Misperception of sexual and romantic interests in opposite‐sex friendships: Four hypotheses

Abstract

Two online studies evaluated the misperception of sexual and romantic interests in established relationships and tested four hypotheses: a simple sex-difference hypothesis, a projection hypothesis, a mate value hypothesis, and a mediation hypothesis. Two hundred thirty-eight (Study 1) and 198 (Study 2) members of young adult opposite-sex friendship dyads indicated their sexual and romantic interests in their friend and their perceptions of their friend’s sexual and romantic interests in them. Participants projected their own levels of sexual and romantic interests onto their opposite-sex friend, mediating the following effects: males overperceived and females underperceived their friends’ sexual (but not romantic) interest, and participants of both sexes misperceived the sexual (but not romantic) interest of friends depending on the friends’ mate value.

Full text

UNCORRECTED PROOF
Misperception of sexual and romantic interests in
opposite-sex friendships: Four hypotheses
BRYAN L. KOENIG,aLEE A. KIRKPATRICK,aAND TIMOTHY KETELAARb
aCollege of William and Mary and bNew Mexico State University
Abstract
Two online studies evaluated the misperception of sexual and romantic interests in established relationships and
tested four hypotheses: a simple sex-difference hypothesis, a projection hypothesis, a mate value hypothesis, and a
mediation hypothesis. Two hundred thirty-eight (Study 1) and 198 (Study 2) members of young adult opposite-sex
friendship dyads indicated their sexual and romantic interests in their friend and their perceptions of their friend’s
sexual and romantic interests in them. Participants projected their own levels of sexual and romantic interests onto
their opposite-sex friend, mediating the following effects: males overperceived and females underperceived their
friends’ sexual (but not romantic) interest and participants of both sexes misperceived the sexual (but not romantic)
interest of friends depending on the friends’ mate value.
Previous research has documented links
between the misperceptions of sexual and
romantic interests, sexual assault, and sexual
harassment. In one study, misperception of
sexual interest was the strongest predictor of
number of sexual assaults, including rape
(Abbey, McAuslan, & Ross, 1998), and other
studies have shown that misperception of sex-
ual interest may contribute to sexual harass-
ment (Johnson, Stockdale, & Saal, 1991;
Sigal, Gibbs, Adams, & Derfler, 1988). Mis-
understandings regarding sexual and romantic
interests are not limited to dating relationships
but can also occur in the context of opposite-
sex friendships (O’Meara, 1989) and could
potentially result in dissolution of such friend-
ships. Approximately 15% of sexual assaults
occur within opposite-sex friendships (Abbey
et al., 1998). Therefore, greater understanding
of misperception of sexual and romantic inter-
ests is a valuable topic for empirical research
and theoretical understanding.
This article presents two studies that extend
previous research in three ways. First, in these
studies, we evaluated the misperception of
sexual interest in long-standing relationship
dyads, that is, opposite-sex friendships. Sec-
ond, we examined misperception of romantic
interest. Finally, we evaluated the mediation of
sex differences in misperception of sexual and
romantic interests by perceiver level of sexual
and romantic interests, respectively. This arti-
cle is organized as follows. After a brief liter-
ature review, we present two conclusions:
(a) to demonstrate the misperception of sexual
interest, researchers should compare perceived
level of sexual interest with actual level of
sexual interest and (b) previous researchers
have focused on men’s, and largely ignored
women’s, misperception of sexual interest.
We then provide a rationale for differentiating
sexual and romantic interests. Next, we iden-
tify opposite-sex friendships as useful subjects
B
PERE0163
Journal Name Manuscript No.
Dispatch: 22.6.07 Journal: PERE CE: J. Rajalakshmi
Author Received: No. of pages: 19 ME: Senthil
Bryan L. Koenig, Department of Psychology, College of
William and Mary; Lee A. Kirkpatrick, Department of
Psychology, College of William and Mary; Timothy
Ketelaar, Department of Psychology, New Mexico State
University.
Bryan L. Koenig is now at the Department of Psychol-
ogy, New Mexico State University.
The authors thank two anonymous reviewers of an
earlier draft of this manuscript for their helpful comments.
The material presented here is a part of the first author’s
master’s thesis under the supervision of the second author.
Correspondence should be addressed to Bryan L. Koe-
nig, New Mexico State University, Department of Psychol-
ogy, Las Cruces, NM 88003, e-mail: blkoen@nmsu.edu.
Personal Relationships,14 (2007), 411–429. Printed in the United States of America.
Copyright Ó2007 IARR. 1350-4126=07
411
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55

UNCORRECTED PROOF
of study for misperception of sexual and
romantic interests. We then review potential
causes for misperception and articulate four
hypotheses: the simple sex-difference hypoth-
esis, the projection hypothesis, the mate value
hypothesis, and the mediation hypothesis. We
then present two studies designed to test these
hypotheses.
Studies typically find that, compared to
women, men perceive both men and women
to have more sexual interest (e.g., Abbey,
1982; Abbey & Harnish, 1995; Fisher &
Walters, 2003; Goodchilds & Zellman, 1984;
Shea, 1993; Shotland & Craig, 1988). Re-
searchers usually interpret these findings as
resulting from men’s, rather than from wom-
en’s, misperceptions (e.g., Abbey, 1982;
Haselton & Buss, 2000; Johnson et al., 1991;
Saal, Johnson, & Weber, 1989; but see Abbey,
1987; Donat & Bondurant, 2003; Shotland &
Craig for exceptions). Logically, however, the
conclusion that misperception has occurred is
not justifiable when participants rate the sex-
ual interest of hypothetical targets. Because
hypothetical targets do not exist, they do not
have levels of sexual interest that can be accu-
rately or inaccurately perceived. That is, hypo-
thetical targets do not have an actual level of
sexual interest that can be used as a standard of
comparison for determining overperception or
underperception. Instead, such results merely
demonstrate that men reliably perceive more
sexual interest in third party targets than do
women, or conversely, that women perceive
less sexual interest in third parties than do
men. Thus, studies using hypothetical targets
are limited to the identification of sex differ-
ences in perception. Measurement of under-
perception and overperception of sexual
interest per se, in contrast, requires actual lev-
els of sexual interest to use as a comparison,
and thus actual targets.
To identify the misperception of sexual
interest, researchers can have each member
of actual male-female dyads rate their own
sexual interest in the other person and the sex-
ual interest that they think the other person has
toward them. With the assumptions that par-
ticipants (a) accurately perceive their own
sexual interest, (b) honestly report their per-
ception of their own sexual interest, and
(c) honestly report their perception of the sex-
ual interest of the other person in them,1it is
logical to conclude that when a man perceives
a woman to have more sexual interest in him
than she self-reports, he is overperceiving her
sexual interest in him. Likewise, if a woman
perceives a man to have less sexual interest in
her than he self-reports, she is underperceiving
his sexual interest in her.
The first published study to evaluate mis-
perception of sexual interest employed a design
with actual targets and found evidence that
both men overperceive and women underper-
ceive the sexual interest of opposite-sex tar-
gets (Abbey, 1982). In that study, a man and
woman (the actors), who did not know each
other, talked for 5 min and then rated their
own and the other actor’s sexual interest in
each other. Hidden observers also rated the
actors’ sexual interest in each other. Men’s
ratings of the female actors’ sexual interest
were higher than those self-reported by the
female actors, supporting an interpretation of
men overperceiving women’s sexual interest.
Female actors and observers rated the male
actor’s sexual interest lower than male actors
self-reported, which suggests that women
underperceived the sexual interest of the male
actors. Abbey (1982) did not explicitly recog-
nize female underperception, instead conclud-
ing that men ‘‘read sexual intent into friendly
behavior...becauseofageneralmalebias....
Evidently, females are not subject to this bias’’
(p. 838). This conclusion implies that men, but
not women, misperceive sexual interest.
Researchers have replicated Abbey’s
(1982) seminal study at least five times. Each
replication has provided evidence consistent
1. These assumptions, implicit in almost all research on
the misperception of sexual interest, are merely stated
explicitly here (for alternative assumptions, see, e.g.,
Fisher & Walters, 2003 or Haselton & Buss, 2000).
Nonetheless, there may be somedegree of self-deception
in the perception of one’s own sexual or romantic inter-
est (Paulhus, 1984; Trivers, 2000). Also, impression
management may influence participant ratings of their
own sexual and romantic interests, or those of another
person (Fisher & Walters; Haselton & Buss; but see
Abbey, 1982; Abbey, et al., 2000; Shea, 1993 for con-
tradictory evidence). Empirical evaluation of these
assumptions is critical to establish the validity of the
corpus of sexual misperception research to date.
412 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
with men overperceiving and women under-
perceiving the sexual interest of opposite-sex
targets (Abbey, Zawacki, & McAuslan, 2000;
Edmondson & Conger, 1995; Harnish, Abbey, &
DeBono, 1990; Saal et al., 1989; Shea, 1993).
Yet, in only two of these six studies using
actual targets did the researchers explicitly
mention women underperceiving the men’s
sexual interest (Abbey et al., 2000; Shea).
Thus, while researchers of the misperception
of sexual interest are beginning to recognize
the possibility that women misperceive men’s
sexual interest, the standard interpretation
appears to be that men, not women, misper-
ceive levels of sexual interest. In order to eval-
uate the misperception of sexual interest by
men and women, instead of just sex differen-
ces in perception, the current studies utilized
actual targets of both sexes.
Extending misperception research to
romantic interest
Mating relationships can be of varying tempo-
ral durations, with short-term mating (e.g.,
brief affairs or one night stands) anchoring
the short-lived end of this continuum and
long-term mating (e.g., marriage) anchoring
the other end (Buss & Schmitt, 1993). Studies
on the perception of sexual interest have con-
founded short-term and long-term mating
strategies in various ways. Participants rated
target sexuality using the adjectives ‘‘roman-
tic’’ and ‘‘sexy’’ (Abbey & Harnish, 1995);
participants used these four topics to indicate
degree of sexual interest: sexual attraction,
sexual-advance receptivity, interest in having
sex, and interest in dating (Abbey et al., 2000);
a mate value measure included both short-term
and long-term mate value (Haselton, 2003); and
a manipulation of mate-search motivation
included a first date movie that may have been
more sexual for men and romantic for women
(Maner et al., 2005). Sexual strategies theory
emphasizes the importance of differentiating
between short-term and long-term mating
strategies (Buss & Schmitt). Research on mate
value has empirically supported the impor-
tance of the distinction between short-term
and long-term mating (Kenrick, Sadalla,
Groth, & Trost, 1990). Short-term and long-
term mating appear to be distinct phenome-
non; therefore, the current studies explicitly
differentiate sexual and romantic interests.
To the authors’ knowledge, almost no
research has explored the misperception of
romantic interest (see Haselton & Buss, 2000,
for misperception of commitment intent).
Therefore, we designed the current research
to shed light on the misperception of romantic
interest. Two factors limited the prediction of
the effects of perceiver sex (i.e., maleness or
femaleness). First, stereotypically, people
expect women to be more interested in love
than are men (Basow, 1992), yet one study
found that men fall in love faster than do
women (Rubin, Peplau, & Hill, 1981). Second,
research based on Lee’s (1973/1976) six styles
of love has found that the sexes differ in their
approaches to love, with neither men nor
women dominant across all types of love.
Erotic love, which includes early attraction,
physical attraction, emotional intensity, and
strong commitment to one’s lover, is arguable
the style of love most closely related to roman-
tic interest. Studies have found levels of erotic
love to be either the same for men and women
(Hendrick & Hendrick, 1986; Hendrick,
Hendrick, Foote, & Slapion-Foote, 1984) or
higher for women (Sprecher & Toro-Morn,
2002). Together, these mixed findings do not
allow a clear prediction of which sex will have
greater romantic interest. We applied all
hypotheses developed for sexual interest to
romantic interest, but romantic interest predic-
tions were nondirectional with regard to sex
differences.
Opposite-sex friendships as subject of study
in misperception research
Researchers may be able to arouse high levels
of sexual interest in the lab, especially in men,
but inducing passionate love in the lab is both
difficult and unethical. One solution to this
challenge is to utilize naturally occurring sex-
ual and romantic interests. A brief review of
the literature suggests that sexual interest is
common in opposite-sex friendships. Sexual
attraction is often an important component in
the formation of opposite-sex friendships
(Rose, 1985); a majority of participants in
Misperception of mating interests 4131
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
one study reported physical or sexual attrac-
tion toward an opposite-sex friend (Reeder,
2000), and two of five students in one college
sample reported having had sexual intercourse
with at least one person who was their oppo-
site-sex friend at the time (Afifi & Faulkner,
2000). Research has also found romantic inter-
est between opposite-sex friends. Opposite-
sex friendships often start with the hope that
the friendship will develop into a romantic
relationship (Kaplan & Keys, 1997); 52% of
participants indicated romantic attraction
toward an opposite-sex friend (Reeder), and
long-term mating potential was on the top 10
list of the most frequent aspects of opposite-
sex friendships for both men and women (see
Bleske & Buss, 2000, Table 3, p. 141). The
prevalence of sexual and romantic interests
in opposite-sex friendships provides ample
opportunity for their misperception. Indeed,
Abbey (1987) found that half of all self-
reported experiences of one’s sexual interest
being misperceived occurred within an
opposite-sex friendship. The high prevalence
of sexual and romantic interests in opposite-
sex friendships suggests that opposite-sex
friendships might be useful alternatives to
actual or hypothetical strangers for the evalu-
ation of the misperception of sexual and
romantic interests.
Most studies on the misperception of sexual
interest have been conducted in the lab and
have used strangers or hypothetical targets
(see Abbey, 1987 and Haselton, 2003 for
exceptions). We know of no reason to expect
patterns of misperception to be different
between opposite-sex friends as compared to
those between strangers. Replication of pat-
terns of misperception of sexual interest
between individuals in established relation-
ships would provide evidence for the external
validity of the cumulative laboratory findings
about the misperception of sexual interest.
Causes of misperception
Although previous research has explored the
role of contextual factors on the misperception
of sexual interest, the literature does not con-
verge on a single common causal pathway for
the misperception of sexual interest. Some
authors suggest that socialization, primarily
through the media and stereotypes, produces
men with traditional gender attitudes and val-
ues as well as cognitive schemas that depict
a highly sexualized view of the world, which
in turn result in men interpreting ambiguous
information as sexual interest, especially in
combination with men’s lower accuracy than
women at decoding nonverbal communication
(e.g., Abbey et al., 2000; Kowalski, 1993). In
line with this socialization view, researchers
have suggested that men may have a lower
threshold than women for perceiving sexual
interest (e.g., Kowalski; Saal et al., 1989).
Evolutionary theorists have suggested that
men overperceive women’s sexual interest so
as not to miss mating opportunities, the pri-
mary constraint on men’s reproductive success
(Haselton & Buss, 2000). Evolutionary theory,
in conjunction with self-fulfilling prophecy,
the tendency for false beliefs to cause others
to behave so as to make the belief come true
(Merton, 1968), may be able to explain wom-
en’s underperception of men’s sexual interest.
Women who perceive sexual interest in a man
may inadvertently behave in ways that result in
sexual interest and sexual advances from that
man, which could be costly in terms of sexual
harassment from that man or jealousy from her
mate. A final proposed mechanism for the mis-
perception of sexual interest is that men pro-
ject their higher level of sexual interest onto
women. That is, men assume women have the
same level of sexual interest as they them-
selves have (Shotland & Craig, 1988).
The simple sex-difference hypothesis
Researchers have emphasized the perceiver’s
sex as a critical variable in the misperception
of sexual interest for diverse theoretical rea-
sons, including socialization (e.g., Abbey,
1982) and evolved sex differences in cognitive
biases (Haselton & Buss, 2000). In her seminal
study, Abbey (1982) proposed the basic logic
for the socialization approach:
Certainly the stereotypes of our culture, as
evidenced by the mass media’s depiction of
men and women, portray men as having
greater interest in sexual matters than do
414 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
women. Once men develop this sexual ori-
entation, it may act as a generalized expec-
tancy, causing them to interpret ambiguous
information ...asevidence in support of
their beliefs (p. 837).
This socialization hypothesis predicts that
men overperceive women’s sexual interest.
Presumably, socialization explanations can
accommodate the complementary findings
that women underperceive men’s sexual inter-
est by arguing that women are supposed to
have less interest in sexual matters and, there-
fore, interpret ambiguous information as lack
of sexual interest (Abbey & Melby, 1986).
These theoretical considerations suggest the
simple sex-difference hypothesis: men will
overperceive women’s sexual interest and
women will underperceive men’s sexual inter-
est regardless of their degree of sexual interest
in the target.
The projection hypothesis
Shotland and Craig (1988) first suggested that
the sex difference in misperception occurs
because people use their own level of sexual
interest to estimate the sexual interest of
others. Men assume women have the same
level of sexual interest as men, men have more
sexual interest than do women (for a review,
see Baumeister, Catanese, & Vohs, 2001), thus
men overperceive women’s sexual interest.
Maner et al. (2005) recently proposed a more
general model of projection, functional projec-
tion, in which people’s motivational states
lead them to overperceive others to be in emo-
tional states that it is functional to overper-
ceive. For example, a mate-search motive
increased men’s perception of sexual interest
in attractive women and fear increased percep-
tion of anger in potential antagonists. The
logic of functional projection is based on error
management theory (EMT; Haselton and
Buss, 2000), which argues that, over evolu-
tionary time, overperceptions or underpercep-
tions with less costs or more benefits were
likely to be selected and become species typ-
ical. According to the projection hypothesis,
high levels of sexual interest should predict
overperception of sexual interest, whereas
low levels of sexual interest should predict
underperception.2
The mate value hypothesis
Haselton (2003) found that people with
higher attractiveness as a mate (i.e., mate
value) were more likely to report that their
friendly behavior had been misinterpreted by
a member of the opposite sex as sexual inter-
est. According to EMT, underperceiving the
sexual interest of a high mate value target is
more costly than overperceiving because
missing a chance to mate with a high mate
value individual (underperception) is more
costly than wasted mating effort (overper-
ception). Thus, the mate value hypothesis
predicts that the sexual interest of targets
with high short-term mate value will be over-
perceived, whereas the sexual interest of tar-
gets with low short-term mate value will be
underperceived. Furthermore, it follows that
the short-term and long-term mate values of
the target individual may be an important
variable beyond perceiver sex for evaluating
the mediation hypothesis, to which we now
turn.
The mediation hypothesis
A corollary of the projection hypothesis is
that the perceiver’s levels of sexual interest
in the target may mediate the effects of per-
ceiver sex and target mate value (Shotland &
Craig, 1988). That is, for perceiver sex, over-
perception is the result of the higher level of
sexual interest of perceivers who are men. For
mate value, overperception is the result of
higher levels of sexual interest toward targets
2. One reviewer noted that it is unclear why self-fulfilling
prophecies would not also explain such results. Self-
fulfilling prophecies might explain how the perception
that a friend has a high level of sexual interest would
lead that friend to increase their level of sexual interest,
and therefore be a proximal mechanism resulting in
fitness benefits in addition to those suggested by
EMT. Nonetheless, we are not here concerned with
such effects. That is, we are looking at how a person’s
level of sexual interest affects the perception of
a friend’s sexual interest, not how a person’s level of
sexual interest affects the friend’s actual sexual interest.
Misperception of mating interests 4151
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
with higher short-term mate value. In sum,
the mediation hypothesis predicts that con-
trolling perceiver level of sexual interest will
reduce the ability of perceiver sex and target
mate value to predict misperception of sexual
interest.
The current studies tested four hypotheses.
First, the simple sex-difference hypothesis
predicts that men will overperceive women’s
sexual interest and women will underperceive
men’s sexual interest, irrespective of level of
sexual interest of the perceiver. Second, the
projection hypothesis predicts that high levels
of perceiver sexual interest will result in over-
perception of target sexual interest, whereas
low levels of perceiver sexual interest will
result in underperception of target sexual inter-
est. Third, the mate value hypothesis predicts
that participants will overperceive the sexual
interest of targets with high short-term mate
value and participants will underperceive the
sexual interest of targets with low short-term
mate value. Finally, the mediation hypothesis
predicts that controlling for perceiver level of
sexual interest will reduce or eliminate the
effects of perceiver sex and target mate value
on misperception. We also applied each
hypothesis to romantic interest, excluding
directional predictions based on perceiver
sex. First, the simple sex-difference hypothesis
predicts sex differences in misperception of
romantic interest, irrespective of the per-
ceivers’ level of romantic interest. Second,
the projection hypothesis predicts that high
levels of perceiver romantic interest will result
in overperception of target romantic interest
and low levels of perceiver romantic interest
will result in underperception of target roman-
tic interest. Third, the mate value hypothesis
predicts that participants will overperceive the
romantic interest of targets with high long-
term mate value and participants will under-
perceive the romantic interest of targets with
low long-term mate value. Finally, the media-
tion hypothesis predicts that perceiver level of
romantic interest will mediated any effects of
perceiver sex or target long-term mate value.
Study 1 tested the simple sex-difference
hypothesis, the projection hypothesis, and the
mediation hypothesis. Study 2 replicated
Study 1 and tested the mate value hypothesis.
Study 1
The goal of the current project was to evaluate
predictors of misperception of sexual and
romantic interests. Using online question-
naires, members of opposite-sex friendship
dyads (a) indicated their sexual and romantic
interests in the other member of the dyad and
(b) estimated that friend’s sexual and romantic
interests in them. This design allowed us to
evaluate concurrently the unique predictive
abilities of perceiver sex and perceivers’ own
interests.
Method
Participants. Participants were introduc-
tory psychology students at the College of
William and Mary, a middle class, academi-
cally competitive, southeastern public univer-
sity in the United States with a majority
of Caucasian students, and their closest oppo-
site-sex friend, whom each student recruited.
Demographic data collected from students’
opposite-sex friends did not include their
occupation; thus, we do not know what pro-
portion of them were also students at College
of William and Mary. The use of a convenience
and volunteer sample in the current study was
necessary because students had to have an
opposite-sex friend who would participate in
the study and no sampling frame for such par-
ticipants exists. Students participated in partial
fulfillment of a course requirement for intro-
ductory psychology. The study’s homepage
provided the following information to partic-
ipants: ‘‘Welcome to the Relationships & Sex-
uality study. We will be asking you to
complete a number of online questionnaires
regarding your beliefs, attitudes, and history
in relationships.’’ We excluded three dyads
from analysis because at least one member
was a gay male or lesbian. We excluded four
additional dyads because the student’s oppo-
site-sex friend and their romantic partner had
the same first name and therefore may have
been the same person (the first names of
romantic partners were collected as part of
another study). In five cases, two students each
had the other as their opposite-sex friend,
resulting in duplicative data. We excluded
416 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
one of each set of data at random. The final
sample for Study 1 consisted of 68 female and
51 male students, each with a friend (238 par-
ticipants total). The mean age of students was
18.79 years (SD ¼1.25) and the mean friend-
ship length was 31.16 months (SD ¼33.92).
Note that opposite-sex friend refers to the
friends recruited by the students to participate
in this study and from the perspective of either
member of the opposite-sex friendship dyad,
the term refers to the other member. When
used in the first meaning, it includes students’
as a clarifier, for example, the students’
opposite-sex friend. When used in the second
meaning, it will be expressed without a clari-
fier, for example, the participants’ romantic
interest in their opposite-sex friend. Partici-
pant refers to both the students and the oppo-
site-sex friends, which they recruited, since all
were participants in the study.
Questionnaires. Participants completed
questionnaires on the Internet. Students pro-
vided their own name and the name of their
friend. The Web pages embedded these names
in the questionnaires. After providing relation-
ship history details, participants answered
questions about their sexual and romantic
interests in their opposite-sex friend and their
perceptions of that friend’s sexual and roman-
tic interests in them, resulting in four scales
(presented to participants in this order): per-
ception of friend’s romantic interest in self,
perception of friend’s sexual interest in self,
romantic interest in friend, and sexual interest
in friend. Participants rated their friend’s inter-
est before their own, reducing the likelihood
that participants would be primed by the ques-
tionnaire to use their own level of interest as
a reference point when estimating their
friend’s level of interest. Each scale included
three questions designed to measure a concep-
tually different component of sexual and
romantic interests (presented in this order):
behavioral, affective, and cognitive. An exam-
ple of the behavioral question measuring sex-
ual interest, using ‘‘John’’ as the friend’s
name, follows: ‘‘If you and John were both
single, how likely is it that you would have
casual sex with John if John asked?’’ The cor-
responding affective measure of sexual inter-
est would be, ‘‘How much do you desire to
have casual sex with John?’’ The cognitive
measure of sexual interest would be, ‘‘How
frequently do you think about having casual
sex with John?’’ On the topic romantic inter-
est, questions asked about joining a long-term,
committed romantic relationship instead of
having casual sex. To evaluate perceptions of
friend’s interests, we reversed the name of the
friend and the term ‘‘you.’’ For example, the
behavioral question probing perception of
friend’s romantic interest would be, ‘‘If you
and John were both single, how likely is it that
John would join a long-term, committed
romantic relationship with you if you asked?’’
A 7-point response scale followed all ques-
tions. Anchors were appropriate to the ques-
tion: For behavioral questions, anchors were
very unlikely and very likely; for affective
questions, anchors were none and very much;
and for cognitive questions, anchors were
never and very often. Cronbach alphas ranged
from .92 to .94, demonstrating satisfactory
reliability. Two principal components analy-
ses with varimax rotations, one for self-
reported interests and another for perception
of friend’s interests, each revealed two orthog-
onal (i.e., independent) components with
eigenvalues above one. In both analyses, sex-
ual and romantic interests were independent
constructs.
The questionnaire also included the follow-
ing relationship history questions. ‘‘How long
have you been close friends with [opposite-sex
friend’s name]?’’ ‘‘How close are you to
[opposite-sex friend’s name]?’’ A 7-point
response scale was provided with the anchors,
not close and very close . ‘‘Have you and
[opposite-sex friend’s name] ever seriously
discussed whether or not to begin a long-term,
committed romantic relationship with each
other?’’ Response options were no and yes.
‘‘Have you and [opposite-sex friend’s name]
ever been in an explicitly long-term, commit-
ted romantic relationship with each other?’’
Response options were no and yes. ‘‘How
many times did you have sex (including manual,
oral, anal, and vaginal sex) with [opposite-sex
friend’s name]?’’ Responses were provided in
a text box. Students indicated their relation-
ship status by which of two links they clicked
Misperception of mating interests 4171
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
to start the study: one link was for single par-
ticipants and the other was for dating partici-
pants. Students’ friends indicated their
relationship status with a no or yes response
to the following question: ‘‘Are you currently
in a long-term, committed romantic relation-
ship?’’ A question at the end of the study asked
participants for comments or suggestions and
included a text box for responses.
The following details about the two study’s
online surveys are based on the Checklist for
Reporting Results on Internet E-Surveys
(Eysenbach, 2006
1). An online survey was pref-
erable to a paper-and-pencil survey because it
provided easy access to the study for the stu-
dents’ friends, who may not be near campus.
We recruited students using Experimetrix
(Sona Systems Ltd., 2006), a Web-based sub-
ject pool management service. Students vali-
dated their identity by logging on to the
study’s homepage using their student ID. Each
student’s opposite-sex friend logged on using
the student’s ID and the password generated
by the student. The Web pages automatically
logged student IDs with each response. In
cases of multiple responses to a page, we first
used submissions with the least missing data,
otherwise we used final submissions. The first
author wrote the Web pages using ColdFusion
markup language (Forta, 1998
2), a dynamic
scripting language closely related to hypertext
markup language. Data collection spanned
November and December 2004 for Study 1
and March and April 2005 for Study 2. We
did not calculate response rates because the
study was not open to the public. In order to
reduce socially desirable responding due to the
presence of participants’ opposite-sex friends
or others, the first and final instructions on
each studies’ first page were for participants
to ‘‘Complete questionnaires privately—with
no one else around (a dorm room is perfect).’’
The instructions also informed participants
that a hidden, password-protected file accessi-
ble only by the researchers would store their
data. Questions asked of students were on one
Web page with 20 questions. The 23 questions
asked of friends were on one Web page. Web
pages automatically required all participants to
respond to each question before continuing to
the next page. A no response option was avail-
able for each question. Questions were not
randomized. Participants could change their
answers with the browser’s Back button. Stu-
dents provided additional data for another
study on attachment after completing Study 1.
Results and Discussion
Overall misperception by sex. See Figure 1
for mean ratings of participant sexual and
romantic interests in their friend and partici-
pant perceptions of their friend’s sexual and
romantic interests in them. A mixed-model 2
2 analysis of variance (ANOVA), with sex of
perceiver a between subjects variable and per-
ception of friend’s sexual interest and friend’s
self-reported sexual interest as within subjects
variables, revealed a significant interaction,
F(1, 222) ¼12.10, p,.05, g
p
2¼.05, indi-
cating that the pattern of misperception of sex-
ual interest differed for men and women. Two
planned contrasts evaluated misperception of
Figure 1. Mean levels (6SE) of self-reported
and perceived sexual and romantic interests of
opposite-sex friends in one another in Study 1.
(A) Sexual interest. (B) Romantic interest.
Note. Inset pvalues are planned comparisons
determined using paired samples ttests com-
paring self-reported interest with friend’s per-
ception of that interest.
418 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
men’s and women’s sexual interest. Replicat-
ing the key findings of previous researchers
and providing evidence that the current opera-
tionalization of sexual interest (as casual sex) is
comparable with those of previous studies,
men overperceived women’s sexual interest,
t(111) ¼2.37, p,.05, d¼.22 and women
underperceived men’s sexual interest, t(111) ¼
22.56, p,.05, d¼.24 (see Figure 1). The
interaction and main effects were not signifi-
cant (ps..05) for a parallel mixed-model 2 
2 ANOVA for romantic interest variables (see
Figure 1), corroborating the results of previous
researchers (e.g., Buss & Schmitt, 1993
3) that
short-term and long-term mating are distinct.
Regression analysis strategy. Previous sex-
ual misperception research has failed to take
into account that the levels of sexual interest
of both the perceiver and the target can poten-
tially influence the perception of the target’s
sexual interest because these perceptions occur
within dyads. Fortunately, the Actor-Partner
Interdependence Model (APIM; Kashy &
AKenny, 2000
4), a multilevel statistical analysis
strategy, allows researchers to estimate the
effects of both the actor (actor effect) and their
partner (partner effect) on the dependent vari-
able. For example, in the current research, the
actor effect is the effect of the perceiver’s
(actor’s) level of sexual interest on the actor’s
perception of their friend’s (partner’s) sexual
interest, controlling for their partner’s level of
sexual interest. The actor effect tests the pro-
jection hypothesis by evaluating the actor’s
perception of their partner’s sexual interest
while controlling for the partner’s self-reported
level of sexual interest. The partner effect is
the effect of the level of the partner’s sexual
interest on the actor’s perception of the part-
ner’s sexual interest. The partner effect can be
interpreted as the degree of accuracy of the
actor’s perceptions of the partner’s sexual
interest.
We conducted the APIM analyses according
to Campbell and Kashy (2002) using hierar-
chical linear modeling (HLM; Raudenbush &
Bryk, 2002). We modeled all predictors at
Level 1, unless otherwise noted. We modeled
opposite-sex friendship dyads at Level 2. We
centered continuous variables around the
grand mean and entered dichotomous varia-
bles uncentered and contrast coded. Although
multilevel models with random effects require
nontraditional mediation procedures, the
APIM uses fixed effects when estimating
regression coefficients (although the error
term for the intercept is allowed to vary ran-
domly; Campbell & Kashy); therefore, tradi-
tional mediation procedures are appropriate
(Kenny, Korchmaros, & Bolger, 2003). HLM
provides unstandardized bvalues.
Tests of mediation include four steps
(Baron & Kenny, 1986; Kenny, 2006). In Step
1, show that the initial variable is correlated
with the outcome variable. To test the media-
tion hypothesis, perceiver sex is the initial var-
iable and perception of friend’s sexual interest
is the outcome variable. If the path coefficient
from the initial variable to the outcome vari-
able when controlling for the mediator is oppo-
site in sign from the product of (a) the path
coefficient from the initial variable to the
mediator and (b) the path coefficient from
the mediator to the outcome variable, then
Step 1 may not be met, but there may still be
mediation. In this scenario, the mediator also
acts as a suppressor variable (Kenny). In Step
2, show that the initial variable is correlated
with the mediator (perceiver’s level of sexual
interest). In Step 3, show that the mediator is
correlated with the outcome variable. In Step
4, show that the effect of the initial variable on
the outcome variable is reduced or eliminated
when controlling for the mediator. For all anal-
yses in this article, we control for friend’s self-
reported sexual interest (the partner effect) as
a covariate at each step so that other variables
predict misperception, that is, perception of
friend’s sexual interest above and beyond their
self-reported sexual interest. We apply the
same logic to romantic interest as we applied
here to sexual interest.
Correlations among primary variables.
Table 1 shows the correlations among primary
variables. Correlations for Study 1 are above
the diagonal. Correlations for Study 2 are
below the diagonal.
The simple sex-difference hypothesis. As
can be seen in Figure 2, the unstandardized path
Misperception of mating interests 4191
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

UNCORRECTED PROOF
coefficient from perceiver sex to perceived
sexual interest was not positive and significant
(which would indicate that men overperceived
and women underperceived) when controlling
for perceiver’s level of sexual interest and
friend’s self-reported sexual interest. The
unstandardized path coefficient from perceiver
sex to perceived romantic interest, controlling
for perceiver’s level of romantic interest and
friend’s self-reported romantic interest, was
not significantly different from zero, b¼
20.01, t(195) ¼20.11, p..05. Thus, the
simple sex-difference hypothesis was not sup-
ported for sexual or romantic interest.
The projection hypothesis. Supporting the
projection hypothesis, while controlling for
perceiver sex and friend’s self-reported inter-
est, the path coefficient between perceiver
level of interest and perceived level of interest
is positive and significant for both sexual interest
Perceiver
Sexual
Interest
(.84*) .42*
Perceiver
Sex
-.21* (.15)
.49* (.63*)
Actual
Sexual
Interest
Perceived
Sexual
Interest
Figure 2. Path diagram for the mediation of
the effect of perceiver sex on the mispercep-
tion of sexual interest by perceiver level of
sexual interest in Study 1.
Note. Unstandardized coefficients in parenthe-
ses are from two models: in one, perceiver sex
predicted perceiver sexual interest; in the
other, perceiver sex and friend’s self-reported
sexual interest (actual sexual interest) pre-
dicted perceived sexual interest. Coefficients
outside parentheses are from a model in which
perceiver sex, perceiver sexual interest, and
friend’s self-reported sexual interest predicted
perceived sexual interest. All variables repre-
sent perceiver data, except actual sexual inter-
est, which the perceiver’s opposite-sex friend
provided.
*p,.05.
Table 1. Correlations for primary variables within opposite-sex friendship dyads in Study 1 and Study 2
Variables
Perceiver
sexa
Perceiver
sexual interest
Perceived
sexual interest
Actual
sexual interest
Perceiver
romantic interest
Perceived
romantic interest
Actual
romantic interest
Short-term
mate value
Perceiver sexa— .35* 2.16* 2.35* .11 2.04 2.11 —
Perceiver sexual interest .31* — .44* .22* .51* .35* .22* —
Perceived sexual interest 2.07 .70* — .63* .26* .56* .30* —
Actual sexual interest 2.31* .40* .65* — .22* .36* .51* —
Perceiver romantic interest .08 .56* .45* .29* — .60* .45* —
Perceived romantic interest .09 .50* .55* .34* .63* — .56* —
Actual romantic interest 2.08 .29* .42* .56* .35* .58* — —
Short-term mate value 2.05 .28* .28* .21* .28* .19* .14 —
Long-term mate value .04 .16* .03 .03 .30* .11 .06 .35*
Note. Correlations for Study 1 are above the diagonal, and correlations for Study 2 are below the diagonal.
aFemale ¼21, male ¼1.
*p,.01.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
420 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar

UNCORRECTED PROOF
(see Figure 2) and romantic interest, b¼0.43,
t(195) ¼7.05, p,.05.
The mediation hypothesis. The mediation
hypothesis predicts that men overperceiving
women’s sexual interest and women underper-
ceiving men’s sexual interest is the result of
members of each sex projecting their different
level of sexual interest onto their friend. As
Figure 2 illustrates, for sexual interest, Step 1
was not met. Nonetheless, when controlling
for the mediator and friend’s self-reported sex-
ual interest, the path coefficient for perceiver
sex was opposite in sign from the product of
(a) the path coefficient from the perceiver sex
to the perceived sexual interest and (b) the path
coefficient from perceiver sexual interest to
perceived sexual interest, suggesting that per-
ceiver sexual interest was acting as a suppres-
sor as well as a mediator of the effects of
perceiver sex (Kenny, 2006).3Steps 2 and 3
are met. Evaluation of Step 4 suggests that
perceiver level of sexual interest mediated
the men’s overperception and women’s under-
perception of sexual interest found in the
ANOVA reported above. A Sobel test con-
firmed that the effect of perceiver sex was sig-
nificantly changed when we added perceiver
level of sexual interest to the model, t¼4.40,
p,.05. After controlling for perceiver level
of sexual interest, the effect of perceiver sex is
significantly different than zero, indicating
that perceiver level of sexual interest partially
mediated the effect of perceiver sex. The neg-
ative sign of the coefficient for perceiver sex
suggests that there was an additional, sup-
pressed effect of sex in which women
overperceived and men overperceived their
friends’ sexual interests. Previous studies have
not used the necessary statistical analyses (or
experimental designs) to find this effect of sex.
Thus, perceiver level of sexual interest not
only mediated men’s overperception of women’s
sexual interest and women’s underperception of
men’s sexual interest but also suppressed an
additional effect of sex in which men under-
perceive and women overperceive the sexual
interest of opposite-sex targets.
Parallel analyses for romantic interest
found that perceiver sex was not a significant
predictor of the perception of romantic interest
in any step, ps..05; therefore, there was no
effect of perceiver sex to be mediated by per-
ceiver level of romantic interest.
Accuracy of perceptions. Recall that, in
these models, the partner effect can be inter-
preted as the degree of accuracy of the actor’s
perceptions of the partner’s sexual and roman-
tic interests. While controlling for perceiver
sex and perceiver level of sexual interest, the
path coefficient for the partner effect, termed
actual interest, is positive and significant for
both sexual interest (see Figure 2) and roman-
tic interest, b¼0.31, t(195) ¼4.47, p,.05.
These findings suggest that perceptions of
a friend’s sexual and romantic interests in one-
self are not pure fantasy, but to a substantial
degree reflect that friend’s actual interests.
Relationship history variables. To exclude
the possibility that relationship history variables
caused the current results, we reran the regres-
sion analyses with the additional relationship
history variables as predictors, including sexual
and romantic histories between friends. As seen
in Table 2, coefficients for critical variables
maintained significance when statistically con-
trolling for relationship history variables. Fur-
thermore, the meaningfulness of the distinction
between romantic and sexual interest is indi-
cated by the inability of perceiver romantic
interest to account for the effect of perceiver
sexual interest, and vice versa. This result also
argues against the possibility that the results sup-
porting the projection hypothesis could be
explained by a common method effect (Kenny,
2006), that is, perceivers rating both their own
level of interest (a predictor) and their perception
of their friend’s interest (the outcome variable)
3. To verify that this result was not due to multicollinear-
ity, the Level 1 data were imported into SPSS
9. Multi-
collinearity analyses for this and all subsequent
mediation models indicated that multicollinearity was
not a problem, VIFs ,.160.
10 Also, as can be seen in
Table 1, the zero-order correlation between perceiver
sex and perceiver sexual interest is positive and signif-
icant in both studies; thus, the reversal in the direction
of the regression coefficient for perceiver sex from
positive to negative when controlling for perceiver
level of sexual interest is not a result of an initial neg-
ative correlation between perceiver sex and perceived
sexual interest.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Misperception of mating interests 421

UNCORRECTED PROOF
because perceiver romantic and sexual interests
were not equally able to predict the mispercep-
tion of both romantic and sexual interests.
Study 2
Study 2 was a replication of Study 1 that also
tested the mate value hypothesis and the
mediation hypothesis with regard to mate
value. Also, Study 2 included several meth-
odological improvements. First, because
someone’s closest opposite-sex friend could
be a relative or a boyfriend or girlfriend (Hen-
drick & Hendrick, 1993), and these relation-
ships are not part of the traditional definition
of opposite-sex friendship (see, e.g., Mon-
sour, 2002), the study instructions explicitly
prohibited participants from recruiting oppo-
site-sex friends who were romantic partners
or relatives. Second, because 3 participants in
Study 1 indicated that one member of the
opposite-sex friendship dyad was a gay male
or lesbian, a question directly asked partici-
pants their sexual orientation. Finally, a ques-
tion at the end of the study asked participants
if their data should be excluded for any rea-
son, with assurance that exclusion would be
free from penalty.
Method
Participants. We excluded the data for five
opposite-sex friendship dyads because one
member was a gay male or lesbian. For five
Table 2. Multiple regression analyses in Study 1 in which hypothesis variables and relationship
history variables predicted the misperception of friend’s sexual and romantic interests (i.e., the
perception of friend’s sexual and romantic interests in the perceiver while controlling for the
friend’s actual sexual and romantic interests in the perceiver, respectively)
Predictor variables
Sexual interest Romantic interest
Unstandardized coefficient SE Unstandardized coefficient SE
Actual interest 0.48* 0.07 0.29* 0.07
Closeness 0.22* 0.10 0.18* 0.09
Length of friendshipa20.37 0.19 20.34* 0.16
Romantic relationship
with friend in past?b, c
0.30 0.25 0.23 0.25
Perceiver relationship
statusd
0.20 0.11 0.12 0.09
Discussed long-term
relationship?
20.10 0.15 0.03 0.13
Times friends
had sexb, e
0.31 0.30 0.24 0.24
Perceiver sexf20.21* 0.09 20.06 0.10
Perceiver romantic
interest in friend
20.03 0.06 0.35* 0.07
Perceiver sexual
interest in friend
0.44* 0.07 0.11* 0.05
Note. Unstandardized regression coefficients for hypotheses variables are italicized.
aBecause length of relationship was positively skewed, it was log transformed.
bThis variable was entered at Level 2.
cNo romantic relationship in past ¼21, romantic relationship in past ¼1.
dSingle ¼21, in a romantic relationship ¼1.
eBecause many participants chose to not reply to this question, a response from either member of the friendship was used
for the sexual history of the friendship. Because times friends had sex was positively skewed, it was log transformed.
fFemale ¼21, male ¼1.
*p,.05.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
422 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar

UNCORRECTED PROOF
other dyads, both members were students,
resulting in duplicative data, so we excluded
one of each set of data at random. Two partic-
ipants indicated their data should be excluded,
so we excluded it. The final sample for Study 2
consisted of 50 female and 49 male introduc-
tory psychology students at the College of
William and Mary, each with a recruited friend
(198 participants total). The mean student age
was 18.85 (SD ¼.72) and the mean friendship
length was 28.81 months (SD ¼27.71).
Questionnaires. Questionnaires used in
Study 2 were identical to those used in Study
1. Chronbach alphas ranged from .93 to .95 for
self-reported and perceived sexual and roman-
tic interest scales, showing satisfactory reli-
ability. Two principal components analyses
with varimax rotations, one for self-reported
interests and another for perception of friend’s
interests, revealed that romantic and sexual
interest items were independent constructs.
In addition, short-term mate value and long-
term mate value questions were based on
Haselton (2003). The Web pages modified
the mate value questions by embedding the
friend’s first name. For a participant whose
friend’s name is John, the question assessing
short-term mate value would be, ‘‘Compared
with others you know who are the same sex as
you and about your age, how desirable do
members of the opposite sex find John as
a short-term mate or casual sex partner?’’
The question assessing long-term mate value
included the words long-term mate or mar-
riage partner instead of short-term mate or
casual sex partner. A 7-point response scale
with the anchors very undesirable and very
desirable followed each mate value question.
Results and Discussion
Overall misperception by sex. Figure 3
presents the mean ratings of participant sexual
and romantic interests in their friend and par-
ticipant perceptions of their friend’s sexual and
romantic interests in them. A mixed-model 2 
2 ANOVA, with sex of perceiver as a between
subjects variable and perception of friend’s
sexual interest and friend’s self-reported sexual
interest as within subjects variables, revealed
a significant interaction, F(1, 186) ¼18.76,
p,.05, g
p
2¼.08, indicating that the pattern
of misperception of sexual interest differed for
men and women. Planned contrasts replicated
the results of Study 1 and the key finding of
previous research: men overperceived wom-
en’s sexual interest, t(93) ¼2.71, p,.05,
d¼.28, and women underperceived men’s
sexual interest, t(93) ¼23.05, p,.05, d¼
.31 (see Figure 3). Parallel analyses for roman-
tic interest variables produced a significant
interaction, F(1, 190) ¼7.75, p,.05, g
p
2¼
.04. Men overperceived women’s romantic
interest, t(94) ¼2.72, p,.05, d¼.28; women
did not misperceive men’s romantic interest,
t(96) ¼21.26, p..05 (see Figure 3).
The simple sex-difference hypothesis. As in
Study 1, perceiver sex was not a positive pre-
dictor of perception of sexual interest while con-
trolling for perceiver level of sexual interest and
Figure 3. Mean levels (6SE) of self-reported
and perceived sexual and romantic interests of
opposite-sex friends in one another in Study 2.
(A) Sexual interest. (B) Romantic interest.
Note. Inset pvalues are planned comparisons
determined using paired samples ttests com-
paring self-reported interest with friend’s per-
ception of that interest.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Misperception of mating interests 423

UNCORRECTED PROOF
friend’s self-reported sexual interest, thus the
simple sex-difference hypothesis was not sup-
ported for sexual interest (see Figure 4). Failing
to replicate the finding of Study 1, perceiver sex
was a positive, significant predictor of romantic
interest while controlling for perceiver level of
romantic interest and friend’s self-reported
romantic interest, b¼0.22, t(161) ¼2.40,
p,.05; therefore, the simple sex-difference
hypothesis was supported for romantic interest.
The projection hypothesis. Consistent with
the projection hypothesis, perceiver interests
were positive, significant predictors of misper-
ception of their friend’s corresponding inter-
ests while controlling for perceiver sex and
friend’s self-reported interests, for both sexual
interest (see Figure 4) and romantic interests,
b¼0.48, t(161) ¼7.88, p,.05, replicating
the findings of Study 1.
The mediation hypothesis. Recall that the
mediation hypothesis predicts that men’s over-
perception of women’s sexual interest and
women’s underperception of men’s sexual
interest is the result of perceivers projecting
their own level of sexual interest onto the tar-
get. As Figure 4 illustrates, Steps 1, 2, and 3
were met for sexual interest. Evaluation of
Step 4 indicated that, replicating the results
found in Study 1, perceiver level of sexual
interest mediated the effect of sex and sup-
pressed an additional effect of sex. The Sobel
test confirmed that the ability of perceiver sex
to predict the misperception of sexual interest
was significantly changed when adding per-
ceiver level of sexual interest to the model,
t¼5.25, p,.05. Thus, the mediation hy-
pothesis was supported for sexual interest.
Parallel mediation analyses for romantic
interest found that, controlling for friend’s
self-reported romantic interest, perceiver sex
was a significant predictor of perception of
romantic interest before, b¼0.26, t(165) ¼
2.22, p,.05, and after, b¼0.22, t(161) ¼
2.40, p,.05, adding perceiver level of roman-
tic interest to the model. Mediation Steps 2 and
4 are not supported and the Sobel’s tis not
significant (p..05). Thus, the mediation
hypothesis was not supported for romantic
interest. This is the only analysis that does
not support the hypothesis that an effect of
perceiver sex is mediated by the projection
of the perceiver’s level of interest.
Mate value hypothesis. The next set of anal-
yses evaluated whether friends’ short- and long-
term mate values predicted misperception of
their sexual and romantic interests, respectively,
and whether misperception of sexual and roman-
tic interests due to friends’ short- and long-term
mate value was mediated by the perceiver level
of sexual and romantic interests. As Figure 5
shows, while controlling friend’s self-reported
sexual interest, short-term mate value was a
positive and significant predictor of perceived
sexual interest, supporting the mate value
hypothesis. All four steps necessary for indicat-
ing mediation were supported for short-term
mate value. A Sobel test was also consistent with
mediation, t¼2.70, p,.05. While controlling
for friend’s self-reported romantic interest,
Perceiver
Sexual
Interest
(.91*) .63*
Perceiver
Sex
-.33* (.24*)
.30* (.64*)
Actual
Sexual
Interest
Perceived
Sexual
Interest
Figure 4. Path diagram for the mediation of
the effect of perceiver sex on the mispercep-
tion of sexual interest by perceiver level of
sexual interest in Study 2.
Note. Unstandardized coefficients in parenthe-
ses are from two models: in one, perceiver sex
predicted perceiver sexual interest; in the
other, perceiver sex and friend’s self-reported
sexual interest (actual sexual interest) pre-
dicted perceived sexual interest. Coefficients
outside parentheses are from a model in which
perceiver sex, perceiver sexual interest, and
friend’s self-reported sexual interest predicted
perceived sexual interest. All variables repre-
sent perceiver data, except actual sexual inter-
est, which the perceiver’s opposite-sex friend
provided.
*p,.05.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
424 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar

UNCORRECTED PROOF
long-term mate value did not predict the mis-
perception of romantic interest, b¼0.15,
t(159) ¼1.28, p..05, failing to support the
mate value hypothesis for long-term interest and
indicating that there is not an effect of long-term
mate value to be mediated by perceiver level of
romantic interest. In sum, the results supported
the mate value hypothesis and the mediation
hypothesis for short-term mate value but not
long-term mate value.
Accuracy of perceptions. As in Study 1,
while controlling for perceiver sex and per-
ceiver level of interest, friend’s self-reported
interest was a significant predictor of percep-
tion of that interest for both sexual interest (see
Figure 5) and romantic interest, b¼0.37,
t(161) ¼6.48, p,.05. These findings indicate
that participant perceptions of their friend’s
sexual and romantic interests, to a substantial
degree, accurately reflected those interests.
Relationship history variables. Table 3
presents the results of regression analysis that
include theoretically relevant variables as well
as relationship history variables as predictors.
Replicating the findings of Study 1, critical
variables were in the same direction and main-
tained significance when statistically control-
ling for relationship history variables. Notably,
many results for the relationship history vari-
ables themselves did not replicate across stud-
ies, thus deserving replication. Discussion of
these inconsistencies is beyond the scope of
the current article.
General Discussion
Two studies found that men overperceived
women’s sexual interest and women underper-
ceived men’s sexual interest, replicating the
key finding of numerous studies (e.g., Abbey,
1982). Only men in Study 2 misperceived
romantic interest. Further analyses in both
studies provided evidence that perceivers pro-
jected their own levels of sexual and romantic
interests onto their opposite-sex friend. Study
2 found that the short-term mate value of
targets, but not their long-term mate value,
also resulted in systematic misperception.
Mediation analyses were consistent with the
hypothesis that perceiver level of sexual inter-
est in the target resulted in the effects of both
perceiver sex (in Studies 1 and 2) and target
short-term mate value (in Study 2); however,
perceiver level of romantic interest did not
mediate men’s overperception of women’s
romantic interest in Study 2. Surprisingly,
mediation analyses in both studies revealed
that perceiver level of sexual interest was sup-
pressing an additional effect of sex in which
men underperceived women’s sexual interest
and women overperceived men’s sexual interest.
The simple sex-difference hypothesis
Replicating numerous studies, the overall effect
of sex (as indicated in Figures 1 and 3) was that
men overperceived and women underperceived
the sexual interest of an opposite-sex target.
Perceiver
Sexual
Interest
(.28*) .52*
.07 (.21*)
.39* (.56*)
Actual
Sexual
Interest
Perceived
Sexual
Interest
Short-term
Mate
Value
Figure 5. Path diagram for the mediation of
the effect of friend’s short-term mate value on
the misperception of sexual interest by per-
ceiver level of sexual interest in Study 2.
Note. Unstandardized coefficients in parenthe-
ses are from two models: in one, friend’s short-
term mate value predicted perceiver sexual
interest; in the other, friend’s short-term mate
value and friend’s self-reported sexual interest
in the perceiver (actual sexual interest) pre-
dicted perceived sexual interest. Unstandard-
ized coefficients outside parentheses are from
a model in which friend’s short-term mate
value, perceiver sexual interest, and friend’s
self-reported sexual interest predicted per-
ceived sexual interest. All variables represent
perceiver data, except actual sexual interest,
which the perceiver’s opposite-sex friend pro-
vided.
*p,.05.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Misperception of mating interests 425

UNCORRECTED PROOF
Nevertheless, both studies found the reverse
effect when holding perceiver level of sexual
interest constant: women overperceived and
men underperceived the sexual interest of their
opposite-sex friend. No socialization or evolu-
tionary theories known to the authors predicts
women’s overperception and men’s underper-
ception of the sexual interest of opposite-sex
targets. A plausible explanation is that people
are aware that men have higher levels of sexual
interest than do women (Baumeister et al.,
2001), and therefore, in ambiguous situations,
people may assume that a man has a high level
of sexual interest, whereas a woman has a low
level of sexual interest. That is, people may use
common knowledge about actual sex differen-
ces in levels of sexual interest to discount their
estimates of women’s sexual interest and
increase their estimates of men’s sexual interest.
Thus, socially acquired knowledge that men
have more sexual interest than women may lead
to a sex-of-target effect resulting not in men
overperceiving women’s sexual interest (Abbey,
1982) but instead to the opposite: men under-
perceiving (and women overperceiving) of the
sexual interest of members of the opposite sex.
The projection hypothesis
Researchers have interpreted functional pro-
jection (Maner et al., 2005) and EMT (Hasel-
ton & Buss, 2000) to support male, but not
Table 3. Multiple regression analyses in Study 2 in which hypothesis variables and relationship
history variables predicted the misperception of friend’s sexual and romantic interests (i.e., the
perception of friend’s sexual and romantic interests in the perceiver while controlling for the
friend’s actual sexual and romantic interests in the perceiver, respectively)
Predictor variables
Sexual interest Romantic interest
Unstandardized coefficient SE Unstandardized coefficient SE
Actual interest 0.24* 0.07 0.24* 0.06
Closeness 20.08 0.08 20.08 0.10
Length of friendshipa0.02 0.16 20.13 0.18
Romantic relationship
with friend in past?b, c
20.57* 0.15 0.11 0.10
Perceiver relationship
statusd
20.17* 0.09 0.04 0.10
Discussed long-term
relationship?
0.18* 0.06 0.20* 0.05
Times friends
had sexb, e
0.66* 0.19 0.00 0.16
Perceiver sexf20.29* 0.08 0.19 0.10
Friend’s mate value 0.02 0.06 20.04 0.09
Perceiver romantic
interest in friend
20.01 0.06 0.39* 0.08
Perceiver sexual
interest in friend
0.53* 0.06 0.06 0.07
Note. Unstandardized regression coefficients for hypotheses variables are italicized.
aBecause length of relationship was positively skewed, it was log transformed.
bThis variable was entered at Level 2.
cNo romantic relationship in past ¼21, romantic relationship in past ¼1.
dSingle ¼21, in a romantic relationship ¼1.
eMany participants chose to not reply to this question; therefore, a response from either member of the friendship was
used for the sexual history of the friendship. Because times friends had sex was positively skewed, it was log transformed.
fFemale ¼21, male ¼1.
*p,.05.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
426 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar

UNCORRECTED PROOF
female, misperception of the sexual interest of
opposite-sex targets. How can functional pro-
jection and EMT accommodate the current
findings that level of sexual interest, not
maleness or femaleness per se, best predicted
misperception of both sexual and romantic
interests? Consider when sexual interest is
not aroused. An unknown woman with low
current reproductive potential, elderly, prepu-
bescent, or very ill, for example, arouses little
sexual interest in most normal men. The more
indicators that a woman is fertile and of high
mate quality, the more sexually interested men
will be in her, and the more useful it is to over-
perceive her sexual interest so as not to miss
the valuable mating opportunity. Compared to
men, the sexual interest of women is less easily
aroused (Baumeister et al., 2001), reflecting
women’s greater discrimination of sexual part-
ners (Trivers, 1972). Once women are sexually
aroused, women may benefit from overperceiv-
ing the sexual interest of their target just like
men.Thiscanbeunderstoodbythinkingof
arousal of sexual interest as a sexual-opportunity
meter, that is, an indicator that a target is a
valuable potential mate whose interest would
be costly to miss, as in the affect-as-informa-
tion model of emotions (see, e.g., Clore &
Storbeck, in press). From an evolutionary per-
spective, men do not want to miss an opportu-
nity to have sex with most women, and women
do not want to miss an opportunity to have sex
with a high mate value man, thus both men and
women may benefit from overperceiving the
sexual interest of those who activate their sex-
ual interest. Similarly, passionate love may act
as a romantic-opportunity meter, indicating
that a target would be a worthwhile long-term
romantic partner, and therefore, missing the
valuable opportunity by underperceiving their
romantic interest would be more costly than
overperceiving their romantic interest.
The mate value hypothesis
Replicating the finding of Haselton (2003),
target short-term mate value predicted misper-
ception of that target’s sexual interest. This
finding corroborates the argument presented
above that it is functional to overperceive the
sexual interest of those with high short-term
mate value. The effect of mate value on mis-
perception disappeared; however, once we
held participants’ own level of sexual interest
constant, suggesting that the perceivers’ own
level of sexual interest may mediate the mis-
perception of sexual interest related to short-
term mate value. The study used a single item
to measure short-term mate value, however, so
caution is warranted in concluding that medi-
ation occurred. Long-term mate value did not
predict the misperception of romantic interest,
perhaps because, unlike casual sex, romantic
love involves commitment (Frank, 1988;
Gonzaga, Keltner, Londal, & Smith, 2001 5
;
Ketelaar & Goodie, 1998) and therefore
greater potential costs if someone were to
exploit this commitment.
Limitations
First, as in all self-report data, but especially
those about one’s relationship with an
opposite-sex friend (Monsour, Harris, Kurz-
wil, & Beard, 1994 6
), socially desirable
responding, or self-deception (Paulhus, 1984)
may influence the results. Second, these stud-
ies had high cancellation rates (23.6% in Study
1 and 13.2% in Study 2), perhaps because
potential participants wanted to avoid discus-
sing romantic or sexual interests with their
opposite-sex friend (Afifi & Burgoon, 1998).
Third, the use of naturally occurring mating
interests provides evidence for the external
validity of the existing research on the misper-
ception of sexual interest, but necessitates cor-
relational instead of experimental methods for
evaluating causation. Experimental manipula-
tion of perceiver’s level of sexual interest is
necessary to demonstrate the causal role of
perceiver’s level of sexual interest in the mis-
perception of sexual interest. Fourth, the same
survey measured level of sexual interest and
perception of friend’s level of sexual interest;
thus, a common method effect may have
caused the high correlation between the medi-
ator and the outcome variable (Kenny, 2006),
although Tables 2 and 3 present evidence
against this possibility. Finally, the use of volun-
teer, convenience samples does not allow gener-
alization to a known population. Future research
should attempt to address these limitations.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Misperception of mating interests 427

UNCORRECTED PROOF
Conclusions
The current findings suggest a number of
research directions and practical implications.
First, if misperception of sexual interest is pri-
marily due to projection of one’s own level of
interest, then individual differences in level of
sexual interest may be a factor underlying risk
for perpetrating sexual assault, including
acquaintance rape. This assertion is supported
by the finding that men who have committed
acquaintance rape tend to be very sexually
active (Kanin, 1985). Similarly, if the percep-
tion that another person is sexually interested
in oneself is increased by being sexually inter-
ested in that person, then individuals with
a chronically high sex drive are at an increased
risk of unwittingly sexually harassing others.
Education about how people may project their
own level of sexual interest, about what level
of sexual interest is common for different
groups (e.g., women), and about valid indica-
tors that may be used to ascertain another’s
level of sexual interest may reduce the risk
for sexual aggression and harassment for
those with chronically high levels of sexual
interest. Finally, the current studies support
the advice given by others that men may need
to be skeptical of their perceptions of wom-
en’s sexual interest in them (Abbey & Har-
nish, 1995). The current research also
suggests that a complementary set of advice
is warranted. That is, women may tend to
underperceive the sexual interest of men
who do not arouse their sexual interest and
therefore, experience an unreciprocated sense
of nonsexuality or friendliness, perhaps result-
ing in a false sense of security. Furthermore,
women may be advised to use men’s behav-
iors, such as sexual innuendos or sexual
advances, as indicators not only that he is sex-
ually interested in her but also that he may
think she is sexually interested in him.
Women should be careful not to discount esti-
mates of men’s sexual interest based on their
own lack of sexual interest.
References
Abbey, A. (1982). Sex differences in attributions for
friendly behavior: Do males misperceive females’
friendliness? Journal of Personality and Social Psy-
chology,42 , 830–838.
Abbey, A. (1987). Misperceptions of friendly behavior as
sexual interest: A survey of naturally occurring inci-
dents. Psychology of Women Quarterly ,11, 173–194.
Abbey, A., & Harnish, R. J. (1995). Perception of sexual
intent: The role of gender, alcohol consumption and
rape supportive attitudes. Sex Roles ,32, 297–313.
Abbey, A., McAuslan, P., & Ross, L. T. (1998). Sexual
assault perpetration by college men: The role of alco-
hol, misperception of sexual intent, and sexual beliefs
and experiences. Journal of Social and Clinical Psy-
chology,17 , 167–195.
Abbey, A., & Melby, C. (1986). The effects of nonverbal
cues on gender differences in perceptions of sexual
intent. Sex Roles,15, 283–289.
Abbey, A., Zawacki, T., & McAuslan, P. (2000). Alco-
hol’s effects on sexual perception. Journal of Studies
on Alcohol,61 , 688–697.
Afifi, W. A., & Burgoon, J. K. (1998). We never talk about
that: A comparison of cross-sex friendships and dating
relationships on uncertainty and topic avoidance.
Personal Relationships,5, 255–272.
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, 205–222.
Baron, R. M., & Kenny, D. A. (1986). The moderator-
mediator variable distinction in social psychological
research: Conceptual, strategic and statistical consid-
erations. Journal of Personality and Social Psychol-
ogy,51 , 1173–1182.
Basow, S. A. (1992). Gender: Stereotypes and roles (3rd
ed.). Pacific Grove, CA: Wadsworth;Brooks/Cole. 7
Baumeister, R. F., Catanese, K. R., & Vohs, K. D. (2001).
Is there a gender difference in strength of sex drive?
Theoretical views, conceptual distinctions, and
a review of relevant evidence. Personality and Social
Psychology Review,5, 242–273.
Bleske, A. L., & Buss, D. M. (2000). Can men and women
be just friends? Personal Relationships,7, 131–151.
Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies
theory: An evolutionary perspective on human mating.
Psychological Review,100 , 204–232.
Campbell, L., & Kashy, D. A. (2002). Estimating actor,
partner, and interaction effects for dyadic data using
PROC MIXED and HLM: A guided tour. Personal
Relationships,9, 327–342.
Clore, G. L., & Storbeck, J. (in press). Affect as informa-
tion about liking, efficacy, and importance. In
J. Forgas (Ed.), Hearts and minds: Affective influences
on social cognition and behavior. New York: Psychol-
ogy Press. 8
Donat, P., & Bondurant, B. (2003). The role of sexual
victimization in women’s perceptions of others’ sexual
interest. Journal of Interpersonal Violence ,18, 50–64.
Edmondson, C. B., & Conger, J. C. (1995). The impact of
mode of presentation on gender differences in social
perception. Sex Roles ,32, 169–183.
Eysenbach, G. (2004 September 29). Improving the qual-
ity of Web surveys: The Checklist for Reporting
Results of Internet E-Surveys (CHERRIES). Journal
of Medical Internet Research,6, e34. Retrieved
September 12, 2006, http://www.jmir.org/2004/3/e34/
Fisher, T. D., & Walters, A. S. (2003). Variables in addi-
tion to gender that help to explain differences in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
428 B. L. Koenig, L. A. Kirkpatrick, and T. Ketelaar

UNCORRECTED PROOF
perceived sexual interest. Psychology of Men and
Masculinity,4, 154–162.
Forta, B. (1998). The ColdFusion 4.0 web application con-
struction kit. Indianapolis, IN: Que Publishing.
Frank, R. (1988). Passions within reason: The strategic
role of the emotions. New York: Norton.
Goodchilds, J. D., & Zellman, G. L. (1984). Sexual sig-
naling and sexual aggression in adolescent relation-
ships. In N. M. Malamuth & E. Donnerstein (Eds.),
Pornography and sexual aggression (pp. 233–243).
Orlando, FL: Academic Press.
Gonzaga, G., Keltner, D., Londal, E., & Smith, D. (2001).
Love and the commitment problem in romantic rela-
tions and friendship. Journal of Personality and Social
Psychology,81, 247–262.
Harnish, R. J., Abbey, A., & DeBono, K. G. (1990). Toward
an understanding of the ‘‘sex game’’: The effects of
gender and self-monitoring in initial interactions. Jour-
nal of Applied Social Psychology,20, 1333–1344.
Haselton, M. G. (2003). The sexual overperception bias:
Evidence of a systematic bias in men from a survey of
naturally occurring events. Journal of Research in
Personality,37, 43–47.
Haselton, M. G., & Buss, D. M. (2000). Error management
theory: A new perspective on biases in opposite-sex
mind reading. Journal of Personality and Social Psy-
chology,78 , 81–91.
Hendrick, C., & Hendrick, S. S. (1986). A theory and
method of love. Journal of Personality and Social
Psychology,50, 392–402.
Hendrick, C., Hendrick, S., Foote, F. H., & Slapion-Foote, M.
J. (1984). ‘‘Do men and women love differently?’’ Jour-
nal of Social and Personal Relationships ,1, 177–195.
Hendrick, S. S., & Hendrick, C. (1993). Lovers as friends.
Journal of Social and Personal Relationships,10 ,
459–466.
Johnson, C. B., Stockdale, M. S., & Saal, F. E. (1991).
Persistence of men’s misperceptions of friendly cues
across a variety of interpersonal encounters. Psychol-
ogy of Women Quarterly,15 , 463–475.
Kanin, E. J. (1985). Date rapists: Differential sexual
socialization and relative deprivation. Archives of Sex-
ual Behavior,14, 219–231.
Kaplan, D. L., & Keys, C. B. (1997). Sex and relationship
variables as predictors of sexual attraction in cross-sex
platonic friendships between young heterosexual
adults. Journal of Social and Personal Relationships ,
14, 191–206.
Kashy, D., & AKenny, D. A. (2000). The analysis of data
from dyads and groups. In H. T. Reis & C. M. Judd
(Eds.), Handbook of research methods in social and
personality psychology (pp. 451–477). New York:
Cambridge University Press.
Kenny, D. A. (2006). Mediation. Retrieved September 12,
2006, http://davidakenny.net/cm/mediate.htm
Kenny, D. A., Korchmaros, J. D., & Bolger, N. (2003).
Lower level mediation in multilevel models. Psycho-
logical Methods,8, 115–128.
Kenrick, D. T., Sadalla, E. K., Groth, G., & Trost, M. R.
(1990). Evolution, traits, and stages of human court-
ship: Qualifying the parental investment model. Jour-
nal of Personality,58 , 97–115.
Ketelaar, T., & Goodie, A. S. (1998). The satisficing role
of emotions in decision-making. Psykhe ,7,
63–77.
Kowalski, R. M. (1993). Inferring sexual interest from
behavioral cues: Effects of gender and sexually rele-
vant attitudes. Sex Roles ,29, 13–36.
Lee, J. A., (1973). The colors of love: An exploration of the
ways of loving. Don Mills, Ontario, Canada: New
Press, (Popular Edition, 1976).
Maner, J. K., Kenrick, D. T., Becker, D. V., Robertson, T.
E., Hofer, B., Neuberg, S. L. et al. (2005). Functional
projection: How fundamental social motives can bias
interpersonal perception. Journal of Personality and
Social Psychology,88, 63–78.
Merton, R. K. (1968). Social theory and social structure.
New York: Free Press.
Monsour, M. (2002). Women and men as friends: Rela-
tionships across the lifespan in the 21st century.
Mahwah, NJ: Lawrence Erlbaum.
Monsour, M., Harris, B., Kurzwil, N., & Beard, C. (1994).
Challenges confronting cross-sex friendships: Much
ado about nothing? Sex Roles ,31, 55–77.
O’Meara, J. D. (1989). Cross-sex friendship: Four basic
challenges of an ignored relationship. Sex Roles ,21,
525–543.
Paulhus, D. L. (1984). Two-component models of socially
desirable responding. Journal of Personality and
Social Psychology,46, 609.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical
linear models: Applications and data analysis methods
(2nd ed.). Thousand Oaks, CA: Sage.
Reeder, H. M. (2000). ‘I like you .as a friend’: The role
of attraction in cross-sex friendship. Journal of Social
and Personal Relationships,17, 329–348.
Rose, S. M. (1985). Same- and cross-sex friendship and the
psychology of homosociality. Sex Roles ,12, 63–74.
Rubin, Z., Peplau, L. A., & Hill, C. T. (1981). Loving and
leaving: Sex differences in romantic attachment. Sex
Roles,7, 821–835.
Saal, F., Johnson, C., & Weber, N. (1989). Friendly or
sexy? It may depend on whom you ask. Psychology
of Women Quarterly,13 , 263–276.
Shea, M. E. C. (1993). The effects of selective evaluation
on the perception of female cues in sexually coercive
and noncoercive males. Archives of Sexual Behavior ,
22, 415–433.
Sigal, J., Gibbs, M., Adams, B., & Derfler, R. (1988). The
effect of romantic and nonromantic films on percep-
tion of female friendly and sexual behavior. Sex Roles ,
19, 545–554.
Shotland, R. L., & Craig, J. M. (1988). Can men and
women differentiate between friendly and sexually
interested behavior? Social Psychology Quarterly ,
51, 66–73.
Sona Systems Ltd. (2006). Experimetrix. Retrieved May
25, 2007, http://www.experimetrix.com
Sprecher, S., & Toro-Morn, M. (2002). A study of men and
women from different sides of Earth to determine if
men are from Mars and women are from Venus in their
beliefs about love and romantic relationships. Sex
Roles,46 , 131–147.
Trivers, R. L. (1972). Parental investment and sexual
selection. In B. Campbell (Ed.), Sexual selection and
the decent of man: 1871-1971 (pp. 136–179). Chicago:
Aldine.
Trivers, R. L. (2000). The elements of a scientific theory of
self-deception. Annals of the New York Academy of
Sciences,907 , 114–131.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Misperception of mating interests 429

UNCORRECTED PROOF
Author Query Form
Journal: Personal Relationships
Article : pere_163
Dear Author,
During the copy-editing of your paper, the following queries arose. Please respond to
these by marking up your proofs with the necessary changes/additions. Please write your
answers on the query sheet if there is insufficient space on the page proofs. Please write
clearly and follow the conventions shown on the attached corrections sheet. If returning the
proof by fax do not write too close to the paper’s edge. Please remember that illegible mark-
ups may delay publication.
Many thanks for your assistance.
Query
No.
Query Remark
1Please note that ‘‘Eysenbach (2004) has been changed to
‘‘Eysenbach (2006)’’ as per the list. Please check.
2Please note that spelling of the author ‘‘Forte (1998)’’
has been changed to ‘‘Forta (1998)’’ as per the list.
Please check.
3Please note that spelling of the author ‘‘Buss and Scmitt
(1993)’’ has been changed to ‘‘Buss and Schmitt
(1993)’’ as per the list. Please check.
4Please note that spelling of the author ‘‘Kashy and
Kenny (2000)’’ has been changed to ‘‘Kashy and
AKenny (2000)’’ as per the list. Please check.
5Please note that spelling of the author ‘‘Londahl’’ in
‘‘Gonzaga, Keltner, Londahl, and Smith (2001)’’ has
been changed to ‘‘Londal’’ as per the list. Please check.
6Please note that spelling of the author ‘‘Kurzweil’’ in
‘‘Monsour, Harris, Kurzweil, and Beard (1994)’’ has
been changed to ‘‘Kurzwil’’ as per the list. Please check.
7Please retain any one publisher name as per the journal
style for ‘‘Basow (1992).’’
8Please update ‘‘Clore and Storbeck (in press).’’

UNCORRECTED PROOF
9Please spell out ‘‘SPSS’’ in the sentence ‘‘To verify that
this result ..’’
10 Please spell out ‘‘VIF’’ in the sentence ‘‘Multicollinear-
ity analyses for ..’’