455. Hacker, H. M. (1981). Blabbermouths and clams: Sex differences in self disclosures in same-sex

and cross-sex dyads. Psychology of Women Quarterly, 5, 385-401.

Page 16: Gender role orientations and friendship: Some attenuation, but gender differences abound

566 Wright and Scanlon

Hill, C. T., & Stull, D. E. (1981). Sex differences in effects of social and value similarity in same-sex friendship. Journal of Personality and Social Psychology, 41, 488-5052.

Johnson, F. L., & Aries, E. J. (1983). The talk of women friends. Women's Studies Interna- tional Forum, 6, 353-361.

Lavine, L. O., & Lombardo, J. P. (1984). Self-disclosure: Intimate and nonintimate disclosures to parents and best friends as a function of Bern sex-role category. Sex Roles, 11, 735-744.

Lea, M. (1989). Factors underlying friendship: An analysis of responses on the Acquaintance Description Form in relation to Wright's friendship model. Journal of Social and Per- sonal Relationships, 6, 275-292.

Lombardo, J. P., & Lavine, L. O. (1981). Sex-role stereotyping and patterns of self-disclosure. Sex Roles, 7, 403-411.

McAdams, D. P., Healy, S., & Kraus, S. (1984). Social motives and patterns of friendship. Journal o f Personality and Social Psychology, 47, 828-838.

Rose, S. M. (1985). Same- and cross-sex friendships and the psychology of homosociality. Sex Roles, 12, 63-74.

Rubin, L. (1985). Just friends. New York: Harper & Row. Sherrod, D. (1989). The influence of gender on same-sex friendships. In C. Hendrick (Ed.),

Close relationships. Newberry Park, CA: Sage. Tognoli, J. (1980). Male friendship and intimacy over the life span. Family Relations, 29, 273-279. Underwood, B. J. (1957). Psychological research, new York: Appleton-Century-Crofts. Walker, L. S., & Wright, P. H. (1976). Self-disclosure in friendship. Perceptual and Motor

Skills, 42, 735-742. Weiss, L., & Lowenthal, M. F. (1975). Life course perspective on friendship. In M. Thurnher

& D. Chiriboga (Eds.), Four stages of life. San Francisco: Jossey-Bass. Williams, D. G. (1985). Gender, masculinity-femininity, and emotional intimacy in friendship.

Sex Roles, 12, 587-600. Winstead, B. A. (1988). Sex differences in same-sex friendships. In V. J. Derlega & B. A. Win-

stead (Eds.), Friendship and social interaction. New York: Springer-Verlag. Winstead, B. A., & Derlega, V. J. (1984). The therapeutic value of same-sex friendships. Paper

presented at the annual meeting of the Southwestern Psychological Association, New Orleans, LA.

Wright, P. H. (1982). Men's friendships, women's friendships, and the alleged inferiority of the latter. Sex Roles, 8, 1-20.

Wright, P. H. (1984). Self-referent motivation and the intrinsic quality of friendship. Journal o f Social and Personal Relationships, 1, 115-130.

Wright, P. H. (1985). The acquaintance description form. In S. Duck & D. Perlman (Eds.), Understanding personal relationships: an interdisciplinary approach. London: Sage.

Wright, P. H. (1988). Interpreting research on gender differences in friendship: A case for moder- ation and a plea for caution. Journal of Social and Personal Relationships, 5, 367-373.

Wright, P. H. (1989). Gender differences in adults' same- and cross-gender friendships. In R. G. Adams & R. Blieszner (Eds.), Older adult friendship: Structure and process. New- berry Park, CA: Sage.

Wright, P. H., & Bergloff, P. J. (1984). The acquaintance description form and the study of relationship differentiation. Paper presented at the Second International Conference on Personal Relationship, Madison, WI.

Wright, P. H., & Crawford, A. C. (1981). Agreement and friendship: A close look and some second thoughts. Representative Research in Social Psychology, 2, 52-69.

Yoon, G. H. (1978). The natural history of friendship: Sex differences in best friendship pat- terns. Dissertation Abstracts, 39(3-B), 1553.