Western Journal of Communication
Vol. 69, No.4, October 2005, pp. 339–358
ISSN 1057–0314 (print)/ISSN 1745–1027 (online) © 2005 Western States Communication Association
Relational Maintenance in Cross-Sex
Friendships Characterized by Different
Types of Romantic Intent: An
Laura K. Guerrero & Alana M. Chavez
Taylor and Francis LtdRWJC130530.sgm10.1080/10570310500305471Western Journal of Communication1057-0314 (print)/1745-1027 (online)Original Article2005Western States Communication Association
694000000October 2005Professor LauraGuerreroHugh Downs School of Human CommunicationArizona State UniversityPO Box 871205TempeAZ 85287-1205USALaura.Guerrero@asu.edu
This study investigates whether perceptions of maintenance behavior in cross-sex friend-
ships vary as a function of romantic intent, biological sex, and uncertainty. Individuals
recalled the maintenance behaviors they had used over the past month in a cross-sex friend-
ship characterized by one of the following situations: mutual romance (i.e., both partners
want the friendship to turn romantic), strictly platonic (i.e., neither partner wants the
friendship to turn romantic), desires romance (i.e., the individual wants romance but
fears the friend does not), or rejects romance (i.e., the individual does not want romance
but thinks the friend does). Individuals in the mutual romance situation generally reported
the most maintenance behavior. Those in the rejects romance and strictly platonic situa-
tions reported less routine contact and activity, less flirtation, and more talk about outside
romance. Individuals in the rejects romance and mutual romance situations reported the
most relationship talk. Sex differences were also found. Finally, individuals who were
uncertain about the status of their cross-sex friendship were less likely to report using some
types of relational maintenance. These findings underscore the dynamic nature of rela-
tional maintenance in cross-sex friendships.
Keywords: Relational Maintenance; Maintenance Behavior; Cross-Sex Friendships;
Laura Guerrero (PhD, University of Arizona) is a professor in the Hugh Downs School of Human Communica-
tion at Arizona State University. Alana Chavez (MPA, Arizona State University) is currently working as the Project
Management Assistant for Bioindustry and Downtown Projects for the City of Phoenix. A portion of the data
presented in this paper was collected as part of the second author’s honors thesis, which was directed by the first
author. We are grateful for Sandra Petronio’s feedback as a member of the honors thesis committee. A version of
this paper based on the portion of the data collected for the honors thesis was presented at the 1999 Western States
Communication Association conference in Vancouver, British Columbia. Correspondence to: Dr Laura Guerrero,
Hugh Downs School of Human Communication, Arizona State University, Tempe, AZ 85287-1205, USA. Email:
340 L. K. Guerrero & A. M. Chavez
Friendships are a significant source of happiness in people’s lives, yet friends often
take one another for granted (Fehr, 1996, 2000). Wiseman (1986) described friend-
ship as fragile because there is little societal pressure for friends to maintain or repair
their relationships. Similarly, Canary, Stafford, Hause, and Wallace (1993) found that
although people reported using many of the same maintenance behaviors with
friends, romantic partners, and family members, less maintenance was reported in
friendships. Yet prosocial behaviors such as shared activity, self-disclosure, and
supportiveness appear to be key ingredients associated with emotional closeness and
relational satisfaction in friendships (Afifi, Guerrero, & Egland, 1994; Bippus &
Rollin, 2003; Fehr, 2000).
Cross-sex friendships, which are becoming more prevalent and significant within
people’s social networks (Fehr, 1996; Monsour, 2002), may be particularly fragile and
difficult to maintain (Werking, 1997). Rose (1985) found that people generally report
engaging in less maintenance behavior in cross-sex friendships than same-sex friend-
ships. Cross-sex friendship can be complicated, with ambiguity about the potential
romantic or sexual nature of the relationship creating uncertainty. As Dainton,
Zelley, and Langan (2003) observed, ‘For heterosexual individuals, maintaining a
cross-sex friendship involves the affection, companionship, intimacy, and assistance
found in same-sex relationships, but it also involves downgrading sexuality’ (p. 91).
Sometimes, however, one or both of the friends may want the friendship to turn
romantic. Sexuality may then begin to be emphasized. By contrast, individuals in
cross-sex friendships may downgrade sexuality more than usual when one of the
friends is in the position of rejecting romance. Indeed, promoting or discouraging
movement toward romance is likely to be an important and sometimes difficult part
of relational maintenance in cross-sex friendships, especially when cross-sex friends
have different romantic intentions and experience uncertainty about the state of their
The present study explores how romantic intention (or the lack thereof), biological
sex, and uncertainty associate with maintenance behavior in cross-sex friendships.
Four types of cross-sex friendships are examined—mutual romance (the participant
believes that both s/he and the friend want the friendship to turn romantic), strictly
platonic (the participant believes that both s/he and the friend want the relationship to
stay platonic), desires romance (the participant wants the relationship to turn romantic
but perceives that the friend does not), and rejects romance (the participant does not
want the relationship to turn romantic but perceives that the friend does). Because men
and women often perceive cross-sex friendships differently (Dainton et al., 2003; Rose,
1985), we also examine how reported use of maintenance behavior varies as a function
of sex and the interaction between sex and friendship type. Finally, due to ambiguity
about romantic intent, some cross-sex friends are likely to experience high levels of
relational uncertainty (Afifi & Burgoon, 1998), which could inhibit their use of proso-
cial maintenance behavior (Dainton, 2003). Thus, we investigate whether people in
cross-sex friendships differ in relational uncertainty as a function of romantic intent,
as well as whether there are associations between uncertainty and maintenance behav-
ior in cross-friendships.
Western Journal of Communication 341
Relational Maintenance as a Dynamic Process
A central assumption guiding this paper is that maintenance is a dynamic process that
involves adapting to the changing needs and goals that characterize a relationship.
Dindia (1994, 2003) proposed that maintenance behaviors are multiphasic; in other
words, the same behavior can be used to accomplish varied goals across different
phases of a relationship. For example, positivity can be used as an impression manage-
ment strategy in the beginning stages of a relationship, as a routine maintenance behav-
ior in established relationships, and as a repair strategy after a conflict or relationship
problem has occurred. Similarly, we contend that cross-sex friends can use mainte-
nance behaviors in an attempt to maintain the status quo, move the friendship into a
new romantic phase, or discourage the friend from pursuing a romantic relationship.
By engaging in behaviors that help accomplish these goals, cross-sex friends enhance
their chances of sustaining a comfortable and satisfying relationship.
Relational maintenance has been defined in terms of different goals or relational
outcomes. Dindia and Canary (1993; Dindia, 2003) identified four common concep-
tualizations of maintenance: keeping a relationship in existence, keeping a relationship
in a specified state or stable condition, keeping a relationship satisfying, and keeping a
relationship in repair. Dindia (2003) argued that the second definition—keeping the
relationship in a stable condition—seems at odds with dialectic perspectives on
maintenance (e.g., Baxter & Simon, 1993) that focus on adapting to changing needs (or
dialectical tensions) within a relationship. However, Dindia noted that the term rela-
tional maintenance ‘need not imply that a relationship is static and unchanging’ (p. 3).
On the contrary, keeping a relationship stable and satisfying often requires coping
effectively with change. We take the position that relational maintenance encompasses
all four of the definitions reviewed by Dindia and Canary, with partners adjusting to
different situations by either maintaining the status quo or changing the relationship
based on their goals and desires. If partners are ineffective in adapting to changing
goals, the relationship is more likely to become dissatisfying or to end. Thus, our
perspective privileges the position that maintenance behaviors maximize the satisfac-
tion level within relationships (to the extent possible given the circumstances), with
behaviors that stabilize or change a relationship defined as maintenance as long as they
contribute to sustaining the relationship.
Maintenance in Cross-Sex Friendship
Research on cross-sex friendship suggests that people engage in different behaviors
depending on their goals for the relationship. For example, Messman, Canary, and
Hause (2000) demonstrated that cross-sex friends in platonic relationships avoid flirt-
ing with one another as a way of maintaining the status quo and preventing the rela-
tionship from turning romantic. Afifi and Burgoon (1998) found that cross-sex friends
avoid talking about sensitive issues (such as the state of their relationship) if they worry
that such talk could threaten the existence of their friendship. Of course, enacting
appropriate and effective maintenance behavior may be most complicated in situations
342 L. K. Guerrero & A. M. Chavez
where opposite-sex friends are at cross purposes, with one friend feeling dissatisfied
because he or she wants romance, and the other partner trying to maintain the status
quo or reject the friend’s advances without ruining the friendship.
O’Meara (1989) discussed four challenges that cross-sex friends sometimes face,
three of which are particularly relevant to the romantic tension that exists in some
cross-sex friendships—the emotional bond challenge (i.e., confusing closeness and
liking with romantic attachment), the public presentation challenge (i.e., having to
explain the nature of the friendship to others), and the sexual challenge (i.e., negotiat-
ing the sexual boundaries of the relationship). Although most studies suggest that
sexual activity is relatively rare between cross-sex friends (e.g., Fuiman, Yarab, & Sensi-
baugh, 1997; Messman et al., 2000), Afifi and Faulkner (2000) found that about 50%
of the college population they sampled could recall having had sex with a friend at least
once. They also found that having sex with a friend was associated with aversive uncer-
Tensions associated with these challenges likely vary depending upon perceptions of
romantic intent in cross-sex friendships. As Dainton et al. (2003) claimed, ‘Just as
variations exist in the ways that casual daters, serious daters, and married couples
maintain their relationships,’ so too should there be variations based on the type of
friendship people share (p. 97). Research on friendship already suggests that people
maintain various types of friendships differently. For example, Rose and Serafica
(1986) found that proximity (e.g., routine contact) was the most frequently mentioned
maintenance behavior for casual friends. In contrast, moderate levels of affection and
high levels of interaction (e.g., communication, activity) were reportedly needed to
maintain relationships between close and best friends. Other studies have demon-
strated that people generally report using more maintenance in same-sex friendships
than cross-sex friendships (e.g., Afifi et al., 1994; Rose, 1985). However, these studies
also suggest that people are more likely to report using affection (Rose, 1985) and flir-
tation (Afifi et al., 1994) in cross-sex friendships, possibly because of the potential
romantic or sexual involvement that characterizes some of these friendships.
Cross-sex friends should differ in their maintenance behavior depending on which
of the four situations (e.g., strictly platonic, desires romance) they believe characterizes
their relationship. Relevant to the mutual romance situation, research suggests that
romantic partners increase their use of positive maintenance strategies as their relation-
ships become more serious (Dindia, 1994; Guerrero, Eloy, & Wabnik, 1993). Similarly,
cross-sex friends might use especially high levels of relational maintenance as they
prepare to develop the friendship into a romance. When individuals are confident that
their partners also want to escalate the friendship to romance, they may feel especially
free to use high levels of maintenance behavior. When an individual perceives that her
or his romantic intentions do not match the friend’s intentions, more complex patterns
of maintenance behavior may occur. For example, those in the desires romance situa-
tion might avoid talking about their romantic feelings because they are afraid of rejec-
tion. Messman et al. (2000) found that some people identified ‘fear of rejection’ and
‘safeguarding the relationship’ as common goals in platonic cross-sex friendships. Fear
of rejection involves being worried about the emotional risks of disclosing romantic
Western Journal of Communication 343
feelings to a friend, whereas safeguarding the relationship involves being cautious about
the possible consequences of changing the nature of the friendship (e.g., from platonic
Those in the rejects romance situation, on the other hand, might avoid engaging in
too much relational maintenance so their friends will not misinterpret their behavior
as romantic interest. In support of this claim, an early study by Ayres (1983) showed
that people reported using avoidance in response to a hypothetical situation in which
their partners wanted the relationship to become more intimate but they wanted the
relationship to stay the same. Thus, people in the rejects romance situation may be
less likely to use positive maintenance behaviors than those in the other friendship
Based on this reasoning, we expect that cross-sex friends will report differences in
maintenance behavior based on whether they want the relationship to develop into a
romance, or whether they want it to remain platonic. However, because this is the first
study to examine this issue in the context of cross-sex friendships, we are uncertain as
to what the specific differences will be. Therefore, we ask:
RQ1: Will individuals report using different levels of relational maintenance
depending on whether they perceive themselves to be in the strictly platonic,
mutual romance, desires romance, or rejects romance situation?
If cross-sex friends have not arrived at a mutual understanding about issues related to
romantic intent, uncertainty is likely to exist. Indeed, Afifi and Burgoon (1998) found
that cross-sex friends were more likely to report relational uncertainty and to avoid
discussing the state of their relationship than were dating partners. According to Afifi
and Reichert (1996), people who experience relational uncertainty are unsure about
their partner’s intentions for the relationship, the nature of the relationship (e.g.,
romantic vs. platonic), and the future of the relationship (e.g., how long the relation-
ship will last). Similarly, Knobloch and Solomon (1999) defined relational uncertainty
as the extent to which people lack confidence about the accuracy of their perceptions
of the partner’s involvement. They also discussed a specific form of relational uncer-
tainty—mutuality uncertainty—that occurs when partners are unsure whether or not
their feelings are reciprocated. Given the possibility of ambiguity in cross-sex friend-
ships, we ask the following research question:
RQ2: Will relational uncertainty differ depending on whether individuals perceive
themselves to be in the strictly platonic, mutual romance, desires romance, or
rejects romance situation?
The level of ambiguity or uncertainty present is also likely to be associated with
self-reported use of maintenance behavior. Indeed, scholars have argued that people
can use specific types of maintenance behavior to help manage relational uncertainty
(Dainton, 2003) or repair relationships after an uncertainty-increasing event
(Emmers & Canary, 1996). In a study that included both long-distance and proximal
344 L. K. Guerrero & A. M. Chavez
dating relationships, Dainton and Aylor (2001) found that relational uncertainty
associated negatively with the maintenance strategies of openness, assurances, positiv-
ity, task sharing, and social networking. In another study focusing on long-distance
relationships, uncertainty was negatively correlated with openness, assurances, and
positivity (Ficara & Mongeau, 2000). A third study (Dainton, 2003) examined associ-
ations between maintenance behaviors and four types of relational uncertainty identi-
fied by Knobloch and Solomon (1999): behavioral norms uncertainty (i.e.,
uncertainty about what is acceptable vs. unacceptable behavior), mutuality uncer-
tainty (i.e., uncertainty about whether or not one’s feelings are reciprocated), defini-
tional uncertainty (i.e., uncertainty about the state of the relationship), and future
uncertainty (i.e., uncertainty about the directions the relationship will take in the
future). Dainton (2003) demonstrated that all four forms of relational uncertainty
were negatively associated with the maintenance behaviors of advice, assurances, inte-
grative conflict management, social networking, openness, positivity, and task sharing
within the context of romantic relationships. Because cross-sex friendships are some-
times ambiguous in terms of romantic interest, the individuals in these relationships
may need to manage uncertainty as much or more than romantic partners. Thus, it is
important to ask:
RQ3: Are there associations between relational uncertainty and the reported use of
maintenance behaviors within the context of cross-sex friendship?
Undergraduate students were recruited from lower- and upper-division communi-
cation courses at a large southwestern university. Of the 456 students who
completed questionnaires, 16 who identified themselves as gay/lesbian were excluded
from the study because the issue of romantic interest would be different for these
participants than for those who identified themselves as heterosexual (n = 436) or
bisexual (n = 4).
The average age of these 440 respondents (female n = 226, male n
= 214) was 20
years old (range = 18–46 years old; md = 20 years old). Respon-
dents had known their cross-sex friends for an average of 38 months (range = 3–144
The investigators visited various undergraduate communication classes to recruit
volunteers, who were informed that the questionnaire examined a wide range of behav-
iors that people use in friendships between men and women. Questionnaires were then
distributed to those who wished to participate. Respondents completed the question-
naires during or after class time and dropped them into a box as they exited the class-
room. Students were told not to place their names on the questionnaires to ensure
Western Journal of Communication 345
Respondents first read the following instructions:
This questionnaire involves recalling and reporting the activities you use in a rela-
tionship with a cross-sex friend. Cross-sex friends are defined as friends of the oppo-
site-sex whom you spend time with but do not currently date. Please think of a friend
fitting this description and place her or his initials here: ________.
Respondents were asked to keep this friend in mind as they answered all questions.
Relational maintenance behaviors
The first part of the questionnaire instructed respondents to report on the types of
behaviors they had used in their cross-sex friendship over the past month using seven-
point Likert-type scales (1 = disagree strongly, 7 = agree strongly). The most popular
typology for measuring relational maintenance in romantic relationships was devel-
oped by Stafford and Canary (1991), who delineated five behaviors: positivity (i.e.,
being pleasant and complimentary toward the partner), openness (i.e., sharing private
information through self-disclosure), assurances (i.e., verbalizing commitment to one
another and planning together for the future), social networking (i.e., integrating one
another’s circle of friends and family), and task sharing (i.e., helping each other with
tasks in a fair and equitable fashion). Stafford and Canary’s (1991) scales were
included in the present study, with items modified to reflect the context of cross-sex
friendships. Based on past research, we also included items representing additional
maintenance behaviors that may be applicable to cross-sex friendships, including
supportiveness, activity sharing, humor, flirting, avoidance, and antisocial behaviors
such as acting jealous or trying to change the partner (see Stafford, 2003 for a review).
A total of 55 relational maintenance items were included in the questionnaire.
We subjected these items to a principal components analysis with oblique rotation,
which is an appropriate method for identifying an underlying factor structure while
allowing for moderate correlations among factors (Park, Dailey, & Lemus, 2002). Items
were retained if they had primary loadings of at least .60 as well as secondary loadings
at least .20 less than their primary loading. After removing 21 items that did not meet
these criteria during a first and second run, a 10-factor solution accounting for 68.16%
of the variance emerged. The Kaiser–Meyer–Oklin measure of sampling adequacy was
.91. The results of the principal components analysis, including items, primary load-
ings, variance accounted for by each factor, means, standard deviations, and inter-item
reliability are reported in Table 1. Correlations between the scales measuring various
maintenance behaviors are reported in Table 2.
After completing the relational maintenance section, respondents checked the box
that best described their cross-sex friendship. The options were as follows: (a) neither
of us wants to escalate our friendship to a romantic relationship; (b) both of us want to
escalate our friendship to a romantic relationship; (c) I would like to escalate our
346 L. K. Guerrero & A. M. Chavez
Ta b le 1 Principal Components Analysis Results
Factor/items Variance Reliability
Routine contact and activity (M = 4.66, SD = 1.38) 12.59% .82
I call my friend on a regular basis .72
I go places with my friend on a regular basis .80
We visit one another’s homes or apartments on a regular basis .78
I initiate phone calls to my friend .80
Emotional support and positivity (M = 5.85, SD = .83) 12.55% .81
I act cheerful and positive when with my friend .83
I try hard to listen to my friend’s problems .78
I try to be supportive and caring .63
I comfort my friend in times of trouble .69
I present myself as cheerful and optimistic when with my friend .63
Relationship talk (M = 4.17, SD = 1.40) 11.51% .82
I tell my friend how I feel about our friendship .78
I attempt to talk to my friend about the quality of our
I tell my friend what I want from our friendship .81
We have periodic talks about our friendship .81
Instrumental support (M = 5.52, SD = 1.00) 10.39% .75
I give my friend advice .83
I let my friend know I am available to help with tasks or chores .74
I help my friend solve problems .76
I help my friend accomplish tasks and get things done .64
Social networking (M = 5.16, SD = 1.26) 9.51% .79
We spend time with mutual friends .83
We focus on common friends and affiliations .78
I show that I’m willing to do things with her/his circle of friends .63
I include our common friends in our activities .82
Antisocial behavior (M = 4.29, SD = 1.48) 9.26% .82
We argue about differences in opinion .90
I communicate my frustrations about our friendship .77
I often complain to my friend .83
Humor and gossip (M = 5.68, SD = .98) 8.71% .71
I tease my friend good-naturedly .70
We share ‘inside jokes’ .69
I joke around a lot with my friend .78
We frequently ‘gossip’ together .70
Talk about outside romance (M = 5.00, SD = 1.54) 7.45% .67
I tell my friend about my past and/or current romances .88
I tell my friend about my romantic encounters .77
Western Journal of Communication 347
friendship to a romantic relationship, but my friend probably does not; and (d) my
friend would like to escalate our friendship to a romantic relationship, but I would
not. These statements were used to operationalize the strictly platonic (n = 215, female
n = 109, male n = 106), mutual romance (n = 82, female n = 43, male n = 39), desires
romance (n = 53, female n = 24, male n = 29), and rejects romance (n = 90, female
n = 50, male n = 40) situations, which served as an independent variable.
To validate these categorical measures, two seven-point Likert-type items
(1 = disagree strongly, 7 = agree strongly) that read ‘I would like this friendship to
develop into a romantic relationship’ and ‘I have romantic feelings for my friend’
= .97) were included with some general measures of relational closeness
later in the questionnaire. A contrast comparing the mutual romance (M = 6.24, SD
Ta b le 1 Continued
Factor/items Variance Reliability
Flirtation (M = 4.64, SD = 1.59) 5.50% .74
I am flirtatious with my friend .85
I avoid flirting with my friend (recoded) .89
Avoidance of negativity (M = 4.21, SD = 1.55) 4.04% .71
I avoid conflict with my friend .86
I avoid criticizing my friend .86
Notes: Reliabilities were calculated using Cronbach’s alpha statistic. The sum of the variance accounted for by each
factor equals more than 68.16% (the amount of total variance accounted for by the solution) because oblique
rotation takes the correlations between factors into account.
Ta b le 2 Correlations among Maintenance Behaviors
RCA ESP RT IS SN AB HG TOR F AN
ESP .40** 1.0
RT .39** .37** 1.0
IS .40** .56** .39** 1.0
SN .46** .36** .22** .33** 1.0
AB .35** .20** .22** .23** .16* 1.0
HG .36** .46** .24** .44** .29** .31** 1.0
TOR .19** .23** .24** .27** .26** .38** .35** 1.0
F .27** .14* .20** .23** .09 .09 .18*** .07 1.0
AN −.05 .01 −.04 .01 −.08 −.35** −.04 −.16** −.07 1.0
Notes: RCA = routine contact and activity; ESP = emotional support and positivity; RT = relationship talk; IS =
instrumental support; SN = social networking; AB = antisocial behavior; HG = humor and gossip; TOR = talk
about outside romance; F = ﬂirtation; AN = avoidance of negativity.
**p < .001; *p < .01, two-tailed.
348 L. K. Guerrero & A. M. Chavez
= 1.73) and desires romance situations (M = 5.80, SD = 1.20) to the strictly platonic
(M = 2.24, SD = 1.64) and rejects romance situations (M = 2.77, SD = 1.63) showed
that individuals who had marked one of the former two descriptions reported a higher
desire to develop the friendship into a romantic relationship than did those in the latter
two groups, t(438) = 21.67, p < .001,
Relational uncertainty and sex
Next, respondents were asked how certain they were about their friend’s feelings
toward them. Relational uncertainty was assessed with two questions (inter-item
= .93). The first read as follows: ‘Sometimes people are uncertain about how their
cross-sex friends really feel about them. How certain are you that the box you checked
above really describes the way that you and your friend feel about each other?’ There
were six possible responses: very certain (1), certain (2), somewhat certain (3), some-
what uncertain (4), uncertain (5), and very uncertain (6). The second item read:
‘Sometimes people are confused about the nature of their cross-sex friendships. How
confident are you that the box you checked above is an accurate reflection of the
current state of your friendship?’ Responses were: very confident (1), confident (2),
somewhat confident (3), somewhat unconfident (4), unconfident (5), and very uncon-
fident (6). At the end of the survey, participants completed a few questions assessing
the quality of their cross-sex friendships as well as demographic questions including
sex, age, dating/marital status (for themselves and their friend), and sexual orientation.
Research Questions 1 and 3
Ten separate 4 (friendship type) by 2 (sex) analyses of covariance (with uncertainty as
the covariate) were utilized to test RQ1 and RQ3. Biological sex was included because
past research suggests that men and women differ somewhat in their reported use of
maintenance behavior (see Dainton et al., 2003 for a review). Three main effects
producing small effect sizes emerged for sex. Women (M = 6.01, SD = .75) reported
using more emotional support and positivity than men (M = 5.68, SD = .91), F(1, 431)
= 13.88, p < .001,
= .04. Women (M = 5.59, SD = .98) also reported giving more
instrumental support than men (M = 5.45, SD = 1.02), F(1, 431) = 8.05, p < .05,
= .02. Finally, women (M = 5.34, SD = 1.46) reported talking more about outside
romance than men (M = 4.72, SD = 1.59), F(1, 431) = 9.76, p < .01,
Two of the ANCOVAs produced significant friendship type by sex interactions: social
networking, F(3, 431) = 6.73, p < .001,
= .11; and antisocial behavior, F(3, 431) =
3.87, p < .01,
= .05. As graphed in Figure 1, men did not report significantly different
levels of social networking based on friendship type, F(3, 205) = .117, p > .05. Women,
by contrast, did report different levels of social networking based on friendship type,
F(3, 227) = 9.22, p < .001,
= .09, with a Tukey-B range test demonstrating that
women in the rejects romance situation perceived themselves to use less social network-
ing (M = 4.55, SD = 1.42) than women in the strictly platonic (M = 5.55, SD = 1.14),
Western Journal of Communication 349
Strictly PlatonicMutual RomanceRejects RomanceDesires Romance
Figure 1 Interaction on Social Networking.
Strictly PlatonicMutual RomanceRejects RomanceDesires Romance
Figure 2 Interaction on Antisocial Behavior.
350 L. K. Guerrero & A. M. Chavez
mutual romance (M = 5.34, SD = 1.23), and desires romance (M = 5.28, SD = 1.72) situ-
ations. A significant main effect for friendship type also emerged for social networking,
but since this effect only held for women, it was subsumed by the interaction effect.
Figure 1 Interaction on Social Networking.
As graphed in Figure 2, men and women were similar in how much antisocial behav-
ior they reported in the rejects romance, mutual romance, and strictly platonic situa-
tions. However, within the desires romance condition, women (M = 5.13, SD = 1.31)
reported significantly more antisocial behavior than men (M = 3.67, SD = 1.20), t(51)
= 3.32, p < .01,
= .18. Significant main effects for friendship type and sex also
emerged for antisocial behavior, however, the disordinal nature of the interaction
rendered these effects uninterpretable.
Figure 2 Interaction on Antisocial Behavior.
Effects of friendship type
Main effects for friendship type (RQ1) emerged for routine contact and activity, F(3,
432) = 7.92, p < .001,
= .06; emotional support and positivity, F(3, 431) = 7.90, p <
= .06; relationship talk, F(3, 431) = 10.38, p < .001,
= .08; instrumental
support, F(3, 431) = 4.56, p < .01,
= .04; talk about outside romance, F(3, 431) =
10.46, p < .05,
= .08; and flirtation, F(3, 431) = 34.98, p < .001,
= .20. As the means
in Table 3 show, respondents in the mutual romance situation generally reported the
most relational maintenance behavior, with the exception of ‘talk about outside
romance,’ which was reported more by those in the strictly platonic and reject romance
situations. Individuals in the desires romance and rejects romance conditions also
tended to differ from one another. Specifically, those in the desires romance condition
reported using more routine contact and activity, less relationship talk, more talk about
outside romance, and more flirtation than those in the rejects romance condition.
Associations between maintenance behavior and relational uncertainty
The ANCOVAs revealed significant negative associations between six of the mainte-
nance behaviors and uncertainty (RQ3), albeit most effect sizes were very small.
Ta b le 3 Means and (Standard Deviations) Associated with Signiﬁcant Main Effects for
Routine contact and activity 4.91 (1.50)
Emotional support and positivity 5.94 (.83)
Relationship talk 3.93 (1.29)
Instrumental support 5.63 (.97)
Talk about outside romance 4.45 (1.71)
Flirtation 5.50 (1.27)
Note: Tukey-B range tests showed that means that do not share a common subscript when viewed across rows
were signiﬁcantly different from one another.
Western Journal of Communication 351
Individuals who reported high levels of uncertainty also tended to report low levels of
relational talk, F(1, 431) = 27.67, p < .001, r = −.30; talk about outside romance, F(1,
431) = 3.78, p < .05, r = −.12; routine contact and activity, F(1, 431) = 6.09, p < .05, r
= −.12; social networking, F(1, 431) = 6.60, p < .05, r = −.11; instrumental support, F(1,
431) = 4.56, p < .05, r = −.11; and humor/gossip, F(1, 431) = 4.51, p < .05, r = −.11.
Research Question 2: Uncertainty across Friendship Types
RQ3 examined whether individuals in the four friendship situations would differ in
their reports of relational uncertainty.
A one-way ANOVA showed that relational
uncertainty did indeed vary as a function of friendship type, F(3, 436) = 11.75, p < .001,
= .07. A Tukey B range test (p < .05) showed that individuals in the desires romance
situation reported more relational uncertainty (M = 3.38, SD = 1.33) than those in the
strictly platonic (M = 2.43, SD = 1.36), rejects romance (M = 2.24, SD = 1.17), or
mutual romance (M = 2.16, SD = 1.06) situations. None of the groups, however,
reported very high levels of relational uncertainty.
The purpose of this investigation was to determine how relational maintenance func-
tions in cross-sex friendships that are perceived to vary in terms of romantic intent.
Overall, the results suggest that perceptions related to both romantic intent and the
mutuality (or non-mutuality) of that intent make a difference. Individuals who
considered themselves to be in the mutual romance, strictly platonic, desires romance,
and rejects romance conditions reported differences in the maintenance behaviors
they used and the extent to which they experienced relational uncertainty. Relational
uncertainty, in turn, was associated with lower levels of relational talk, routine contact
and activity, social networking, talk about outside romance, instrumental support, and
humor/gossip. The present study also uncovered some interesting categories of rela-
tional maintenance behavior that may be especially relevant to cross-sex friendships.
Maintenance Behaviors in Cross-Sex Friendship
Our factor analysis suggests that there may be some unique categories of maintenance
behavior in cross-sex friendships. Some of the maintenance behaviors we examined—
social networking, antisocial behavior, and flirtation—were wholly consistent with
those found in past research (see Stafford, 2003, for a review). However, new or refined
categories also emerged. In contrast to other studies utilizing a single category of open-
ness (e.g., Stafford & Canary, 1991), two categories related to specific types of openness
surfaced in the present study: relationship talk and talk about outside romance.
Within the context of cross-sex friendships, relationship talk (which focuses on
discussing feelings about the friendship) and talk about outside romance (which
focuses on talking about relationships other than the friendship) appear to represent
potentially sensitive forms of talk that vary based on romantic intent and uncertainty
352 L. K. Guerrero & A. M. Chavez
(see also, Afifi & Burgoon, 1998). On the other end of the spectrum, avoidance of nega-
tivity focuses on refraining from engaging in communication that could cause conflict
or hurt feelings, and thus represents a lack of openness.
In addition to openness, scholars have discussed spending time together and staying
connected as critical to the maintenance of friendships (Dainton et al., 2003; Fehr,
1996; Rawlins, 1994). Within the present study, items related to going places together,
visiting one another’s homes, and calling one another worked together to produce a
factor we termed ‘routine contact and activity.’ Scholars have also pointed to support-
iveness as a key maintenance behavior in friendship (e.g., Afifi et al., 1994; Burleson &
Samter, 1994; Dainton et al., 2003; Fehr, 1996; Messman et al., 2000) and positivity as
a key maintenance behavior in romantic relationships (e.g., Canary & Stafford, 1994;
Dainton, Stafford, & Canary, 1994; Stafford & Canary, 1991). In the present study,
these two types of behaviors merged into a broad category we termed ‘emotional
support and positivity.’ Behaviors falling under this category promote a positive
emotional atmosphere that allows friends to enjoy one another’s company and alleviate
one another’s stress.
Friends can support each other instrumentally as well as emotionally. In some past
work on friendship maintenance, giving advice has been considered part of general
supportiveness (e.g., Messman et al., 2000; Stafford, 2003). In work on romantic
relationships, task sharing has been discussed as a distinct category of maintenance
behavior (e.g., Stafford & Canary, 1991). In the present study, however, advice and task
sharing fell under a common factor, which we termed instrumental support. Thus,
scholars studying maintenance in friendships may want to distinguish between
emotional support (i.e., displays of comfort, empathetic listening, positivity, and
caring) and instrumental support (e.g., advice giving, problem solving, and task shar-
ing). Such a distinction is consistent with research and theory on social support, which
has separated emotion- and problem-focused behavior (e.g., Barbee, Rowatt, &
Finally, although other scholars have discussed humor as a maintenance behavior in
friendships and other relationships (e.g., Afifi et al., 1994; Canary et al., 1993; Stafford,
2003), humor and gossip have not combined as a factor in other studies. Perhaps teas-
ing each other good-naturedly, joking around, and gossiping together all function as
forms of light, entertaining conversation that maintain friendships. Of course, some
humor and gossip might reveal information about a person’s true attitudes and
feelings, which could explain the small but significant negative association between
uncertainty and humor/gossip.
Perceptions of Romantic Intent
Consistent with our perspective that relational maintenance is a dynamic process that
reflects individuals’ goals and desires, people reported different levels of maintenance
behavior as a function of the friendship situation. Individuals who believed that their
friendship was characterized by a mutual desire to become romantic reported the most
consistently high pattern of relational maintenance among the four groups. The means
Western Journal of Communication 353
suggest that emotional support and positivity, routine contact and activity, flirtation,
and instrumental support were especially characteristic of individuals in the mutual
romance situation. This result complements Guerrero et al.’s (1993) finding that
romantic couples who reported especially high levels of maintenance behavior during
an initial data collection were likely to rate their relationships as more serious and
committed eight weeks later. Individuals in both romantic relationships and cross-sex
friendships may increase their use of positive maintenance behaviors when they are
preparing to escalate the intimacy level of their relationship.
There was, however, one exception. Individuals in the mutual romance and desires
romance situations reported less talk about outside romance than did those in either
the strictly platonic or reject romance situations. One potential explanation for this
finding, which was supported by a chi-square analysis,
is that individuals who wish to
move their friendship toward romance may be less likely to have current outside
romantic relationships to talk about. Still, the moderate effect size connected to the chi-
square analysis, coupled with the fact that the scale measuring ‘talk about outside
romance’ included items about past as well as current romances, suggests that there
may be other complementary explanations. Those in the strictly platonic and rejects
romance situations might sometimes talk about their romantic relationships as a way
of signaling that they are already involved with someone else. Such talk could provide
an indirect yet fairly effective way of signaling disinterest in romance. For those in the
mutual romance and desires romance situations, talking about current or past roman-
tic relationships may be an especially sensitive topic since jealousy or competitiveness
could stem from such discussion.
In addition to reporting more talk about outside romance, individuals in the rejects
romance and strictly platonic situations reported using less routine contact and activ-
ity, as well as less flirtation. Perhaps individuals in these situations avoid spending
considerable time with their cross-sex friends as a way of discouraging romantic inter-
est. Work on unrequited love suggests that people who are in the position of rejecting
romantic advances may be especially careful not to encourage the would-be lover’s
romantic feelings by spending too much time with him or her. As Bratslavsky,
Baumeister, and Sommer (1998) put it, if polite attempts at keeping the relationship
platonic are ineffective, individuals in the rejecter position are likely to withdraw from
interaction so that they can say they ‘did nothing to encourage their admirers’ (p. 317).
Similarly, individuals in the rejects romance and strictly platonic situations may avoid
flirting with one another as a way of keeping their relationships from turning romantic
(Messman et al., 2000). Flirting is a primary mechanism for signaling romantic attrac-
tion and sexual availability (Givens, 1983), so it is natural that individuals who want to
turn their friendships romantic would be most likely to report these behaviors. More-
over, flirtatious behaviors are usually indirect and ambiguous (e.g., demure gaze and
smiling), so flirting may be a relatively safe way to show one’s romantic feelings in the
desires romance situation.
Furthermore, in the present study, individuals in both the desires romance and
strictly platonic situations reported the least relationship talk. For individuals who
desire romance, expressing thoughts about the relationship is likely seen as risky
354 L. K. Guerrero & A. M. Chavez
communication that could lead to rejection or ruin the friendship (Afifi & Burgoon,
1998). However, if a perception of mutual romance exists, an individual is likely to feel
confident about discussing the relationship. In fact, such discussion may help friends
turn their relationship romantic. For individuals in the strictly platonic situation, in
contrast, there may not be a pressing need to discuss the state of the relationship since
both friends are probably happy to maintain the status quo. For individuals in the
rejects romance situation, on the other hand, discussing the state of the friendship may
be necessary, especially if other indirect attempts to signal disinterest have failed.
Interestingly, women in the rejects romance situation reported less social network-
ing than women in the other three situations. This finding could relate to O’Meara’s
(1989) public presentation challenge. If a woman is especially sensitive to the percep-
tions of others, she might be aware that people could view her friend as a potential
boyfriend if their circle of friends becomes too integrated. If she is already in a romantic
relationship with another person, she might also avoid social networking so as not to
invoke jealousy in her romantic partner. Finally, women might be sensitive to the
possibility that allowing a friend to spend considerable time with her social network
could be misinterpreted as a signal that she is romantically interested.
Finally, women who desired romance (but perceived that their partner did not)
reported higher levels of antisocial behavior (e.g., arguing about differences, commu-
nicating frustrations, and complaining) than did men who desired romance. Perhaps
women are more likely than men to show negative affect and to communicate about
potential problems when tension occurs in cross-sex friendships. Women may also be
more likely than men to become frustrated if their cross-sex friend does not reciprocate
their romantic interest. Indeed, Motley and Reeder (1995) suggested that women are
less accustomed than men to having their romantic advances rejected.
Individuals in the desires romance situation also appear to experience the most uncer-
tainty about the status of their friendships. These individuals are in a relational quan-
dary. They are romantically attracted to their cross-sex friends, but they doubt that the
attraction is mutual. Thus it is natural that they should feel some level of uncertainty.
Importantly, however, none of the four friendship situations were characterized by
high levels of uncertainty (see also Afifi & Burgoon, 1998), which could have
suppressed some of the associations between uncertainty and maintenance behavior.
Nonetheless, individuals experiencing uncertainty reported less relationship talk with
their cross-sex friends. Of course, the causal direction of this association is unclear. As
Afifi and Burgoon (1998) suggested, people might avoid discussing the state of the rela-
tionship precisely because they are uncertain of the outcome of such talk and fear rejec-
tion or other negative relational ramifications. Conversely, cross-sex friends could
experience uncertainty precisely because they have spent relatively little time talking
about the status of their friendship.
In addition to relationship talk, individuals were less likely to report talk about
outside romance, social networking, routine contact and activity, instrumental
Western Journal of Communication 355
support, and humor/gossip if they were experiencing uncertainty about the status of
their cross-sex friendship (albeit effect sizes were very small). Dainton (2003) proposed
that uncertainty and inequity might be linked, thereby providing a theoretical explana-
tion for negative associations between uncertainty and maintenance behavior. Specifi-
cally, based on her finding that uncertainty was a stronger predictor of maintenance
behavior than (in)equity, Dainton argued that perceived inequity might lead to uncer-
tainty, which in turn might affect ‘the performance of maintenance’ (p. 181). Although
we did not measure inequity in the present study, it is plausible to expect that percep-
tions of inequity would be greatest in the desires romance situation, wherein individu-
als may feel that they are underbenefited because they want more from and perhaps
invest extra effort into the relationship compared to their friend. Thus, investigating
how inequity, uncertainty, and friendship type work together to predict maintenance
behavior appears to be an important area for future research.
This study represents a first step toward determining how maintenance behaviors
function in the context of cross-sex friendships that vary in terms of romantic intent.
The study is limited to one person’s perceptions of maintenance behaviors as well as
one person’s view of the friendship situation (e.g., desires romance vs. strictly
platonic). Thus, in the future researchers would benefit from examining actual main-
tenance behaviors and by including the perspectives of both cross-sex friends. Indeed,
in some situations friends might have very different perceptions of the friendship. For
example, one friend might characterize the friendship in terms of mutual romance,
whereas the other friend might characterize her or himself as in the rejects romance
position. Nonetheless, the present study suggests that reports of maintenance behavior
vary based on an individual’s own perception of the friendship. Biological sex and rela-
tional uncertainty also play minor roles in predicting how much maintenance behavior
people report in their cross-sex friendships. These findings provide a starting point for
understanding how maintenance behaviors reflect the changing landscape of cross-sex
friendships that involve different goals and desires regarding romantic intent.
 Data from participants who classified themselves as gay/lesbian will be analyzed in a separate
study comparing maintenance behaviors reported by individuals with different sexual orien-
 Respondents also reported on their dating/marital status. Ten individuals were married, 126
were in serious dating relationships, 95 were dating someone casually, and 209 were not
dating anyone in particular. When reporting on their cross-sex friend’s dating/marital status,
respondents reported that 11 were married, 75 were dating someone seriously, 103 were
dating someone casually, and 250 were not dating anyone in particular.
 Although we did not make any predictions regarding the effect of biological sex on uncer-
tainty, we checked to see whether there was a main effect for biological sex or an interaction
effect (biological sex by friendship type) on uncertainty. These effects were nonsignificant.
356 L. K. Guerrero & A. M. Chavez
 Interestingly, the general items related to openness (e.g., I encourage my friend to disclose
thoughts and feelings to me) produced split loadings in the factor analysis, making them too
general to consider as a distinct category.
 We conducted a chi-square analysis to determine whether people were less likely to be in a
serious romantic relationship if they characterized their cross-sex friendship as falling
under the mutual romance or desires romance condition. The resultant chi-square showed
that people differed in their likelihood of being in a serious romantic relationship based
on friendship type,
(3) = 75.18, C = .38. Those in the mutual romance situation were
less likely to be in romantic relationships with other people than were respondents in the
other situations. For specific information regarding this analysis, please contact the first
Afifi, W. A., & Burgoon, J. K. (1998). ‘We never talk about that’: A comparison of cross-sex friend-
ships and dating relationships on uncertainty and topic avoidance. Personal Relationships, 5,
Afifi, W. A., & Faulkner, S. L. (2000). On being ‘just friends’: The frequency and impact of
sexual activity in cross-sex friendships. Journal of Social and Personal Relationships, 17,
Afifi, W. A., Guerrero, L. K., & Egland, K. L. (1994, June). Maintenance behaviors in same- and oppo-
site-sex friendships: Connections to gender, relational closeness, and equity issues. Paper
presented at the annual meeting of the International Network on Personal Relationships,
Iowa City, IA.
Afifi, W. A., & Reichert, T. (1996). Understanding the role of uncertainty in jealousy experience and
expression. Communication Reports, 9, 93–103.
Ayres, J. (1983). Strategies to maintain relationships: Their identification and perceived usage.
Communication Quarterly, 31, 62–67.
Barbee, A. P., Rowatt, T. L., & Cunningham, M. R. (1998). When a friend is in need: Feelings about
seeking, giving, and receiving social support. In P. A. Andersen & L. K. Guerrero (Eds.),
Handbook of communication and emotion: Research, theory, applications, and contexts
(pp. 281–301). San Diego, CA: Academic Press.
Baxter, L. A., & Simon, E. P. (1993). Relationship maintenance strategies and dialectical contradic-
tions in personal relationships. Journal of Social and Personal Relationships, 10, 225–242.
Bippus, A. M., & Rollin, E. (2003). Attachment style differences in relational maintenance and
conflict behaviors: Friends’ perceptions. Communication Reports, 16, 113–123.
Bratslavsky, E., Baumeister, R. F., & Sommer, K. L. (1998). To love or be loved in vain: The trials and
tribulations of unrequited love. In B. H. Spitzberg & W. R. Cupach (Eds.), The dark side of
close relationships (pp. 307–326). Mahwah, NJ: Lawrence Erlbaum Associates.
Burleson, B. R., & Samter, W. (1994). A social skills approach to relationship maintenance: How
individual differences in communication skills affect the achievement of relationship func-
tions. In D. J. Canary & L. Stafford (Eds.), Communication and relational maintenance
(pp. 61–90). San Diego, CA: Academic Press.
Canary, D. J., & Stafford, L. (1994). Maintaining relationships through strategic and routine inter-
action. In D. J. Canary & L. Stafford (Eds.), Communication and relational maintenance
(pp. 3–22). San Diego, CA: Academic Press.
Canary, D. J., Stafford, L., Hause, K. S., & Wallace, L. A. (1993). An inductive analysis of rela-
tional maintenance strategies: Comparisons among lovers, relatives, friends, and others.
Communication Research Reports, 10, 5–14.
Dainton, M. (2003). Equity and uncertainty in relational maintenance. Western Journal of
Communication, 67, 164–186.
Western Journal of Communication 357
Dainton, M., & Aylor, B. (2001). A relational uncertainty analysis of jealousy, trust, and mainte-
nance in long-distance versus geographically close relationships. Communication Quarterly,
Dainton, M., Stafford, L., & Canary, D. J. (1994). Maintenance strategies and physical affection as
predictors of love, liking, and satisfaction in marriage. Communication Reports, 7, 88–98.
Dainton, M., Zelley, E., & Langan, E. (2003). Maintaining friendships throughout the lifespan. In
D. J. Canary & M. Dainton (Eds.), Maintaining relationships through communication
(pp. 79–102). Mahwah, NJ: Lawrence Erlbaum Associates.
Dindia, K. (1994). A multiphasic view of relationship maintenance strategies. In D. J. Canary & L.
Stafford (Eds.), Communication and relational maintenance (pp. 91–112). San Diego, CA:
Dindia, K. (2003). Definitions and perspectives on relational maintenance communication. In
D. J. Canary & M. Dainton (Eds.), Maintaining relationships through communication (pp.
1–28). Mahwah, NJ: Lawrence Erlbaum Associates.
Dindia, K., & Canary, D. J. (1993). Definitions and theoretical perspectives on relational mainte-
nance. Journal of Social and Personal Relationships, 10, 163–173.
Emmers, T. M., & Canary, D. J. (1996). The effect of uncertainty reducing strategies on young
couples’ relational repair and intimacy. Communication Quarterly, 44, 166–182.
Fehr, B. (1996). Friendships processes. Thousand Oaks, CA: Sage.
Fehr, B. (2000). The life cycle of friendship. In C. Hendrick & S. S. Hendrick (Eds.), Close relation-
ships: A sourcebook (pp. 71–82). Thousand Oaks, CA: Sage.
Ficara, L. C., & Mongeau, P. A. (2000, November). Relational uncertainty in long-distance college
student dating relationships. Paper presented at the annual meeting of the National Communi-
cation Association, Seattle, WA.
Fuiman, M., Yarab, P., & Sensibaugh, C. (1997, July). Just friends? An examination of sexual, physical,
and romantic aspects of cross-gender friendships. Paper presented at the annual meeting of the
International Network on Personal Relationships, Oxford, OH.
Givens, D. B. (1983). Love signals. New York: Crown Books.
Guerrero, L. K., Eloy, S. V., & Wabnik, A. I. (1993). Linking maintenance strategies to relationship
development and disengagement: A reconceptualization. Journal of Social and Personal
Relationships, 10, 273–283.
Knobloch, L. K., & Solomon, D. H. (1999). Measuring the sources and content of relational uncer-
tainty. Communication Studies, 50, 261–278.
Messman, S. J., Canary, D. J., & Hause, K. S. (2000). Motives to remain platonic, equity, and the use
of maintenance strategies in opposite-sex friendships. Journal of Social and Personal Relation-
ships, 17, 67–94.
Monsour, M. (2002). Women and men as friends: Relationships across the lifespan in the 21st century.
Mahwah, NJ: Lawrence Erlbaum Associates.
Motley, M. T., & Reeder, H. M. (1995). Unwanted escalation of sexual intimacy: Male and female
perceptions of connotations and relational consequences of resistance messages. Communica-
tion Monographs, 62, 355–382.
O’Meara, D. (1989). Cross-sex friendship: Four basic challenges of an ignored relationship. Sex
Roles, 21, 525–543.
Park, H. S., Dailey, R., & Lemus, D. (2002). The use of exploratory factor analysis and principal compo-
nents analysis in communication research. Human Communication Research, 28, 562–577.
Rawlins, W. K. (1994). Being there and growing apart: Sustaining friendships during adulthood. In
D. J. Canary & L. Stafford (Eds.), Communication and relational maintenance (pp. 275–294).
San Diego, CA: Academic Press.
Rose, S. M. (1985). Same- and cross-sex friendships and the psychology of homosociality. Sex Roles,
Rose, S. M., & Serafica, F. C. (1986). Keeping and ending casual, close, and best friendships. Journal
of Social and Personal Relationships, 3, 275–288.
358 L. K. Guerrero & A. M. Chavez
Stafford, L. (2003). Maintaining romantic relationships: A summary and analysis of one research
program. In D. J. Canary & M. Dainton (Eds.), Maintaining relationships through communica-
tion (pp. 51–77). Mahwah, NJ: Lawrence Erlbaum Associates.
Stafford, L., & Canary, D. J. (1991). Maintenance strategies and romantic relationship type, gender
and relational characteristics. Journal of Social and Personal Relationships, 8, 217–242.
Werking, K. (1997). We’re just good friends: Women and men in nonromantic relationships. New York:
Wiseman, J. P. (1986). Friendship: Bonds and binds in a voluntary relationship. Journal of Social and
Personal Relationships, 3, 191–211.