Content uploaded by Jeffrey A. Hall
Author content
All content in this area was uploaded by Jeffrey A. Hall on Mar 25, 2016
Content may be subject to copyright.
http://spr.sagepub.com/
Relationships
Journal of Social and Personal
http://spr.sagepub.com/content/early/2010/11/25/0265407510386192
The online version of this article can be found at:
DOI: 10.1177/0265407510386192
published online 29 December 2010Journal of Social and Personal Relationships
Jeffrey A. Hall
Sex differences in friendship expectations: A meta-analysis
Published by:
http://www.sagepublications.com
On behalf of:
International Association for Relationship Research
can be found at:Journal of Social and Personal RelationshipsAdditional services and information for
http://spr.sagepub.com/cgi/alertsEmail Alerts:
http://spr.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Article
Sex differences in
friendship expectations:
A meta-analysis
Jeffrey A. Hall
University of Kansas, USA
Abstract
Friendship expectations are prescriptive normative behaviors and highly valued qualities
in ideal same-sex friends. This paper reports the results of five meta-analyses of sex
differences from 37 manuscripts (36 samples, N¼8825). A small difference favoring
females was detected in overall friendship expectations (d¼.17). Friendship expectations
were higher for females in three of four categories: symmetrical reciprocity (e.g., loyalty,
genuineness; d¼.17), communion (e.g., self-disclosure, intimacy; d¼.39), solidarity
(e.g., mutual activities, companionship; d¼.03), but agency (e.g., physical fitness, status;
d¼.34) was higher in males. Overall expectations and symmetrical reciprocity showed
small effect sizes. Medium effect sizes for communion favoring females and for agency
favoring males support predictions of evolutionary theory.
Keywords
Evolutionary psychology, friendship expectations, ideal same-sex friends, meta-analysis,
sex difference
Friendship expectations are cognitive conceptualizations about attributes individuals
would like friends to possess (Bigelow & La Gaipa, 1980; Wiseman, 1986) and
intimacy-producing behaviors individuals would like friends to enact (Fehr, 2004).
Through repeated interactions with several friends, unwritten contracts of behavioral
expectations develop (Wiseman, 1986). Friendship expectations not only play an impor-
tant role in the formation (Bigelow & La Gaipa, 1980; La Gaipa, 1977), maintenance
Corresponding author:
Jeffrey A. Hall, Ph.D., Assistant Professor, University of Kansas, Communication Studies Department, Bailey
Hall, 1440 Jayhawk Boulevard, Room 102, Lawrence, KS 66045-7574, USA
Email: hallj@ku.edu
Journal of Social and
Personal Relationships
000(00) 1–25
ªThe Author(s) 2010
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0265407510386192
spr.sagepub.com
J S P R
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
(Oswald, Clark, & Kelly, 2004), and dissolution (Clark & Ayers, 1993) of friendships,
they represent a standard to which individuals will behaviorally conform to please others
and will use to judge their friends. Violations of friendship norms (Felmlee, 1999),
expectations (Johnson, 2005), and rules (Argyle & Henderson, 1984) can diminish the
quality of friendship and may endanger its continuance (Johnson et al., 2004).
Researchers have long been interested in sex differences in friendship (Bakan, 1966;
Duck & Wright, 1993; Wright, 1982). For such inquiries, Dindia (2006) and Wright (1988)
recommend theoretically grounded and sober interpretation of effect sizes, such as those
provided by meta-analyses (Allen, 2009). Guided by the qualitative reviews of Wright
(2006) and Hartup and Stevens (1997), this manuscript will present the results of a
meta-analysis of sex differences in overall friendship expectations, and four meta-
analyses of sex differences in theoretically derived dimensions of friendship expectations:
symmetrical reciprocity, communion, solidarity, and agency. For each dimension, hypoth-
eses will be informed by past research on friendship expectations and behaviors (e.g.,
Reis, 1998; Wright, 2006) as well as by evolutionary accounts of friendship (e.g., Geary,
Bird-Craven, Hoard, Vigil, & Numtee, 2003; Taylor et al., 2000).
Ideal expectations and sex differences
Ideal friendship expectations define the social meaning or ‘‘essence’’ of friendship
(Hartup & Stevens, 1997, p. 356), and are the standard upon which present and future
friendships are judged (Fehr, 1996). Expectations of friends are cultivated through
experiences with former and present friends, creating a cycle that modifies and rein-
forces existing expectations (Elkins & Peterson, 1993; Wiseman, 1986). Once developed
during adolescence, ideal expectations remain relatively unchanged throughout life
(Bigelow & La Gaipa, 1980; Weiss & Lowenthal, 1975). These ‘‘built up expectations’’
become unspoken social norms for the ideal friend – a friend that individuals may never
have but nonetheless desire and prefer (Wiseman, 1986, p. 196).
Ideal friend expectations are also relevant to evolutionary approaches to friendship.
The ability to estimate the value of particular friendship qualities or behaviors is an
important adaptation for the comprehension of social exchange and altruism (Cosmides
& Tooby, 2005; Tooby & Cosmides, 1996). Ideal friends are desirable because they
possess qualities that yield the greatest long-term benefits to the individual. Actual
friends, however, are not equally endowed with desirable attributes. Just as individuals
have different reproductive value as potential mates, they also have different association
value as friends (Kendrick, Maner, & Li, 2005; Tooby & Cosmides, 1996). The partic-
ular qualities sought in an ideal friend will differ depending upon the needs of the indi-
vidual. Biological sex may influence which qualities in ideal friends are most needed.
Specifically, sex differences in expectations are likely to have evolved in cases where
females and males differ in outcomes needed or desired from friendship associations
(MacDonald, 1996). When one type of friendship expectation offers benefits to or serves
the needs of females more than males, females should place greater value on these beha-
viors or qualities in an ideal friend.
Although sex differences in friendship expectations have not been systematically
assessed, females appear to hold higher expectations for same-sex friends than males
2Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
(Clark & Ayers, 1993; Fuhrman, Flannagan, & Matamoros, 2009; Sapadin, 1988). As a
consequence of greater expectations, females’ same-sex friendships provide more emo-
tional and informational support (Hays, 1989), experience more self-disclosure and trust
(Jones, 1991), and engage in more support, openness, and interaction maintenance beha-
viors than males’ same-sex friendships (Oswald et al., 2004). Because expectations have
been defined in a variety of ways across studies, it is difficult to generate a prediction
derived from evolutionary theory for overall sex differences. Instead, particular cate-
gories of expectations are more likely to demonstrate sex-specific evolved preferences
due to their match to particular stress and resource-attainment processes (discussed
below). To verify the magnitude and direction of sex difference in overall expectations
found in past research, we predict:
H
1
: Females will have higher overall friendship expectations than males.
Symmetrical reciprocity: trust, loyalty, genuineness, commitment
The possible dimensions of expectations are manifold: some studies have suggested as
few as two (Zarbartany, Conley & Pepper, 2004) and others in excess of 20 dimensions
(Bigelow, 1971; La Gaipa, 1987). The majority of expectation studies have explored
relational and emotional qualities of friendships (e.g., Bigelow & La Gaipa, 1980), but
others have included material benefits of friendships (e.g., wealth, business connections)
(Lusk, MacDonald, & Newman, 1998; Vigil, 2007). This manuscript will propose four
dimensions of expectations.
The first proposed dimension is symmetrical reciprocity, which includes the most
important features of friendship: loyalty, mutual regard or authenticity, trustworthiness,
and support (Hartup & Stevens, 1997; Rawlins, 2009; Sapadin, 1988). Loyalty, trust, and
support are the most prototypical behaviors in producing intimacy (Fehr, 2004), and can
distinguish a close friend from an acquaintance (Newcomb & Bagwell, 1995). Sym-
metrical reciprocity is highly valued by both sexes (Hartup & Stevens, 1997). Regardless
of activities enjoyed, friendship origination and development, and sex composition of the
dyad, these qualities are present in best friendships (Sapadin, 1988; Wright, 2006). These
qualities are both necessary and sufficient conditions for close friendship; without them a
close friendship would not exist (La Gaipa, 1987; Wright, 2006). Hartup and Stevens
(1997) argue there should not be sex differences in friendship expectations for sym-
metrical reciprocity because both males and females seek out, maintain, and find con-
siderable value in best friendships marked by trust, commitment, and genuineness.
In contrast, evolutionary accounts suggest that females should have higher expecta-
tions for reciprocal altruism (Geary et al., 2003). Although Hartup and Stevens’ (1997)
concept is not synonymous with reciprocal altruism, both share trust and honesty as
central characteristics. Geary et al. (2003) argue that because female coalitions were
more likely to be unrelated to one another throughout human history (i.e., non-kin),
females would be more likely to develop mechanisms of obtaining mutual and reciprocal
benefits from unrelated others (i.e., female friends). This would produce less tolerance
for non-reciprocity in friendships, and higher expectations of trust and loyalty from ideal
friends. On the other hand, male coalitions were more competitive and non-reciprocal,
Hall 3
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
due to associating primarily with hierarchically differentiated male kin. Under these con-
ditions, reciprocity would be a critical component of female–female friendships, but less
so for male–male friendships (Geary et al., 2003). To test competing hypotheses, the
evolutionary prediction of sex difference will be hypothesized:
H
2
: Females will have higher friendship expectations of symmetrical reciprocity.
Communion: intimacy, self-disclosure, and empathic understanding
Communion is a traditional dimension of friendship that includes emotional availability
and self-disclosure given and received (Bakan, 1966, Wright, 1988). Intimacy, emo-
tional disclosure, and empathic understanding are conceptually related and often highly
correlated in friendship (Hussong, 2000; Laurenceau, Rivera, Schaffer, & Pietromonaco,
2004). Communality is distinct from symmetrical reciprocity, in that aspects of symme-
trical reciprocity can be achieved through emotional self-disclosure, but without trust
and loyalty, disclosure is rarely attempted (Fehr, 1996, Hartup & Stevens, 1997).
In Reis’s (1998) meta-analyses of sex-differences in friendship behaviors, women expe-
rienced more intimacy than men (d¼.66). Intimacy is also more highly valued by girls
in comparison with boys (Bigelow & La Gaipa, 1980; La Gaipa, 1987). Differences in
communion do not result from differences in the definition or desirability of intimacy.
Both sexes agree on how intimacy and closeness are defined in friendship, and the value
of intimacy in friendship (Fehr, 2004; Sapadin, 1988).
Evolutionary accounts have suggested that sex differences in communion may have
resulted from evolved mechanisms of managing stress (Taylor et al., 2000). The tendency
toward a tend-and-befriend response to stress evolved in females due to asymmetric invest-
ment in offspring and a need to form and maintain protective female coalitions. The
befriending response is supported by females’ greater use of emotional support from
friends to manage stress (Tamres, Janicki, & Helgeson, 2002). Because ideal friends must
be able to meet individuals’ needs, females would desire friends capable of meeting high
communion expectations. Therefore:
H
3
: Females will have higher friendship expectations of communion than males.
Solidarity: inclusion, mutual activities, and companionship
This paper will use Hartup and Stevens’ (1997) term solidarity to refer to sharing mutual
activities and the companionship of friends. Sharing common activities is one of the first
expectations to develop in children (Bigelow & La Gaipa, 1980), and throughout ado-
lescence both sexes spend ample time in the company of same-sex friends (Fehr, 1996).
Time spent together distinguishes close friends from non-friends (Newcomb & Bagwell,
1995). Although males’ friendships have been characterized as more activity focused,
recent reviews of friendship suggest no sex difference in solidarity (see Wright,
2006). Males and females both agree that being with friends for the sake of company and
mutual enjoyment is important (Duck & Wright, 1993; Fehr, 1996), and both sexes
4Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
have similar expectations of inclusion in mutual activities (Clark & Ayers, 1993, Clark &
Bittle, 1992).
If males and females are similar in the value placed on time spent in the company of
friends or the desire to be included (Geary et al., 2003; Taylor et al., 2000), then group-
level dynamics may be a factor in the evolution of sex differences (Kendrick et al.,
2005). When certain expectations are valued more by one sex in comparison with the
other, similar expectations of solidarity reinforce sex differences through behavioral
conformation and peer socialization. Individuals who behaviorally conform to the
expectations of same-sex peers are more likely to be included and liked within friend-
ship networks (Bigelow & La Gaipa, 1980). Because solidarity is valued, individuals
will attempt to possess and demonstrate the characteristics expected by their same-sex
peers. This process reinforces friendship expectations particular to each same-sex
friendship network. We offer the following question:
RQ
1
: Will there be sex differences in solidarity friendship expectations?
Agency: personal and financial resources, status, and reward value
Agency expectations arise when friends are regarded as objects from which benefits can
be derived (Bakan, 1966). Because friendship is typically understood in terms of trust,
disclosure, and companionship, this expectation dimension is not often reported (e.g.,
La Gaipa, 1987). However, friends are often considered desirable because of what they
have or what they offer the individual, especially in public and instrumental friendships
(Rawlins, 2009). A friendship can offer the individual assistance in achieving personal
success and providing access to resources or information (Zarbatany et al., 2004).
Agency expectations pertain to what a friend can do, has access to, and is able to
offer associates. Research in friendship and agency has typically contrasted ideal
expectations for different types of relationships (e.g., romantic partnership vs.
best friend) (Lusk et al., 1998; Sprecher & Regan, 2002). Vigil (2007) includes
intelligence, financial resources, and athletic ability in his measure of personal
capacity. Athleticism and popularity also arise in children’s (Bigelow, 1977) and
adolescents’ friendship expectations (Gonzales et al., 2004). Overall, in comparison
with females, males tend to assign greater value to agency expectations (Lusk et al.,
1998; Vigil, 2007; Zarbartany et al., 2004).
From an evolutionary perspective, there are several reasons why males tend to have
higher agency expectations. Within males’ social networks, friendships reflect a bal-
ance of cooperative and competitive behaviors (Geary & Flinn, 2002; Geary et al.,
2003). Men and boys are more likely to organize their social environments hier-
archically, wherein competition between males, once resolved, results in a stable
coalition structure (Geary et al., 2003). By asserting a place within a hierarchy, males
are assured greater access to external resources derived from cooperative male
activities (e.g., hunting, victory in sports). Because group unity and cooperation is
necessary to achieve group-level goals, the stabilizing bonds of friendship are neces-
sary. By befriending a higher status male associates are granted special access to
Hall 5
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
influential persons and high quality resources (Geary et al., 2003; Kendrick et al.,
2005). Therefore:
H
4
: Males will have higher friendship expectations of agency than females.
Five meta-analyses will be conducted to evaluate the existence and magnitude of sex dif-
ferences. Biological arguments of difference will be examined by testing the moderating
effects of age and race/ethnicity. If sex differences are present and moderated by age,
then cultural and peer influence processes are likely operative, but if no age moderation
is found then differences remain constant over time. If sex differences are moderated by
race/ethnicity (i.e., percent of sample non-White), then there is evidence for cultural
variation in sex differences. If there is a lack of moderation, then sex differences are
similar between cultures and more likely to be biologically based. In addition to these
sample-based moderators, the publication source of the manuscript (i.e., published vs.
unpublished), year of publication, and measurement of expectations (i.e., quantitative
vs. qualitative) will be tested to rule out other explanations of variation in sex differences.
Method
Study identification
Study identification began by searching PsychINFO,Dissertation Abstracts,andERIC
with the key term ‘‘friend*’’ with ‘‘expect*’’, ‘‘desir*’’,’’ ideal’’, or ‘‘need*’’. A subse-
quent search for ‘‘friend*’’ and ‘‘human sex differences’’ was also conducted. Quantitative
studies using methods approximating the working definition prescriptive characteristics of
ideal same-sex friends were included. Studies where participants rated the importance of
ideal friendship characteristics, and where wanted, desired, or needed qualities in friends
were identified or rated were included. Relevant studies were then used to conduct a for-
ward search using Web of Science. This process identified additional articles that cited the
articles included in the sample.
Studies were excluded if they met one of several criteria. First, studies that evaluated
how friends actually were, actually behaved, or were predicted to be or to behave, but
failed to evaluate how ideal friends ought to or should behave were excluded (e.g.,
Bukowski, Nappi, & Hoza, 1987; Lempers & Clark-Lempers, 1993; Mancini & Simon,
1984; Roy, Benenson, & Lilly, 2000). Second, the present study explored ideal expecta-
tions, not expectation violations, and Lipsey and Wilson (2001) caution against combin-
ing experimental and non-experimental studies. Therefore, studies were excluded if they
evaluated confirmations or violations of friendship expectations through experimental
methods (e.g., Felmlee, 1999; Felmlee & Muraco, 2009; Flannagan et al., 2005; Johnson,
2005). Third, studies were excluded if they rated the relative importance of qualities of
friendship in general but not for ideal friends, or friends in general but not same-sex
friends specifically (e.g., Cann, 2004; Horenczyk & Tatar, 1998; Sapadin, 1988). Fourth,
studies that evaluated same-sex friendships for only one sex (e.g., either for males or
females only) were excluded because no sex differences could be estimated. Finally,
studies were excluded when sex differences were not reported in any way that allowed
the data to be imputed, and contact with the authors or attempts to recover the data to
6Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
compute effect sizes proved futile (Banikiotes, Neimeyer, & Lepkowsky, 1981; Burgio,
1987; Cheng, Bond, & Chan, 1995; Weiss, & Lowenthal, 1975). In the final set of rel-
evant studies, 36 unique effect sizes (i.e., 36 distinct samples) were calculated from 37
manuscripts. Three manuscripts reported two samples (i.e., Chance, 2008; Fuhrman
et al., 2009; Zarbatany et al., 2004). Five manuscripts were written using two unpub-
lished datasets (i.e., Bigelow, 1971, 1977). For the meta-analyses, the unpublished data-
sets were used.
Effect sizes
Meta-analytic procedures combine effect sizes from multiple studies treating each study
as an observation. Cohen’s (1988) dstatistic was computed for each sample, which is the
recommended effect size statistic when group comparisons of quantitative variables are
conducted (Johnson, Scott-Sheldon, Snyder, Noar, & Huedo-Medina, 2008). Positive
dvalues were reported when differences favored females and negative values when
differences favored males. Twenty-one of 36 effect sizes were calculated from sample
means and standard deviations provided in the manuscript or through contact with the
lead author. Fourteen effect sizes were calculated from sample size and tor Fstatistics,
and one effect size was calculated from Pearson’s r. In cases where pvalues were pro-
vided but tor Fstatistics were not, lower bound estimates of effect sizes were estimated
(i.e., the lowest possible effect size given the significance level and sample size). When
descriptive statistics were not provided and significant differences were not reported, a
zero effect size was assumed. These conservative approaches to detecting sex differences
mean that actual effect sizes were likely somewhat larger than those reported here
(Lipsey & Wilson, 2001).
Effect sizes and standard errors were corrected for the artifact of unreliability to
decrease the likelihood of misrepresenting sex differences (Allen, 2009). Artifact cor-
rection permits the estimation of effect sizes as they would appear under ideal research
circumstances (Lipsey & Wilson, 2001). In studies where reliability was reported, either
by Cronbach’s alpha or Cohen’s intercoder kappa, artifacts were corrected. In studies
where single item measures were used, there was no artifact correction. In studies where
reliability was not reported but a multiple item measure was used, the average reliability
of all studies reporting reliability was imputed for artifact correction (M
alpha
¼.765).
The present investigation sought to explore sex differences in ideal friendship
expectations overall, and sex differences in four categories of expectations. When a
single effect size was reported, the study was included only in the overall expectation
meta-analysis. When the study measured multiple types of expectations, multiple effect
sizes were calculated. For each study, when sex differences for an overall expectation
measure were not reported, an average effect size was calculated by aggregating the
effect sizes of all categorical expectations (Lipsey & Wilson, 2001). All 36 studies pro-
vided one effect size for overall friendship expectations.
Four categories of expectations were identified, categorical definitions were created,
and two independent coders identified categorical expectations (kappa ¼.97). When
expectations failed to be included in any of four categories by both coders, rather than
forcing a misfit, these effects were excluded from categorical meta-analyses. Less than
Hall 7
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
7%of possible categorical expectation effect sizes were excluded. Finally, some studies
reported sex differences for two separate sub-constructs that belonged to a single expec-
tation category. To ensure that each study contributed only one categorical effect size to
a categorical meta-analysis, the effect sizes for two sub-constructs were averaged before
included. For example, the sex differences in empathic accuracy and self-disclosure
reported separately by Amalfitano (1980) were averaged to form a single effect size for
communion. See Table 1 for all studies, expectation categories, and effect sizes.
Moderators
Five study characteristics were coded as potential moderating variables. Two moderators
were chosen because of their theoretical relevance to the study of sex difference: mean
age of sample, percent of sample non-White. One moderator explored the measurement
methods employed. Expectations measured on numeric scales were coded as quantita-
tive, and expectations measured by frequency counts of the presence or absence of
expectations in text or transcribed interviews were coded as qualitative. Publication
d Overall Expectations Symmetrical Communion Solidarity Agency
2.4 8
.
.
.
1.5 7
1.4 7
1.3 7
1.2
1.1 1 2
1
.9 38 3
.8 3 8
.7 3 8 59 9
.6 4 17
.5 9 6
.4 1117 249 58 1
.3 39 5 23557 0
.2 1235567 6 577 7
.1 1 445678 47
.0 124479 114799 0 0000 0
.0 116 277 2 225666 19
.1 5 25
.2 4 69
.3 6 6
.4 26
.5
.6 8
k¼36 k¼21 k¼31 k¼21 k¼9
Figure 1. Stem-and-leaf plots of sex differences.
8Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. Summary of all studies (positive effect size values where sex differences favor women)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Amalfitano, 1980 13 25 56.5 N/A No Yes Overall 7 Friendship Expectances (La Gaipa, 1979) 0.41
Reciprocity Support, Genuineness, Positive Regard 0.49
Communion Self Disclosure & Empathic Understanding 0.40
Solidarity Similarity 0.05
Argyle & Henderson,
1984; Study 1
30 30 30.0 N/A Yes Yes Overall 14 Relationship Rules (Author Derived) 0.47
Bank, 1994 553 622 14.0 N/A Yes Yes Overall Closeness Norms & Assertiveness Norms
(adapted Argyle & Henderson, 1984)
0.06
Communion Closeness norms 0.14
Agency Assertiveness norms 0.26
Basu & Mukhopad-
hyay, 1986
30 30 20.0 1.00 Yes No Overall 15 Single Item expectancies
(Author derived)
0.27
Reciprocity Honest, Loyal, Genuine, Understanding,
Sacrificing, Reliable
0.40
Communion Open communication 0.16
Solidarity Similar interests 0.07
Agency Physical attractiveness 0.01
Bigelow, 1971 240 240 10.0 N/A No No Overall 21 expectancies (Author derived) 0.04
Reciprocity Reciprocity, Acceptance, Loyalty,
Genuineness, Support
0.09
Communion Intimacy Potential 0.37
Solidarity Common activities, Play, Prior
interaction
0.02
Bigelow, 1974 240 240 10.0 N/A Yes No Overall 21 expectancies (Bigelow, 1971) 0.01
Reciprocity Reciprocity, Acceptance, Loyalty,
Genuineness, Support
0.01
Communion Intimacy Potential 0.20
Solidarity Common activities, Play, Prior interaction 0.06
(continued)
9
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. (continued)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Broderick & Beltz,
1996
105 95 12.2 N/A Yes Yes Overall 10 expectations (Furman & Bierman, 1984) 0.41
Reciprocity Dispositional & Behavioral Support &
Liking
0.56
Communion Behavioral Intimacy 0.79
Solidarity Dispositional & Behavioral Association 0.00
Chance, 2008; Adult
Sample
24 25 20.1 0.65 No Yes Overall 3 Expectations (Author derived) 0.98
Communion Emotional Dependence 1.37
Solidarity Social Solidarity 0.79
Chance, 2008; Child
Sample
12 12 12.8 0.17 No Yes Overall 3 Expectations (Author derived) 1.11
Communion Emotional Dependence 1.57
Solidarity Social Solidarity 0.88
Claes, 1992 170 179 15.9 N/A Yes Yes Overall 4 Friendship Expectances (La Gaipa, 1979) 0.26
Reciprocity Acceptance & Support 0.35
Communion Empathic Understanding 0.35
Solidarity Mutual Activities 0.00
Clark & Ayers, 1993 111 131 12.6 0.27 Yes Yes Overall 4 Friendship Expectances (La Gaipa, 1979) 0.25
Reciprocity Acceptance & Support 0.26
Communion Empathic Understanding 0.45
Solidarity Mutual Activities 0.00
Clark & Bittle, 1992 101 97 10.0 0.26 Yes Yes Overall 4 Friendship Expectances (La Gaipa, 1979) 0.41
Reciprocity Acceptance & Support 0.40
Communion Empathic Understanding 0.83
Solidarity Mutual Activities 0.00
Craven, 1992 108 102 11.5 0.00 No Yes Overall Communality, Empathy, Loyalty
(Author derived)
0.11
Reciprocity Loyalty 0.02
Communion Intimacy & Empathic Understanding 0.48
Solidarity Common or shared activities 0.12
10
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. (continued)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Elkins & Peterson,
1993
58 65 19.0 0.10 Yes Yes Overall Wright’s Acquaintance Description Form 0.93
Fehr, 2004; Study 3 36 53 21.4 N/A Yes Yes Overall Intimacy Expectancies (Author derived) 0.33
Communion Intimacy 0.33
Fuhrman et al. 2009;
Sample 1
151 248 19.8 0.60 Yes Yes Overall Emotional Closeness, Social
Companionship; Relational Positivity
(adapted Argyle & Henderson, 1984)
0.59
Communion Emotional Closeness 0.17
Solidarity Social Companionship 0.17
Fuhrman et al. 2009;
Sample 2
75 90 19.8 0.59 Yes Yes Overall Emotional Closeness, Social
Companionship; Relational Positivity
(adapted Argyle & Henderson, 1984)
0.64
Communion Emotional Closeness 0.27
Solidarity Social Companionship 0.30
Fujiwara, 2004 80 156 21.2 0.35 No Yes Overall 10 expectancies (Derived from
La Gaipa)
0.04
Reciprocity Reciprocity, Loyal, Genuine,
Honesty, Accepting, Trust, Support
0.04
Communion Warmth and Openness 0.20
Furman & Bierman,
1984
42 42 9.0 N/A Yes No Overall 10 Expectations (Author derived) 0.00
Reciprocity Dispositional & Behavioral Support &
Liking
0.00
Gonzales et al., 2004 297 297 13.5 0.50 Yes No Overall 21 Expectancies (Bigelow, 1974) 0.00
Reciprocity Reciprocity, Genuineness, Admiration,
Help
0.00
Communion Intimacy Potential 0.18
Solidarity Common activities, Play, Prior
interaction
0.02
Agency Physical attractiveness, Athleticism,
Education
0.00
(continued)
11
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. (continued)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Hall et al., 2009 100 97 20.7 0.12 No Yes Overall Friendship maintenance
(Oswald et al., 2004)
0.64
Reciprocity Support Maintenance 0.78
Communion Positivity & Openness 0.67
Solidarity Interaction 0.41
Hartmann, 1991 45 45 14.1 N/A No No Overall Total expectations & Intimacy, Self
Disclosure, Understanding
(Author derived)
2.48
Communion Intimacy, Self Disclosure, & Mutual
Understanding
1.47
Johnson, 1997 39 33 23.0 0.22 No Yes Overall Expectations of Emotional & Instrumental
Support (Author derived)
0.09
Communion Emotional Support 0.32
Landberg, 1982 157 196 14.8 N/A No No Overall 7 Friendship Expectances
(La Gaipa, 1979)
0.39
Reciprocity Help & Support, Genuineness, Ego
Reinforcement
0.42
Communion Intimacy & Empathic Understanding 0.35
Solidarity Similarity 0.27
Lusk et al., 1998 93 279 24.2 N/A Yes Yes Overall 6 expectations: Attractiveness,
Athleticism,
Education, Ability, Intimacy,
Conscientiousness (Author derived)
0.10
Reciprocity Conscientiousness 0.09
Communion Intimacy & Warmth 0.75
Agency Athleticism 0.31
McDougall & Hymel,
2007
84 90 13.3 0.12 Yes No Overall 28 Expectations (adapted
Bigelow, 1974)
0.02
Reciprocity Reciprocity, Acceptance, Loyalty,
Genuineness, Support
0.07
Communion Intimacy Potential 0.14
Solidarity Companionship, social activities 0.06
12
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. (continued)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Migliaccio, 2002 98 102 34.0 0.38 No Yes Overall 9 Expectances (Personal Attributes
Questionnaire)
0.21
Reciprocity Trust & Reliability 0.20
Communion Share thoughts, Share feelings,
Emotional Availability
0.27
Solidarity Shared activities and experiences 0.14
Ray & Cohen, 1996 92 87 12.0 0.50 Yes Yes Overall 10 Expectations (Furman &
Bierman, 1984)
0.15
Reciprocity Dispositional & Behavioral Support &
Liking
0.07
Communion Behavioral Intimacy 0.02
Solidarity Dispositional & Behavioral Association 0.36
Schneider & Tessler,
2007
40 36 11.4 0.05 Yes No Overall 21 Expectations (Bigelow, 1974) 0.01
Communion Intimacy Potential 0.25
Agency Sportsmanship 0.46
Sprecher & Regan,
2002
59 80 20.2 N/A Yes Yes Overall 14 Expectations (Author derived) 0.07
Communion Warmth & Expressiveness 0.61
Solidarity Similar and shared activities 0.06
Agency Social status, intelligence, wealth,
physical attractiveness
0.09
Verkuyten & Masson,
1996
114 100 16.0 0.67 Yes No Overall Rules of exchange, intimacy,
coordination, relations (adapted
Argyle & Henderson, 1984)
0.22
Reciprocity Rules of Trust & Confidence 0.44
Communion Rules of Relations 0.00
(continued)
13
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Table 1. (continued)
Study
SS
Male
SS
Female
Mean
Age
Percent
Non-white Published Quantative
Expectation
Type Expectation Measurement
Effect
Size
Vigil, 2007 88 163 18.6 N/A Yes Yes Overall 12 Single-item measures (Author derived) 0.24
Reciprocity Kindness & Cooperation 0.07
Agency Social status, intelligence, wealth,
physical attractiveness, athleticism,
education
0.29
Wood, 1972 215 173 12.0 N/A No Yes Overall 4 Friendship Expectances
(La Gaipa, 1979)
0.01
Reciprocity Acceptance & Support 0.01
Communion Empathic Understanding 0.15
Solidarity Mutual Activities 0.15
Zarbatany et al., 1992 28 34 10.5 N/A Yes No Overall 21 Expectations (modified from
La Gaipa)
0.00
Zarbatany et al.,
2004; Adult
Sample
95 217 19.5 N/A Yes Yes Overall 2 expectations: Agency, Communality
(Author derived)
0.23
Communion Self Disclosure, Emotional Support,
Love, Nurturance, Companionship
0.93
Agency Status, connections, independence,
competition, power
0.40
Zarbatany et al.,
2004; Child Sample
222 269 11.9 0.25 Yes Yes Overall 2 expectations: Agency, Communality
(Author derived)
0.25
Communion Self Disclosure, Emotional Support,
Love, Nurturance, Companionship
1.12
Agency Status, connections, independence,
competition, power
0.65
14
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
source (i.e., published vs. unpublished manuscript) and year of publication were also
coded. Coding of mean sample age, percent non-White, methods used, publication
source, and year of publication by two independent coders was reliable (kappa ¼1.00).
Heterogeneity
The effect sizes were weighted by the inverse of the sampling error variance (Lipsey &
Wilson, 2001). Combined effect sizes were evaluated for heterogeneity. Effects were
considered homogenous when the variance could be attributed to sampling error alone
(i.e., Qstatistic was non-significant). When effects were considered homogenous, a fixed
effects model was presumed and no moderators were explored. When effect sizes were
heterogeneous (i.e., significant Qstatistic), moderating analyses were conducted.
Results
Effect sizes distribution
The final sample included 8825 participants (45.8%male). Participants were on average
15.7 years of age (M
sample range
¼10 to 56.5 years). Sample race/ethnicity was reported in
19 samples, wherein 36.8%of participants were non-White. Two-thirds of the studies
were published (N¼24). Quantitative measurement methods were used in 69%of stud-
ies (N¼25). Studies were published between 1971 and 2009 (M
year of publication
¼1996).
A stem and leaf plot presents the distributions of effect sizes for all expectations (see
Figure 1). A single outlier was identified for overall expectations (i.e., Hartmann,
1991). Rather than exclude the data from the outlying study, the main effect for this study
was Winsorized (Lipsey & Wilson, 2001). That is, the main effect for overall expecta-
tions for this study was recoded to be equivalent to the next closest effect size.
The weighted mean effect size for overall ideal friendship expectations was d¼.170,
P(.123 < r< .217) ¼.95, k¼36, N¼8825. This demonstrates that females were
significantly more likely than males to have higher overall expectations for their ideal
friends. The effect sizes demonstrated significant heterogeneity, Q¼126.36, df ¼35,
p< .001, indicating that effect sizes varied across studies. Therefore, moderation
analyses were conducted using weighted regression procedures (Lipsey & Wilson,
2001). Sample age (ß ¼.012, Q
age
¼7.57, df ¼1, p< .001) was related to effect size,
wherein older samples were more likely to have larger effect sizes than younger
samples. For samples 5 years younger than the mean sample age of 15.7 years, the
mean effect size was estimated d¼.134, and for participants 5 years older than
the average age, the mean effect size was estimated d¼.250. Sample race/ethnicity
(ß ¼.150, Q
ethn
¼2.54, df ¼1, p¼ns) did not moderate sex differences in overall
expectations. The year of study (ß ¼.005, Q
year
¼6.03, df ¼1, p< .001) was a sig-
nificant moderator, wherein more recent studies were more likely to have larger effect
sizes. Publication source was not a significant moderator (ß ¼.10, Q
pub
¼3.16, df ¼1, ns).
Finally, the measure of expectations did not moderate sex differences (ß ¼.077,
Q
quant
¼2.13, df ¼1, ns). Multiple regression was conducted using mean sample
age and year of publication as moderators of effect size, and both were unique
moderators: M
year
Z¼2.36, p<.05;M
sample age
Z¼3.55, p< .001.
Hall 15
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Symmetrical reciprocity
Symmetrical reciprocity was the first categorical effect size estimated. H
2
argued that in
comparison with males, females would hold higher expectations of trust, loyalty, com-
mitment, genuineness, and acceptance. The weighted mean effect size for symmetrical
reciprocity expectations was d¼.172, P(.111 < r<.232)¼.95, k¼21, N¼5,499,
indicating that females had higher expectations of symmetrical reciprocity. The effect
sizes demonstrated significant heterogeneity, Q¼50.83, df ¼20,p< .001, therefore
moderation analyses were conducted. Sample age (ß ¼.009, Q
age
¼3.48, df ¼1, p<.10)
and race/ethnicity (ß ¼.075, Q
ethn
¼1.90, df ¼1, ns) were not significant moderators of
sex differences in symmetric reciprocity. The year of study (ß ¼.002, Q
year
¼.753, df ¼1,
ns) and publication source (ß ¼-.080, Q
pub
¼1.63, df ¼1, ns) were not significant
moderators. The method of measurement was not a significant moderator (ß ¼.078, Q
quant
¼1.70, df ¼1, ns). Despite significant heterogeneity, moderators did not further explain
variance in sex differences in symmetrical reciprocity.
Communion
Communion expectations included intimacy, self-disclosure, empathic understanding, and
emotional support. H
3
predicted that females would expect more communion from ideal
friends than males. The weighted mean effect size for communion expectations was signif-
icant, d¼.388, P(.339 < r< .437) ¼.95, k¼31, N¼8245, indicating females had higher
expectations of communion. The effect sizes demonstrated significant heterogeneity, Q¼
170.39, df ¼31, p< .001, therefore moderation analyses were conducted. Mean sample age
(ß ¼.004, Q
age
¼.73, df ¼1, ns) and race/ethnicity (ß ¼.312, Q
ethn
¼3.55, df ¼1, ns)were
not significant moderators of sex differences in communion. The year of study was a signif-
icant moderator of communion (ß ¼.007, Q
year
¼10.25, df ¼1, p< .001), as was the method
of measurement (ß ¼.253, Q
quant
¼18.77, df ¼1, p< .001). These moderations demonstrated
that more recent studies and those collected by quantitative methods reported larger
differences in communion. Finally, publication source (ß ¼.011, Q
pub
¼.039,
df ¼1, ns) was not a moderator of communion. Multiple regression analysis indicated
that year of study and method of data collection were unique moderators of effect sizes:
M
year
Z¼5.23, p< .001; method of collection Z¼8.41, p< .001.
Solidarity
Solidarity expectations were defined as the inclusion of friends in shared activities, time
spent together, and similarity. RQ
1
explored the existence of sex differences in soli-
darity. The weighted mean effect size for solidarity expectations was not significant
because it included zero in the confidence interval, d¼.033, P(.029 < r< .095) ¼.95,
k¼21, N¼5118. Furthermore, the 95%confidence interval indicated a range of dif-
ference that could at most be considered trivial. In response to RQ
1
, females and males
were equally likely to expect solidarity from ideal friends. The effect sizes did not
demonstrate significant heterogeneity, Q¼30.66, df ¼20, ns. Therefore, the effects
were considered homogeneous.
16 Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Agency
Agency measured the degree to which wealth, status, physical attractiveness and fitness,
and intelligence and education were expected from ideal friends. H
4
predicted that males
would expect more agency than females. The weighted average effect size for agency
expectations was significant, d¼.337, P(.420 < r<.255) ¼.95, k¼9, N¼3470,
indicating that males held higher expectations of agency. Effect sizes did not demon-
strate significant heterogeneity, Q¼13.09, df ¼8, ns. Therefore, effects were consid-
ered homogeneous.
Publication bias
To determine whether unpublished studies biased the meta-analyses, a fail-safe Nwas
calculated for each significant mean effect size using Orwin’s (1983) formula. The
criterion effect size was set at |.07|. This is the largest mean effect size including zero
in the 95%confidence interval assuming a standard error equivalent to the largest stan-
dard error found in the present study’s meta-analyses. The smallest fail-safe Nwas for
symmetrical reciprocity (k¼30). That is, 30 unpublished studies with no sex difference
would have to exist to reduce the mean effect size of symmetrical reciprocity to include a
zero in the 95%confidence interval. To reduce the remaining mean effect sizes to non-
significance, the fail-safe Nvalues were 140 for communion, 51 for overall expectations,
and 34 for agency. Given that the lowest fail-safe Nis nearly equivalent to the total num-
ber of studies analyzed in this manuscript, it is unlikely that a significant publication bias
exists. Publication bias is also unlikely because one-third of studies sampled were
unpublished (Lipsey & Wilson, 2001).
Discussion
The meta-analyses demonstrated that expectations of ideal same-sex friends were
slightly greater for females in comparison with males for overall expectations, and
slightly greater for females for the most important expectation in friendship, namely
symmetrical reciprocity. Meta-analyses demonstrated medium-sized sex differences in
communion and agency. The theoretical relevance of these results must be explored
in relation to the effect sizes and moderators to prevent artificially amplifying differ-
ences between the sexes (Dindia, 2006; Wright, 1988).
Overall friendship expectations (d¼.170) and symmetrical reciprocity (d¼.172),
the category including the most central components of friendship, showed small
differences favoring females. The effect size for overall expectations included both
communion and agency, which demonstrated medium sex differences in opposite
directions. This may have offset any sex differences detected in overall expectations.
These differences suggest that males and females are more likely to differ in particular
types of expectations, rather than in overall expectations.
The results relating to symmetrical reciprocity showed some support for evolutionary
accounts of sex differences (Taylor et al., 2000), but provided more support for Hartup
and Stevens’ (1997) argument that males and females may differ in pathways toward
friendship, but similarly value deep structures of friendship. Since less than 3%of the
Hall 17
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
variance in symmetrical reciprocity was explained by sex, differences in expectations are
unlikely to have resulted from evolved mechanisms. This supports Geary and Flinn’s
(2002) critique of Taylor et al. Taylor et al. argue that the mother–child pair bond resulted
in a comforting and befriending adaptation for managing stress. In rebuttal, Geary and
Flinn note that the mother–child relationship is simply not symmetrical, so it is incon-
sistent with the tend-and-befriend hypothesis that females would develop greater expec-
tations for friends in reciprocal altruism. The present study supports Geary and Flinn, who
differentiate adaptive properties of friendship shared by both males and females (e.g.,
symmetrical reciprocity) from those likely to produce sex differences (e.g., communion).
The medium-sized sex difference in communion (d¼.388) is consistent with a past
meta-analysis of females’ greater friendship behavioral intimacy (Reis, 1998) and pro-
vides support for the tend-and-befriend hypothesis (Taylor et al., 2000). It is important
to note that neither age nor race/ethnicity moderated this difference. This suggests that
the sex difference in self-disclosure and intimacy begins at a young age (Bigelow, 1977),
and does not change as children mature. This, in addition to the lack of moderation of
race/ethnicity, offers some support for the argument that these differences are less likely
to be a result of peer modeling or learning, and they may be biologically based. Because
of the strong relationship between what is expected from ideal friends and how actual
friends behave (Hall, Larson, & Watts, 2009), sex differences in communion provide
an explanation for greater behavioral intimacy in female friendships (Reis, 1998). If bio-
logical tendencies increase females’ expectations of intimacy, emotional support, and
empathic understanding from friends, then females may be more likely to choose and
maintain friendships with females who can meet their higher expectations.
The lack of sex differences in solidarity provides meta-analytic support for Wright’s
(2006) contention of similarity in activity orientation for all friendships. This lack of sex
difference is important to both evolutionary and developmental accounts of the role of
same-sex socialization in friendship. Because both sexes place a similar value on inclu-
sion and acceptance by same-sex friends, solidarity expectations may help to produce
and reinforce sex-specific friendship expectations. This meta-analysis provides evidence
of a proximal mechanism for both evolved and developmental sex differences. To con-
firm the socializing influence of solidarity, future research may explore whether greater
inclusion in opposite-sex friendship networks produces behaviors that mimic the expec-
tations of the opposite sex. For example, males accepted in female-dominated friendship
networks may be more likely to value communion over time, and females included in
male-dominated friendship networks may be more likely to value agency over time.
The meta-analysis of the agency expectations offers support for evolutionary
accounts of sex difference, wherein males are more likely than females to seek associ-
ation with high-resource same-sex friends. Because this is the first study to explore
sex differences in agency and the sample of studies was relatively small (i.e., k¼9), the
conclusions of this analysis should be treated with caution. Although it was consistent
with the theoretical account of agency in friendship to include physical attractiveness,
financial resources, physical fitness, and social connections into a single category of
expectations, future research may parcel these into distinct dimensions. Additionally,
agency expectations are consistently among the least important ideal characteristics
(Sprecher & Regan, 2002; Vigil, 2007), and were not identified as an ideal friend
18 Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
characteristic in early research (e.g., Bigelow, 1977). However, the inclusion of this con-
cept in future research is recommended because this dimension may be particularly rel-
evant in research on male–male and cross-sex friendships (Bleske & Buss, 2000).
Moderation analyses
The moderation analyses revealed few moderators of sex differences in friendship
expectations. Only three moderators explained heterogeneity in two of the five
meta-analyses: sample age for overall expectations, year of publication for overall
expectations and communion, and method of data collection for communion. The first
moderator suggests that older samples were more likely than younger samples to demon-
strate larger sex differences in overall expectations. There are several interpretations of
this moderation: expectations increase for females as they age, but do not for males,
expectations decrease for males, but remain constant for females, or expectations
increase or decrease for both sexes over time but do so at different rates. There is evi-
dence that expectations decline for both sexes, particularly as discrepancies between
expectations and friendship behaviors increase (Weiss & Lowenthal, 1975). For males,
employment and gender-role constraints particularly reduce expectations (Fischer &
Oliker, 1983). The present study did not explore the relationship between expectations
and age, only the relationship between sex differences and mean age of sample.
Therefore, a firm conclusion regarding the direction of difference cannot be drawn from
these data. Additionally, a large proportion of sampled studies were of children and ado-
lescents, and only four studies had samples with mean ages over 23 years. Furthermore,
the moderating effect of the age of sample was not detected for any of four categories of
expectations. This suggests moderation itself should be treated with caution. Future
study of friendship expectations in adult and older adult populations and longitudinal
projects demonstrating change in expectations are necessary to provide better evidence
regarding the moderating effect of age on sex differences in overall expectations.
The second moderation demonstrates that year of study moderated sex differences in
overall expectations and communion, wherein more recent studies showed greater sex
differences. Different sample cohorts may demonstrate varying sex differences in
expectations. Perhaps males and females studied in the 1970s were more similar in
overall and communion expectations than males and females sampled in 2009. Although
the data support an increase in sex differences, they do not describe the nature of that
change (i.e., females have increased expectations, males have decreased expectations,
both males and females have increased or decreased but at different rates). Although
sample cohort could explain this moderation, further longitudinal or archival research is
required to identify where that change exists. Additionally, the change in type of research
question may have influenced sex differences. Earlier research focused on child
development (e.g., Bigelow, 1971) and recent research on resource acquisition (e.g.,
Vigil, 2007). The year of publication effect was independent of moderation of mean
sample age for overall expectations. This reduces the likelihood that differences in the
age of samples explain the year of publication moderation. Finally, the significant
measurement moderation suggests that quantitative studies are more likely to report sex
differences in communion than qualitative studies. Future research is required to explain
Hall 19
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
why quantitative measures show greater differences in communion in comparison with
interviews or textual analyses.
Core and peripheral expectations
Males and females share similar expectations about the most critical components of
friendship –trust, commitment, loyalty, and genuineness. Although females expect
slightly more symmetrical reciprocity, the difference is small. It is consistent with this
sex similarity that most studies demonstrate that violations of these core expectations are
very detrimental to all friendships (Argyle & Henderson, 1984; La Gaipa, 1987). The
differences in communion suggest that females expect more of their friends for a central
element of friendship (Fehr, 2004; Wright, 2006). As a consequence, it may be easier for
males than females to meet same-sex friends’ expectations of self-disclosure and inti-
macy. This would suggest that violations of communion would be more damaging to
females’ friendships than males’ – a conclusion supported in experimental studies by
Fehr (2004) and Johnson (2005). Interpreted in combination with symmetrical recipro-
city, it appears that the most important aspects of friendship are expected more in
females’ than males’ same-sex friendships. Therefore, female friendships may require
more effort to meet expectations of symmetrical reciprocity and communion, and may
be more at risk of expectation violation (Clark & Ayers, 1993; Hall et al., 2009). By com-
parison, violations of agency would be unlikely to damage males’ friendships. Since
agency is a peripheral component to friendship (i.e., has low importance), violations
of relatively unimportant expectations may not be detrimental. Wealth, attractiveness,
and physical fitness may be qualities that are nice to have in friendships, but not neces-
sary to have. However, greater expectations of agency in males’ friendships suggest that
it may be difficult for males to meet friends’ expectations in resource acquisition, or in
intrinsic characteristics, such as physical fitness or intelligence. Males may also be more
willing than females to maintain a friendship with a high agency friend who otherwise
does not provide the more core components of friendship (e.g., loyalty, mutual regard).
Males may also be more likely than females to exclude male peers who are resource defi-
cient (e.g., poor, not athletic) from friendship networks. Future work is needed on the
consequences of males’ higher agency expectations.
Limitations and directions for future research
The first limitation is that the meta-analyses sample was relatively young. This reduces
power to detect variation in sex differences past early adulthood. Similarly, this study
had only a limited ability to explore cross-cultural variation in sex differences in expec-
tations, given that only 24%of samples were from countries other than the US and only
53%of studies reported sample race/ethnicity. However, sex differences reported here
are very similar to those reported in an excluded study from China (cf. Cheng et al.,
1995). Although there are cross-culture differences in friendship expectations (e.g.,
Argyle & Henderson, 1984; Bank, 1994; Gonzalez et al., 2004), the results of the present
investigation suggest that the magnitude of sex differences does not vary by race/ethnicity.
Nonetheless, future research conducted in resource-scarce environments could better
20 Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
account for cultural variation in friendship expectations (Kendrick et al., 2005), wherein
excessive investment in non-kin relationships holds greater risks for the survival of off-
spring due to the norms of reciprocity (Benenson & Alavi, 2004). Furthermore, in societies
where patronage and nepotism are common mechanisms for securing employment, the
value of well-connected and socially prominent friends may greatly increase agency
expectations for both sexes. As a final limitation, the present study did not explore
cross-sex friendship expectations. The different desires of males and females for sexual
access in friendship may complicate cross-sex friendships (Halatsis & Chrisakis, 2009;
Weger & Emmett, 2009), but predictable differences in cross-sex expectations may guide
future work (Bleske & Buss, 2000; Fuhrman et al., 2009).
Several types of expectations are also missing from the present study. Fun, a sense of
humor, and having a good personality are important qualities in friends, but are under-
theorized in friendship research (Fehr, 1996). As a consequence, these types of expecta-
tions were well represented in the 7%of expectations excluded from categorical meta-
analyses. Future research should verify the existence of the four dimensions of friendship
expectations and determine whether the excluded expectations merit inclusion.
The moderation of age for overall expectations and the similar value of solidarity
suggest that importance of maturation and peer influence should not be discounted in
future research. Due to the reciprocal influence of expectation and experience (Hall
et al., 2009; Wiseman, 1986), initial biological tendencies in friendship expectations may
produce greater differences in behavioral communion and agency through same-sex
socialization. Future work should document how expectations inform behavior, how
friendship selection and rejection reify same-sex social norms, and whether there are
cohort effects in sex differences in expectations.
Acknowledgements
An earlier version of this manuscript was presented at the 2009 International Association of Rela-
tionship Research conference in Lawrence, KS. My thanks to the editor and the reviewers for care-
ful and useful feedback throughout the revision of this manuscript, and to Diane Felmlee and Noel
Card for help in manuscript preparation, to Viki Zelenak and Kiley Larson for coding assistance,
and to Susan Sprecher, Dorothy Flannagan, and Patti McDougall for sharing data.
Conflict of interest statement
The author(s) declared no conflicts of interest with respect to the authorship and/or publication of
this article.
Funding
This research received no specific grant from any funding agency in the public, commercial, or
not-for-profit sectors.
References
Allen, M. (2009). Meta-analysis. Communication Monographs, 76, 398–407.
*Amalfitano, R. M. M. (1980). A psychological investigation of similarities and differences between
the friendship values and reported numbers of friends of adult women and their mothers and
fathers (Unpublished doctoral dissertation). University of Wisconsin-Madison, Madison, WI.
Hall 21
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
*Argyle, M., & Henderson, M. (1984). The rules of friendship. Journal of Social and Personal
Relationships, 1, 211–237.
Bakan, D. (1966). The duality of human existence. Boston, USA: Beacon.
Banikiotes, P. G., Neimeyer, G. J., & Lepkowsky, C. (1981). Gender and sex-role orientation
effects on friendship choice. Personality and Social Psychology Bulletin, 7, 605–610.
*Bank, B. J. (1994). Effects of national, school, and gender cultures on friendships among adoles-
cents in Australia and the United States. Youth & Society, 25, 435–456.
Benenson, J. F., & Alavi, K. (2004). Sex differences in children’s investment in same-sex peers.
Evolution and Human Behavior, 25, 258–266.
*Bigelow, B. J. (1971). The development of childhood friendship expectations. (Unpublished
master’s thesis). University of Windsor, Windsor, Ontario, Canada.
*Bigelow, B. J. (1977). The development of children’s friendship expectations: A cognitive and
behavioural perspective. (Unpublished doctoral dissertation). University of Dundee, Dundee,
Scotland, UK.
Bigelow, B. J., & La Gaipa, J. J. (1980). The development of friendship values and choice. In
H. C. Foot, A. J. Chapman, & J. R. Smith (Eds.), Friendship and social relations in children
(pp. 15–44). New York: John Wiley & Sons.
Bleske, A. L., & Buss, D. M. (2000). Can men and women be just friends? Personal Relationships,
7, 131–151.
Bukowski, W. M., Nappi, B. J., & Hoza, B. (1987). A test of Aristotle’s model of friendship for
young adults’ same-sex and opposite-sex relationships. The Journal of Social Psychology, 127,
595–603.
Burgio, M. R. (1987). Friendship patterns and friendship expectancies among the successful
aging. (Unpublished doctoral dissertation). New York University, NYC, NY.
Cann, A. (2004). Rated importance of personal qualities across four relationships. Journal of
Social Psychology, 144, 322–335.
*Chance, M. (2008). Behavior expectations in adolescence for same-sex friends, opposite-sex
friends, and romantic relationships. (Master’s thesis). Available from ProQuest Dissertations
and Theses database. (UMI No. 1454551).
Cheng, C., Bond, M. H., & Chan, S. C. (1995). The perception of ideal best friends by Chinese
adolescents. International Journal of Psychology, 30, 91–108.
*Clark, M. L. & Ayers, M. (1993). Friendship expectations and friendship evaluations:
Reciprocity and gender effects. Youth & Society, 24, 299–313.
*Clark, M. L., & Bittle, M. L. (1992). Friendship expectations and the evaluation of present
friendships in middle childhood and early adolescence. Child Study Journal, 22, 115–135.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale,
NJ: Erlbaum.
Cosmides, L., & Tooby, J. (2005). Neurocognitive adaptations designed for social exchange.
In D. M. Buss (Ed.), The evolutionary psychology handbook (pp. 584–627). New York: Wiley.
Dindia, K. (2006). Men are from North Dakota, women are from South Dakota. In K. Dindia &
D. J. Canary (Eds.), Sex differences and similarities in communication, 2nd ed (pp. 3–20).
Mahwah, NJ: Erlbaum.
Duck, S., & Wright, P. H. (1993). Reexamining gender differences in same-gender friendships:
A close look at two kinds of data. Sex Roles, 28, 709–727.
*Elkins, L. E., & Peterson, C. (1993). Gender differences in bestfriendships. Sex Roles, 29, 497–509.
22 Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Fehr, B. (1996). Friendship processes. Thousand Oaks, CA: Sage.
*Fehr, B. (2004). Intimacy expectations in same-sex friendships: A prototypical interaction-
pattern model. Journal of Personality and Social Psychology, 86, 265–284.
Felmlee, D. H. (1999). Social norms in same- and cross-gender friendships. Social Psychology
Quarterly, 62, 53–67.
Felmlee, D., & Muraco, A. (2009). Gender and friendship norms among older adults. Research on
Aging, 31, 318–344.
Fischer, C. S., & Oliker, S. J. (1983). A research note on friendship, gender, and the life cycle.
Social Forces, 62, 124–133.
Flannagan, D., Marsh, D. L., & Fuhrman, R. (2005). Judgments about the hypothetical behaviors
of friends and romantic partners. Journal of Social and Personal Relationships, 22, 797–815.
Fletcher, G. J. O., Simpson, J. A., Thomas, G., & Giles, L. (1999). Ideals in intimate relationships.
Journal of Personality and Social Psychology, 76, 72–89.
Geary, D. C., Byrd-Craven, J., Hoard, M. K., Vigil, J., & Numtee, C. (2003). Evolution and
development of boys’ social behavior. Developmental Review, 23, 444–470.
Geary, D. C., & Flinn, M. V. (2002). Sex differences in behavioral and hormonal response to
social threat: Commentary on Taylor et al. (2000). Psychological Review, 109, 745–750.
*Gonzalez, Y. S., Moreno, D. S., & Schneider, B. H. (2004). Friendship expectations of early
adolescents in Cuba and Canada. Journal of Cross-Cultural Psychology, 35, 436–445.
Halatsis, P., & Christakis, N. (2009). The challenge of sexual attraction within heterosexuals’
cross-sex friendship. Journal of Social and Personal Relationships, 26, 919–937.
*Hall, J. A., Larson, K., & Watts, A. (Nov., 2009). Satisfying friendship maintenance expectations:
The role of friendship standards, sex, and gender. Paper presented at the National Communi-
cation Association Conference in Chicago, IL.
*Hartmann, L. A. (1991). Friendship development in early adolescence: Experience, expecta-
tions, and aspects of loneliness. (Unpublished doctoral dissertation). Miami University,
Miami, OH.
Hartup, W. W., & Stevens, N. (1997). Friendship and adaptation in the life course. Psychological
Bulletin, 121, 355–370.
Hays, R. B. (1989). The day-to-day functioning of close versus casual friendships. Journal of
Social and Personal Relationships, 6, 21–37.
Horenczyk, G., & Tatar, M. (1998). Friendship expectations among immigrant adolescents and
their host peers. Journal of Adolescence, 21, 69–82.
Hussong, A. M. (2000). Distinguishing mean and structural sex differences in adolescent friend-
ship quality. Journal of Social and Personal Relationships, 17, 233–243.
Johnson, H. D. (2005). Conflict goals associated with adolescent perceptions of relationship
expectation violations during conflicts with same-sex friends. In A. V. Lee (Ed.), Psychology
of coping (pp. 41–63). Hauppauge, NY: Nova Science Publishers.
Johnson, A. J., Wittenberg, E., Haigh, M., Wigley, S., Becker, J., Brown, D., et al. (2004). The
process of relationship development and deterioration: Turning points in friendships that have
terminated. Communication Quarterly,52, 54–67.
Johnson, B. T., Scott-Sheldon, L. A. J., Snyder, L. B., Noar, S. M., & Huedo-Medina, T. B. (2008).
Contemporary approaches to meta-analysis in communication research. In A. F. Hayes, M. D.
Slater, & L. B. Snyder (Eds.), Advanced data analysis methods for communication research
(pp. 311–347). Los Angeles, CA: Sage.
Hall 23
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Jones, D. C. (1991). Friendship satisfaction and gender: An examination of sex differences in
contributions to friendship satisfaction. Journal of Social and Personal Relationships, 8,
167–185.
Kendrick, D. T., Maner, J. K., & Li, N. P. (2005). Evolutionary social psychology. In D. M. Buss
(Ed.), The evolutionary psychology handbook (pp. 803–827). New York: Wiley.
La Gaipa, J. J. (1977). Testing a multidimensional approach to friendship. In S. W. Duck (Ed.),
Theory and practice in interpersonal attraction (pp. 249–270). London: Academic Press.
La Gaipa, J. J. (1987). Friendship expectations. In R. Burnett, P. McGhee, & D. Clarke (Eds.),
Accounting for relationships: Explanation, representation and knowledge (pp. 134–157).
London: Methuen.
Laurenceau, J. P., Rivera, L. M., Schaffer, A. R., & Pietromonaco, P. R. (2004). Intimacy as an
interpersonal process: Current status and future directions. In D. J. Mashek & A. Aron (Eds.),
Handbook of closeness and intimacy (pp. 61–78). Mahwah, NJ: Erlbaum.
Lempers, J. D., & Clark-Lempers, D. S. (1993). A functional comparison of same-sex and
opposite-sex friendships during adolescence. Journal of Adolescent Research, 8, 89–108.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.
*Lusk, J., MacDonald, K., & Newman, J. R. (1998). Resource appraisals among self, friend and
leader: Implications for an evolutionary perspective on individual differences. Personality and
Individual Differences, 5, 685–700.
MacDonald, K. (1996). What do children want? A conceptualization of evolutionary influences on
children’s motivation in the peer group. International Journal of Behavioral Development, 19,
53–73.
Mancini, J. A., & Simon, J. (1984). Older adults’ expectations of support from family and friends.
Journal of Applied Gerontology, 3, 150–160.
Newcomb, A. R., & Bagwell, C. L. (1995). Children’s friendship relations: A meta-analytic
review. Psychological Bulletin, 117, 306–347.
Orwin, R. G. (1983). A fail-safe Nfor effect sizes in meta-analysis. Journal of Educational
Statistics, 8, 157–159.
Peretti, P. O., & Venton, W. C. (1986). The influence of functional components of reciprocity of
maintaining and sustaining closest friendship. Journal of Psychological Researches, 30, 83–87.
Rawlins, W. K. (2009). The compass of friendship. Thousand Oaks, CA: Sage.
Reis, H. T. (1998). Sex differences in intimacy and related behaviors: Context and process.
In D. J. Canary & K. Dindia (Eds.), Sex differences and similarities in communication
(pp. 203–231). Mahwah, NJ: Erlbaum.
Roy, R., Benenson, J. F., & Lilly, F. (2000). Beyond intimacy: Conceptualizing sex differences in
same-sex friendships. The Journal of Psychology, 134, 93–101.
Sapadin, L. A. (1988). Friendships and gender: Perspectives of professional men and women.
Journal of Social and Personal Relationships, 5, 387–403.
*Sprecher, S. & Regan, P. C. (2002). Liking some things (in some people) more than others: Part-
ner preferences in romantic relationships and friendships. Journal of Social and Personal Rela-
tionships, 19, 463–481.
Tamres, L. K., Janicki, D., & Helgeson, V. S. (2002). Sex differences in coping behavior: A meta-
analytic review and an examination of relative coping. Personality and Social Psychology
Review, 6, 2–30.
24 Journal of Social and Personal Relationships 000(00)
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
Taylor,S.E.,Klien,L.C.,Lewis,B.P.,Gruenewald,T.L.,Gurung,R.A.R.,&Updegraff,
J. A. (2000). Biobehavioral responses to stress in females: Tend-and-befriend, not fight-or-flight.
Psychological Review, 107, 411–429.
Tooby, J., & Cosmides, L. (1996). Friendship and the banker’s paradox: Other pathways to the
evolution of adaptations for altruism. Proceedings of the British Academy, 88, 119–143.
*Vigil, J. M. (2007). Asymmetries in the friendship preference and social styles of men and
women. Human Nature, 18, 143–161.
Weger, H., Jr., & Emmett, M. C. (2009). Romantic intent, relationship uncertainty, and relationship
maintenance in young adults’ cross-sex friendships. Journal of Social and Personal Relation-
ships, 26, 964–988.
Weiss, L. & Lowenthal, M. F. (1975). Life course perspectives on friendship. In M. Lowenthal,
M. Thurnher, & D. Chiriboga (Eds.), Four stages of life. San Francisco: Jossey-Bass Publishers.
Wiseman, J. P. (1986). Friendship: Bonds and binds in a voluntary relationship. Journal of Social
and Personal Relationships, 3, 191–211.
*Wood, H. D. (1972). The children’s friendship expectancy inventory and the prediction of
sociometric acceptability. (Unpublished master’s thesis). University of Windsor, Windsor,
Ontario, Canada.
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. (1988). Interpreting research on gender differences in friendship: A case for
moderation and a plea for caution. Journal of Social and Personal Relationships, 5, 367–373.
Wright, P. H. (2006). Toward an expanded orientation to the comparative study of women’s and
men’s same-sex friends. In K. Dindia & D. J. Canary (Eds.), Sex differences and similarities in
communication, 2nd ed. (pp. 37–57). Mahwah, NJ: Erlbaum.
*Zarbatany, L., Conley, R., & Pepper, S. (2004). Personality and gender differences in friendship
needs and experiences in preadolescence and young adulthood. International Journal of
Behavioral Development, 28, 299–310.
*Studies denoted with an asterisk were included in this meta-analysis.
Hall 25
at UNIV OF KANSAS on January 13, 2011spr.sagepub.comDownloaded from
