Content uploaded by Omri Gillath
Author content
All content in this area was uploaded by Omri Gillath on Nov 19, 2020
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=hjsr20
The Journal of Sex Research
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/hjsr20
Identifying a Facial Expression of Flirtation and Its
Effect on Men
Parnia Haj-Mohamadi , Omri Gillath & Erika L. Rosenberg
To cite this article: Parnia Haj-Mohamadi , Omri Gillath & Erika L. Rosenberg (2020): Identifying
a Facial Expression of Flirtation and Its Effect on Men, The Journal of Sex Research, DOI:
10.1080/00224499.2020.1805583
To link to this article: https://doi.org/10.1080/00224499.2020.1805583
Published online: 03 Sep 2020.
Submit your article to this journal
View related articles
View Crossmark data
Identifying a Facial Expression of Flirtation and Its Eect on Men
Parnia Haj-Mohamadi
a
, Omri Gillath
a
, and Erika L. Rosenberg
b
a
Department of Psychology, University of Kansas;
b
Center for Mind and Brain, University of California, Davis
ABSTRACT
Internal states may be conveyed to others nonverbally through facial expression. We investigated the
existence of a particular facial cue that may be eectively used by women to indicate interest in a man.
Across six studies, men generally recognized a female facial expression as representing irting. Flirtatious
expressions receiving low recognition by men diered in morphology from the highly recognized irting
expressions. The discrepancies are indicative of individual dierences among women in eectively
conveying a irtatious facial cue and among men in recognizing this cue. The morphology of the highly
recognized irtatious facial expressions, coded using the Facial Action Coding System (FACS), included:
a head turned to one side and tilted down slightly, a slight smile, and eyes turned forward (toward the
implied target). Results from experimental studies showed that irtatious facial expressions, as compared
with happy or neutral expressions, led to faster identication of sex words by men. These ndings support
the role of irtatious expression in communication and mating initiation.
As sexual creatures, people need to secure a mate to repro-
duce. Securing a sexual mate often necessitates the identi-
fication of an appropriate, and preferably available and
willing, mate and the communication of interest in them.
Effective communication should facilitate mating, whereas
ineffective communication could result in negative out-
comes for both sides (e.g., error management theory; Hall
et al., 2015; Haselton & Buss, 2000). For example, effective
communication conveys availability of a specific mate at
a specific time (e.g., the swellings of female baboons;
Dunbar, 2001), as well as the quality of a potential mate
(e.g., peacock’s tail or train; Petrie et al., 1991), all of which
increase reproductive success. Conversely, ineffective com-
munication can result in missed opportunities. For women,
ineffectively conveying interest in a man may cause
a woman to miss the opportunity to mate with a high-
status man, which could lead to a missed opportunity for
acquiring resources for her and her offspring. Ineffective
communication also has negative impacts on men, such
that a man may miss an opportunity to mate with an
available woman, and in turn miss the opportunity to
spread his genes and thus produce more offspring
(Haselton & Buss, 2000).
One channel through which people can convey interest in
a partner is nonverbal communication (Henningsen et al.,
2008; Koeppel et al., 1993). Indeed, Moore (2010) recently
reviewed ways that nonverbal communication is involved in
human courtship. In particular, facial expressions are known to
be an indicator of internal states and to play an important role
in nonverbal communication (e.g., Ekman, 1992; Gottman
et al., 2001), and have been suggested to be an efficient way
to facilitate relationship initiation and mate-selection processes
(see Keltner, 2003). To date, little systematic research has
examined the specific nonverbal facial cues involved in the
initiation of the courtship process.
An effective tactic of conveying interest and getting the
attention of a potential mate is often referred to as irting
(Moore, 2010). Nonverbal flirting behaviors, such as sustained
eye contact, smiling, coy gazing and self-touching were already
found to play an important role in the initiation or courtship
process (Henningsen et al., 2008; Muehlenhard et al., 1986;
Renninger et al., 2004; Tisdale & Sheldon, 2018). Flirting beha-
viors tend to be displayed by both people involved in the
initiation process, providing them with a way to communicate
their interest while possibly evoking interest from their potential
mate (Keltner & Haidt, 1999; Renninger et al., 2004). Thus, one
partner can communicate interest through flirting, potentially
leading to a reciprocal response, in turn facilitating the initiation
of mating or even a full-fledged relationship.
Flirting is communicative, yet subtle (Speer, 2017; White
et al., 2018), leaving open the options for how (or if) to
proceed. That is, using flirting, people can convey interest in
a subtle way that allows them to easily retreat from the encoun-
ter if needed. This subtleness results in ambiguity, such that
perceivers can interpret flirting behaviors in different ways
such as sexual interest, friendliness, or mere academic interest
(e.g., Henningsen et al., 2008). This, in turn, provides time for
the initiator to determine if they want to further engage with
the potential mate.
The ambiguity in the conveyed message is especially
important for women. According to evolutionary theories,
a different set of challenges (e.g., increased investment in
childrearing such as pregnancy and nursing) motivates
women to be more selective than men when choosing
CONTACT Parnia Haj-Mohamadi phajmohamadi@ku.edu Department of Psychology, University of Kansas, Lawrence, KS 66044-7556
Data for this project is available on OSF at https://osf.io/hu735/.
THE JOURNAL OF SEX RESEARCH
https://doi.org/10.1080/00224499.2020.1805583
© 2020 The Society for the Scientific Study of Sexuality
a mate (Trivers, 1972). For example, committing to the
“wrong” partner (e.g., one that would not support her and
her offspring), or not getting the attention of the “right”
partner (e.g., one that could have provided her and her off-
spring’s needs) can be detrimental to women and their off-
spring’s success. To enable selectivity, women should be able
to control – at least to some extent – the initial interaction in
opposite-sex encounters. This allows a woman to choose who
would be encouraged to interact with her and to what extent
the interaction would be allowed to proceed (Givens, 1978).
Hence, a woman who uses cues to increase the likelihood of
a man’s approach while “testing the waters” would have an
advantage over women who do not use such cues. From an
evolutionary perspective (e.g., Buss, 1989), women would
have been expected to have developed such cues – cues that
manifest as flirting behavior.
Men, conversely, had a different set of challenges to cope
with throughout evolution and hence developed different
mating goals and strategies. From an evolutionary perspec-
tive, selectivity is less important for men: their main goal is
to avoid missed mating opportunities. For men, every mat-
ing opportunity can potentially result in the obtainment of
their ultimate evolutionary goal – passing on their genes.
Missing such opportunities carries a genetic cost; therefore,
men are expected to be highly motivated to detect signs of
interest that are conveyed by women (Abbey, 1982;
Haselton & Buss, 2000). At the same time, courting
a woman who is not interested can result in negative out-
comes such as rejection, wasted resources, and even retalia-
tion (Berscheid et al., 1971). Men, therefore, try to avoid
these potential negative outcomes by being as accurate as
possible in identifying women’s cues of interest. Being
successful at correctly identifying interest while minimizing
the chance for false-positives would be advantageous for
men; thus, men are expected to develop the capability to
identify flirtatious cues posed by women. Koeppel et al.
(1993) found that men tend to adopt the belief that flirting
is an indicant of invitation, or a cue to act and avoid
missing an opportunity.
Cross-cultural observational research suggests that female
flirting cues do exist, and specifically in the form of facial
expression: a “coy glance,” involving a downward gaze and
a half-smile (Eibl-Eibesfeldt, 1971). In the laboratory, using
Eibl-Eibesfeldt’s description of a flirting expression, coders
were able to rate interest among women using the number of
coy glances (Simpson et al., 1993). Based on this previous work,
in the current set of studies, using a variety of methods, we
examined whether: (1) a specific facial expression can be gen-
erated by women to represent flirting (defined as a subtle cue
communicating interest in the courtship process; Studies 1–2);
(2) this expression is recognized by men and is distinguishable
from other expressions (Study 3); (3) assuming such an expres-
sion exists, we wanted to define its morphology using the Facial
Action Coding System (FACS; Ekman & Friesen, 1978; Study,
p. 4); and (4) assuming it exists, this expression should convey
meaning relevant to mating, such that exposing people to
flirtatious expressions, compared to happy or neutral expres-
sions, will result in higher accessibility of sex-related words,
especially in men (Studies 5–6).
Method
The studies presented here were designed following the proce-
dures of Tracy and Robins’ (2004) identification of a pride
expression. A total of 482 pictures were taken of nine female
posers. The women were either professional actresses (ones who
had specifically participated in Tracy and Robins’ work) or
women who, in a preliminary interview, reported having flirted
in the past. The pictures consisted of happy and neutral expres-
sions (controls), and flirtatious expressions (experimental) that
were made spontaneously by the poser or as instructed by the
experimenter. The instructions for posing happiness expressions
utilized anatomically-based instructions from the directed facial
action task (Ekman et al., 1983). For flirtation, the exact form of
which had not yet been specified, we provided anatomically-
based suggestions, based on Eibl-Eibesfeldt’s (1971) journals in
some pictures, or instructed the poser to generate what she
thought was a flirtatious facial expression in other pictures. All
posers wore a white t-shirt and were photographed against
a blue background. All pictures were cropped so that nothing
was visible below the shoulders (see Figure 1). Each picture was
then rated by 117 men (M age = 20.57; range 18–41 years) at
a large Midwestern university in one of six studies for course
credit.
Study 1
The goal of Study 1 was to determine whether men perceived
certain facial expressions as indicative of flirting and, if so, to
narrow down the large stimulus pool into a smaller number of
pictures that were rated as highly representative of a flirtatious
facial expression. Specifically, we exposed men to each picture
(500 ms) and recorded their immediate reactions to the facial
expressions conveyed by the female actresses. This allowed us
to capture participants’ ranking of the fit of each picture to each
of the three categories (flirt, happy, neutral).
Participants, Materials, and Procedure
Ten men participated in Study 1. Men were presented with 482
pictures in random order on a computer screen. Because we
wanted to receive participants’ immediate or gut reaction to the
facial expressions, each picture was presented for 500 ms.
Following the presentation of the picture, participants rated how
well each picture fit with the label: “flirtatious,” “happy,” and
“neutral” on a 7-point Likert scale ranging from 1 (not at all) to
7 (extremely). We included neutral pictures by the same posers to
control for the possibility that some women were being perceived
as flirtatious regardless of the expression they posed. We also
included photos of happy facial expressions (i.e., enjoyment smiles,
as per Ekman et al., 1990) to ensure that general positive facial
expressions were not perceived as flirtatious by men. The initial set
of pictures contained approximately 50% flirtatious expressions (n
= 242), 25% happy expressions (n = 120), and 25% neutral expres-
sions (n = 120).
Results
Ratings for flirtatious expressions were averaged across parti-
cipants. For the 482 pictures, the average ratings of flirtatious
2P. HAJ-MOHAMADI ET AL.
expressions ranged from 1.7 to 6.3, with a mean rating of 3.8.
The 18 pictures that were rated above the mean on flirting were
retained for Study 2.
Study 2
Although Study 1 revealed that certain pictures were more
likely than others to be rated as flirtatious, we were unsure of
the standard that men used to inform their ratings for flirta-
tious expressions. Thus, in Study 2, we provided participants
with a definition for flirting and then asked them to rate the set
of pictures selected on the basis of Study 1. The goal of this
study was to further parse our large pool of photographs to
uniquely identify flirtatious expressions from happy and neu-
tral expressions with as little overlap as possible.
Participants, Materials, and Procedure
Twenty-six men participated in Study 2. Before showing the
photos, we provided participants with a common definition
of flirting: “A form of human interaction, usually expressing
a sexual or romantic interest in the other person. It can
consist of conversation, body language, or brief physical
contact. It may be one-sided or reciprocated” (Henningsen
et al., 2008). Participants were then presented with 233
pictures of flirtatious, happy, and neutral facial expressions.
For each picture, men were asked to imagine seeing the
woman with the shown expression at a party or bar and
rate how much she fits with the provided definition of
flirting. Responses were made on a 5-point Likert scale
ranging from 1 (not at all) to 5 (very much). All pictures
were presented on a computer and remained on the screen
until the participant indicated his response. The order of
presentation was randomized across participants. Similar to
Study 1, approximately half of the pictures included
a flirtatious expression (n = 113) and the remainder was
a mixture of happy (n = 60) and neutral expressions
(n = 60).
Results
Pictures that were rated in the top 10% of the stimulus pool as
flirtatious were retained for consideration.
1
Based on ratings
from Studies 1 and 2, a total of 18 pictures of six posers were
retained as representing a flirtatious facial expression.
Study 3
In Studies 1 and 2, our goal was to narrow down a large
stimulus set into a smaller number of pictures rated as repre-
senting a flirtatious expression. In our previous studies, men
were explicitly asked to rate each picture’s fit with conveying
a flirtatious expression. We designed Study 3 to determine
whether men would label the expressions in the pictures
selected in Studies 1 and 2 as flirtatious without initially being
prompted. An additional goal was to verify that men were not
simply rating the women’s more attractive pictures as indica-
tive of flirtation.
Figure 1. The flirting expression. Figure 1a and 1b depict high recognition flirt faces, rated as flirtatious by 77% and 71% of men, respectively. FACS codes for 1a: 6B
+12 C + 52 C + 54B+61D+63 and 1b: 12B+24A+51B+54 C + 62 C + 63. Figure 1c and 1d depict two low recognition faces (8% and 13% recognition, respectively). FACS
codes for 1 c: 12D+25D+51B+54B+55 C +62 C & 1d: 12A+51B+54B+62B.
1
In addition to determining the components of a recognizable flirtatious expression, we also
wanted to ensure that each poser had a highly rated picture for the “happy” and “neutral”
expressions as well. Due to this requirement, some posers were dropped from consideration
after Study 2.
THE JOURNAL OF SEX RESEARCH 3
Participants, Materials, and Procedure
Twenty-six men were presented with 31 pictures from six
female posers (18 of which were rated as flirtatious in the
previous studies; 13 of which were rated as non-flirtatious:
either happy or neutral expressions). Pictures were randomly
presented in two counter balanced blocks and remained on the
screen until a response was made by the participant. In one
block, participants were presented with the 31 pictures and
were asked to think of one word that described what the
woman in each picture was trying to convey or express. They
responded in an open-ended manner. In the other block,
participants were presented with the same pictures and asked
to rate how attractive the woman in the picture was on
a 7-point Likert scale from 1 (not at all) to 7 (very much).
Results
Three research assistants coded the open-ended responses
to the facial expressions. For each one-word response, its fit
with flirtation was rated using a 5-point scale ranging from
1 (not at all) to 5 (very much) (ICC between raters = .85).
Responses that were coded as flirtatious included the fol-
lowing: “flirting,” “attraction,” “interested,” and “aroused.”
To determine whether the flirtatious responses occurred
above chance level, we followed the method of Tracy and
Robins (2004), in which each response was coded as a 1 if
it was at or above the scale midpoint of 3 on flirting or as
a 0 if it was below the midpoint. A binomial test revealed
that 11 of the 18 previously rated flirtatious pictures were
rated as flirtatious more than would be expected by chance
(ps < .05; chance set at 33%).
An additional goal of Study 3 was to verify that flirtatious
ratings were independent of the women’s perceived attractive-
ness. To test whether men labeled expressions as flirtatious as
a function of physical attractiveness, we conducted six
repeated-measures ANOVAs (i.e., one for each poser) with
facial expression as the within-subjects factor (i.e., flirtatious,
happy, neutral). Ratings of attractiveness did not differ across
photos (all ps > .15), suggesting that flirtation ratings were
based on the facial expressions posed by the women rather
than their perceived attractiveness.
Study 4
In Study 4, we had two goals: (1) to use a different methodology
to verify that the pictures chosen as representing flirtation (rated
by men in the previous studies) are indeed perceived as flirta-
tion; (2) to determine the specific morphology of an expression
recognized as flirtation by men. To obtain these goals, we used
only flirtatious expressions and examined which expressions
were more likely to be rated as flirtatious from a multiple-
choice list of 10 response options. We then used a Facial
Action Coding System (FACS; Ekman & Friesen, 1978) to
capture the unique components of the perceived flirtatious facial
expressions from the previous studies. This allowed us to obtain
the highest flirtatious expressions and provide an explicit analy-
sis of what constitutes a flirtatious expression.
Participants, Materials, and Procedure
Based on the previous three studies, 48 men were presented
with 16 pictures and were asked to select from a list what the
woman in the picture was expressing. The pictures were ran-
domly presented on a computer screen. Each picture remained
on the screen until the participant made a response. The
response options were derived from emotions validated in
previous studies (e.g., Ekman et al., 1969; Keltner, 1995) as
well as the open-ended responses from Study 3 (e.g., embar-
rassed, confusion, excitement).
Nine pictures with an expression previously rated as flirta-
tious were included.
2
To ensure variation in the flirtatious
expressions presented, four of the pictures had been rated as
representing flirting in all three previous studies, while five of
the pictures had been rated as flirting in only one previous
study (thus had lower recognizability as a flirtatious expres-
sion). Having variability in the flirtatious expressions was
expected to allow us to compare the morphology of highly
recognizable flirtatious expressions versus less recognizable
ones. For each picture, participants were asked to rate the
expression by choosing one of the following 10 options: happi-
ness, anger, flirtatious, embarrassed, surprise, sadness, pride,
excitement, confusion, or none.
Results
To determine if any expressions were recognized as flirtation,
we converted the responses into binomials, dummy coding
“flirtatious” responses as 1 and all other responses as 0.
A binomial test, with a conservatively set 50% threshold,
revealed that only two expressions were recognized as flirta-
tion significantly more often than would be expected by
chance, both binomial ps < .01. These two pictures had also
been rated as flirtatious in the previous three studies.
The two pictures recognized as flirtation received the
“flirtatious” selection by 77% (Figure 1a) and 71% (Figure
1b) of men, which is within a typical range for agreement
in recognition of facial expressions (see Ekman et al., 1987).
An expert (third author Rosenberg) coded all faces using
FACS. Another coder, blind to the study and hypotheses,
coded a subset of the pictures as well. Inter-coder agree-
ment using a ratio of the total number of agreements
between coders/total number of agreements plus disagree-
ment was .93 (Wexler, 1972). The FACS coding revealed
a common morphology among the two highly recognized
flirtatious expressions, which consisted of: a head turned to
one side, head tilted down slightly, a slight smile, and eyes
turned forward (toward the implied target). When the
morphology of faces with lower recognition as flirtation
(8% and 13%; Figure 1c and 1d, respectively) was con-
trasted with high recognition faces, the lower recognition
faces showed smiles that were either much more intense,
very subtle, less head tilt, or showed the head tilted up
rather than down.
2
A pretest on recognition was conducted in which two of the 11 pictures from
Study 3 were rated as being neutral and were not included in the final study.
4P. HAJ-MOHAMADI ET AL.
Study 5
Studies 1–4 demonstrated that a flirtatious facial expression has
unique characteristics, rendering it different from happy or
neutral expressions. According to theory (Keltner & Haidt,
1999), flirtatious expressions are supposed to communicate
interest and sexual availability to potential mates. In other
words, viewing a flirtatious facial expression should activate
sex-related association networks. Using the identified expres-
sion and pictures, in Study 5 we tested this proposition by
exposing participants to a flirtatious facial expression, as com-
pared to a happy expression, and subsequently measured the
cognitive accessibility for sex-related words. We used a 2 × 2
factorial design with prime type (flirtatious, happy expression)
and target word (sex, neutral) as the within-subjects factors.
We expected that exposure to flirtatious expressions, compared
to happy expressions, would result in higher cognitive accessi-
bility of sex-related words, as demonstrated by faster reaction
times for sex-related words versus neutral words.
Participants, Materials, and Procedure
Fifty-five male participants (age range 18–28; M age =
19.67 years) completed the study in return for class credit.
Participants were invited individually to the laboratory to
take part in a study about “communication strategies.” After
consenting, participants received instructions for
a computerized lexical decision task (see Schvaneveldt &
Meyer, 1973) that stated:
For the following study, your task is to determine as
quickly and accurately as possible, whether a string of letters
appearing on the screen is a word or not. You will see
a ashing image, then a string of letters. Sometimes the
letters will spell out a word, but other times it will not be
a word. We are interested in how quickly you can respond to
whether it is a word or not.
The lexical decision-making task consisted of 131 trials –
five practice trials followed by 126 experimental trials. Each
trial started with a fixation “X” presented for 500 ms,
followed by the prime images, which were either
a flirtatious facial expression (50% of the trials) or
a happy facial expression (50% of the trials) for 25 ms.
The prime presentation was followed by a mask presented
for 500 ms (a visual “noise” pattern that was used to reduce
the possibility that an afterimage remains on the retina).
Participants were then presented with a target letter string
and were asked to indicate whether the letter string was
a proper English word (by pressing 1) or a non-word (by
pressing 3) as fast as they could. Target letter words
appeared in size 20 Tahoma regular font in black lettering
in the center of the screen on a white background. There
was a total of 63 target words across two categories: sex-
related words (e.g., orgasm) and neutral words (e.g., jour-
nal). All words were selected from the Affective Norms for
English Words (ANEW; Bradley & Lang, 1999). The task
was run on an IBM PC Pentium III computer with a SVGA
color screen using SuperLab Pro 2.0 program (Cedrus
Corporation, San Pedro, CA). Participants worked at their
own pace.
Results
To test our prediction that exposure to a flirtatious facial
expression, compared to a happy expression, leads to higher
accessibility of sex-related words, we conducted a repeated
measures ANOVA with target word (sex, neutral) and prime
type (flirtatious, happy expression) as the within-subjects
factors.
3
A significant main effect of target word emerged,
F (1, 54) = 14.87, p = .0001, η
p2
= .22, in which sex words
were identified faster than neutral words. This effect was qua-
lified by a two-way interaction between prime type and target
word, F (1, 54) = 4.64, p = .036, η
p2
= .08. No main effect of
prime emerged, F (1, 54) = .006, p = .94, η
p2
= .00.
A series of pairwise comparisons between the target
words and prime type were conducted to probe the inter-
action. Findings from post-hoc comparisons revealed that
flirtatious expressions led to significantly faster reaction
times for sex target words (M= 601.50 ms, SD = 110.54)
compared to neutral words (M = 653.14 ms, SD = 151.97),
F (1, 54) = 18.64, p = .0001, η
p2
= .26. Happy expression
primes also led to faster reaction times for sex words (M =
616.11 ms, SD = 120.98) compared to neutral words (M =
639.76 ms, SD = 139.50), F (1, 54) = 4.24, p = .044, η
p2
=
.07. Mean reaction times for sex target words in the flirta-
tious prime condition compared to the happy condition
reflected a nonsignificant marginal trend in the predicted
direction, t (54) = −1.55, p = .128. A difference score of
reaction times between sex and neutral target words was
created for the flirt condition and happy condition to
determine whether the difference in reaction times was
stronger in the flirt versus happy condition. Differences in
reaction times between sex and neutral target words were
significantly larger in the flirt condition (M = −51.64, SD =
88.72) compared to the happy condition (M = −23.65, SD =
85.20), t (54) = −2.15, p = .036, 95% CI [−54.05, −1.93]. See
Figure 2 for a depiction of the mean reaction times across
condition.
Study 5 was designed to test whether priming men with
flirtatious vs. happy facial expressions lead to faster reaction
times for identifying sex-related words. Results supported our
hypothesis that men showed significantly faster reaction times
for sex-related target words (e.g., orgasm) compared to neutral
target words (e.g., journal) after being primed with flirtatious
expressions. Although men also identified sex words faster
than neutral words in the happy prime condition, the differ-
ence in reaction times for identifying sex-related target words
between the flirtatious and the happy expression conditions
approached significance in the predicted direction. The identi-
fication of sex target words, compared to neutral words, was
significantly faster after exposure to flirtatious expressions
compared to happy expressions. These results support our
general prediction that being primed with a flirtatious facial
3
The goal of using the lexical decision task was to determine whether priming flirting
would result in faster responses to sex-related words compared to a neutral,
unrelated word. After data collection was complete, we realized that the lexical-
decision task did not include non-words, as noted in the instructions. We do not
expect this omission to affect the findings reported, given that the objective of this
study focused on semantic accessibility rather than semantic processing.
Nonetheless, Study 6 was conducted as a follow-up study that included non-
words in addition to sex and neutral words to address this limitation.
THE JOURNAL OF SEX RESEARCH 5
expression activates sex-related semantic networks in the
brain, which heightens the accessibility of sex-related words,
and hence leads to faster reaction times for identifying sex
target words.
Study 6
In Study 5, we learned that men had significantly faster reac-
tion times for detecting sex target words compared to neutral
words when primed with a flirting facial expression. In Study 6,
we wanted to further test this hypothesis using a mixed-
subjects design. Specifically, in Study 6, we compared the effect
of flirtatious expressions to happy and neutral expressions on
the accessibility of sex target words. Flirtatious expressions,
compared to happy and neutral expressions, were expected to
result in faster identification of sex-related words compared to
neutral or non-words.
Participants, Materials, and Procedure
A total of 72 male undergraduate students from the University
of Kansas were recruited from the university subject pool.
Similar to Study 5, participants were asked to participate in
a study about “communication strategies.” Participants
received course credit for their participation.
Participants came to the lab and were randomly assigned to
one of three priming conditions: flirting facial expression (n =
24), happy facial expression (n = 25), or a neutral facial expres-
sion (n = 23). Next, participants completed the same compu-
terized lexical decision task in SuperLab (Cedrus Corporation,
San Pedro, CA) as in Study 5. Similar to Study 5, each trial
started with a fixation x presented for 500 ms, followed by the
prime images, which were either a flirtatious facial expression,
happy facial expression, or neutral facial expression, depending
on condition, presented for 25 ms. The prime presentation was
followed by a mask presented for 500 ms. Participants then
indicated whether the letter string was a proper English word
(by pressing 1) or a non-word (by pressing 3) as fast as they
could. Across 80 trials participants saw either sex (e.g., naked),
neutral (e.g., key), or non-words (e.g., vderi). Target words
were selected from the ANEW database (Bradley & Lang,
1999).
Results
We hypothesized that people who were primed with
a flirtatious facial expression, compared to a happy or neutral
facial expression, would exhibit faster reaction times for recog-
nizing sex words compared to neutral or non-words. To test
our prediction, we conducted a 3 × 3 mixed factorial analysis of
variance with prime type (flirting expression, happy expres-
sion, and neutral expression) as the between-subjects factor
and target word (sex, neutral, non-word) as the within-subjects
factor. Results revealed a main effect of word type, F (1.84,
126.69) = 9.286, p = .0001, η
p2
= .12, in which sex words were
identified faster than neutral or non-words; but no main effect
of prime type, F (2, 69) = .340, p = .713, η
p2
= .01. The main
effect of word type was qualified by a nonsignificant trend
between word type and prime, F (3.67, 126.69) = 2.18, p = .08,
η
p2
= .06.
As expected, sex words (M = 698.29 ms, SD = 37.87) were
identified faster as words than neutral words (M = 732.10 ms,
SD = 33.91) when primed with a flirtatious expression, p = .02,
95% CI [−63.11, −4.51]. Reaction times for sex and neutral
words did not differ in the happy expression (p = .11) or
neutral expression prime conditions (p = .74). Men responded
faster to sex words (M = 730.42 ms, SD = 37.10) than non-
words (M = 768.15 ms, SD = 38.63) in the neutral expression
condition, p = .05, 95% CI [−74.91, −.54]. Men also responded
faster to sex words (M = 740.28 ms, SD = 38.68) than non-
words (M = 802.51 ms, SD = 40.28) in the happy expression
condition, p = .002, 95% CI [−101.00, −23.47]. Identification of
sex target words did not differ from neutral words in the happy
or neutral prime conditions, ps > .05. See Figure 3 for
a depiction of the mean reaction times across condition.
General Discussion
Six studies, using a variety of methods, allowed us to identify
two facial expressions consistently recognizable by men as
flirtation. The highly recognized flirtatious expressions exhib-
ited distinct similarities with previously observed flirtatious
expressions, including key elements described by Eibl-
Eibesfeldt (1971). The specific morphology of the highly recog-
nized expression included a head turned to one side, head tilted
down slightly, a slight smile, and eyes turned forward (toward
the implied target). Both chosen pictures involved minor body
turns as well. Notably, only one of the expressions contained
features of a Duchenne’s smile, indicative of felt enjoyment
(Figure 1a; Ekman et al., 1990); however, it is unclear if this
difference is suggestive of two forms or motivations for flirting,
or if it is an artifact of posing. The fact that men consistently
indicated a particular form of expression as flirting suggests
that this nonverbal behavior may be a part of an evolved set of
behaviors designed to facilitate the initiation of relationships.
Figure 2. Study 5 findings for mean reaction times across prime condition. Error
bars represent standard errors.
6P. HAJ-MOHAMADI ET AL.
Observable differences in morphology were found between
high and low recognition flirting expressions. Low recogni-
tion expressions appeared either too ambiguous (Figure 1d)
resulting in participants being unable to label the expression
as conveying anything consistently, or too happy (Figure 1c),
potentially resulting in the expression being interpreted as
purely friendly. Thus, expressions that contain some compo-
nents of the flirting expression, but are too subtle, may be
interpreted as flirting only by men with a very low threshold
for perceiving signs of interest. In contrast, flirting in a way
that demonstrates too much of one of these components (e.g.,
Figure 1c), such as smiling, may end up being interpreted as
happiness, missing the courtship-related cue. Further
research is needed to discriminate among these various
explanations.
Flirtatious facial expressions activated sex-related schemas
as shown by faster identification of sex-related words com-
pared to neutral words (Studies 5 and 6). These findings sug-
gest that flirtatious expressions in women convey interest and
are successfully acknowledged by men. Although women’s
motives for flirting may not always be sex-related (e.g., see
Hall et al., 2010; Henningsen, 2004), men are more inclined
to perceive flirtatious expressions as an indication of sexual
interest (e.g., Henningsen et al., 2008) and believe that flirting
is an invitation to initiate sexual relationships with women
(Koeppel et al., 1993). Sex words were also identified faster
than non-words, which may be due to men’s greater inclina-
tion to interpret women’s behaviors as more sexual in cross-sex
interactions (Abbey, 1982; Henningsen, 2004; Koeppel etal.,
1993; Shotland & Craig, 1988).
Flirtatious facial expressions are subtle and ambiguous; this
allows women to communicate their potential interest while
keeping a buffer of safety. Men were able to recognize this
subtlety in flirtatious expressions as compared to happy or
neutral expressions. Men’s ability to recognize ambiguous
and subtle flirtatious cues is consistent with evolutionary per-
spectives that men strive to avoid missing opportunities for
mating.
Notably, flirtatious expressions, compared to happy or neu-
tral expressions, resulted in trends toward faster identification
of sex-related words. In other words, men detected the subtle
differences between the flirtatious and happy expressions and
responded faster to the flirting females. These findings are in
line with work showing that men are most likely to interpret
flirtation as an invitation, or an opportunity to take action and
initiate a relationship with women (Henningsen et al., 2008;
Koeppel et al., 1993).
Limitations and Future Directions
We do not claim the universal existence of a specific flirting
expression or intend to suggest that other behaviors are not
equally or more important in courtship; rather, we contend
that this facial expression may be a particular, recognizable
component of the flirting process and facilitate mate selec-
tion. To further validate the flirting expression, cross-cultural
evidence is needed. Findings should also be replicated using
members of different age groups and non-university samples
to provide further generalizability. In addition, the stimuli in
the current studies were posed expressions generated by
volunteers. Images of spontaneously generated flirting
expressions in a natural setting are needed and would likely
provide not only external validity, but also further informa-
tion on individual differences in the ability to generate this
expression.
Although we focused here on off-line face-to-face flirting,
flirtation can also be conveyed nonverbally in online settings
(e.g., chat rooms) through the use of flirtatious emoticons and
flattery (Ben-Ze’ev, 2004; Whitty, 2003, 2004) or by controlling
self-presentation on online platforms (e.g., Facebook; Abbasi &
Alghamdi, 2017). Indeed, work suggests that virtual flirtation
elicits greater sexual reactions compared to in-person flirtation
(Alapack et al., 2005). Future research should compare our
identified expression with other formats of flirting. Future
work should also consider the facial morphology of online
dating profiles and whether those who display flirtatious facial
expressions have greater success with securing dates, as mea-
sured by the number of likes or messages received from inter-
ested males.
Another limitation of the current work has to do with the
lack of diversity of our posers. Future research should use
posers of different backgrounds, as the posers here were all
European Americans from a similar SES and background.
Ekman and Keltner’s (1997) seminal paper on the universality
of emotions suggested that facial expressions should not vary
cross-culturally. Nevertheless, future work should examine
whether nuanced expressions (i.e., flirtatious expressions)
Figure 3. Study 6 findings for mean reaction times across prime condition. Error bars represent standard errors.
THE JOURNAL OF SEX RESEARCH 7
vary across posers of different backgrounds in terms of the
distinctiveness and recognizability of the expression
among men.
Another limitation is that we did not measure the sexual
orientation of the judges. The argument outlined here is in line
with heterosexual mating behavior; however, we did not mea-
sure sexual orientation or gender identity. We do not know for
certain that all male judges identified only as male or only as
heterosexual, and we do not know how well these findings
generalize to other types of attraction. Future work should
also look at the use of flirting among people of varied gender
identities and the role of sexual orientation (e.g., in same-sex
men relationships).
Some expressions in our research were rated as flirtatious in
some studies, but not in others. This suggests that there is more
than one way to signal interest in a mate and there are indivi-
dual differences in both men’s recognition of a flirting expres-
sion and women’s expression of a flirting facial cue. Women
who are more comfortable in the dating context, for example,
might be more skilled at displaying an effective flirtatious facial
expression. Alternatively, being better able to communicate
interest via a flirtatious expression could be the result of
experience (e.g., determining the facial expression that receives
the most positive feedback). A third option can be that the
flirting expression could be spontaneously generated by some
women more effectively than others, resulting in some women
having a more difficult time signaling interest to men.
In addition to learning more about individual differences in
effective signaling among women, additional research could
also provide a greater understanding of the differences in
interpreting flirtatious expressions in men. Perhaps men with
a higher preference for a short-term mating strategy or those
with higher levels of sociosexuality (i.e., permissive sexual
attitudes) have a lower threshold for recognizing a flirting
expression, allowing them to feel greater confidence in pursu-
ing more mates. Recognition of flirtatious expressions may also
vary based on relationship status (i.e., single, dating, married).
Motivations for flirting have been studied in married couples
(Frisby, 2009; Frisby & Booth-Butterfield, 2012), though the
success of recognizing a flirtatious expression is unknown.
Relationship status was not assessed in the present work, there-
fore there may have been group differences between single and
non-single partners that we are unaware of. Future research,
taking into account the FACS differences between high and low
recognition flirting expressions, could provide insight into the
reasons for discrepant success rates among people interested in
initiating relationships.
The current set of studies focused on flirtatious expressions
signaled by women (presumably) toward men. Based on sexual
selection theory, we would expect that women and men would
have different strategies and goals in courtship. Thus, we would
not expect the same expression to be mimicked in men. Future
research should explore a facial expression among men that is
displayed during courtship and recognized by women as
a flirting cue. While the expression among women has compo-
nents of submissiveness (e.g., head tilted down), a man’s flirta-
tious expression might have more dominant components and
a goal of generating maximum interest and attraction rather
than of maximizing selection.
In conclusion, our results show that men consistently rated
a small subset of pictures as flirtatious. There were other
pictures we used that were rated as flirtatious by some but
not by others, suggesting that the universality of the expression
is highly specific. The FACS based similarity among the highly
rated flirting expressions and their differences from the other
expressions suggest there is indeed a specific morphology
involved in the effective signaling and recognition of female
interest. Exposure to flirtatious expressions, compared to
happy or neutral expressions, resulted in higher accessibility
of sex-related words compared to neutral words in men.
Although further research is needed, our studies serve as an
important first step in identifying the existence and morphol-
ogy of flirtatious cues and their effect on men.
ORCID
Parnia Haj-Mohamadi http://orcid.org/0000-0002-3133-7648
References
Abbasi, I. S., & Alghamdi, N. G. (2017). When flirting turns into infidelity:
The Facebook dilemma. The American Journal of Family Therapy, 45
(1), 1–14. https://doi.org/10.1080/01926187.2016.1277804
Abbey, A. (1982). Sex differences in attributions for friendly behavior: Do
males misperceive females‘ friendliness? Journal of Personality and
Social Psychology, 42(5), 830–838. https://doi.org/10.1037/0022-3514.
42.5.830
Alapack, R., Blichfeldt, M. F., & Elden, A. (2005). Flirting on the internet
and the hickey: A hermeneutic. CyberPsychology & Behavior, 8(1),
52–61. https://doi.org/10.1089/cpb.2005.8.52
Ben-Ze’ev, A. (2004). Flirting on and offline. Convergence: The
International Journal of Research into New Media Technologies, 10(1),
24–42. https://doi.org/10.1177/135485650401000103
Berscheid, E., Dion, K., Walster, E., & Walster, G. W. (1971). Physical
attractiveness and dating choice: A test of the matching hypothesis.
Journal of Experimental Social Psychology, 7(2), 173–189. https://doi.
org/10.1016/0022-1031(71)90065-5
Bradley, M. M., & Lang, P. J. (1999). Aective norms for English words
(ANEW): Instruction manual and aective ratings (Technical report
C-1). The Center for Research in Psychophysiology, University of
Florida.
Buss, D. M. (1989). Sex differences in human mate preferences:
Evolutionary hypotheses tested in 37 cultures. Behavioral and Brain
Sciences, 12(1), 1–49. https://doi.org/10.1017/S0140525X00023992
Dunbar, R. I. (2001). What’s in a baboon’s behind? Nature, 410(6825),
158. https://doi.org/10.1038/35065773
Eibl-Eibesfeldt, I. (1971). Love and hate. Holt Rinehart, & Winston.
Ekman, P. (1992). An argument for basic emotions. Cognition & Emotion,
6(3–4), 169–200. https://doi.org/10.1080/02699939208411068
Ekman, P., Davidson, R. J., & Friesen, W. V. (1990). The Duchenne smile:
Emotional expression and brain physiology: II. Journal of Personality
and Social Psychology, 58(2), 342–353. https://doi.org/10.1037/0022-
3514.58.2.342
Ekman, P., & Friesen, W. V. (1978). Facial action coding system:
A technique for the measurement of facial movement. Consulting
Psychologists Press.
Ekman, P., Friesen, W. V., O’Sullivan, M., Chan, A., Diacoyanni-
Tarlatzis, I., Heider, K., Krause, R., LeCompte, W. A., Pitcairn, T.,
Ricci-Bitti, P. E., Scherer, K., Tomita, M., & Tzavaras, A. (1987).
Universals and cultural differences in the judgments of facial expres-
sions of emotion. Journal of Personality and Social Psychology, 53(4),
712–717. https://doi.org/10.1037/0022-3514.53.4.712
Ekman, P., & Keltner, D. (1997). Universal facial expressions of emotion:
An old controversy and new findings. In U. Segerstrale & P. Molnar
8P. HAJ-MOHAMADI ET AL.
(Eds.), Nonverbal communication: Where nature meets culture (pp.
27–46). Lawrence Erlbaum Associates, Inc.
Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomic nervous
system activity distinguishes among emotions. Science, 221(4616),
1208–1210. https://doi.org/10.1126/science.6612338
Ekman, P., Sorenson, E. R., & Friesen, W. V. (1969). Pan-cultural elements
in facial displays of emotion. Science, 164(3875), 86–88. https://doi.org/
10.1126/science.164.3875.86
Frisby, B. N. (2009). “Without flirting, it wouldn’t be a marriage”:
Flirtatious communication between relational partners. Qualitative
Research Reports in Communication, 10(1), 55–60. https://doi.org/10.
1080/17459430902839066
Frisby, B. N., & Booth-Butterfield, M. (2012). The “how” and “why” of
flirtatious communication between marital partners. Communication
Quarterly, 60(4), 465–480. https://doi.org/10.1080/01463373.2012.704568
Givens, D. B. (1978). The nonverbal basis of attraction: flirtation, court-
ship, and seduction. Psychiatry: Interpersonal and Biological Processes,
41(4), 346–351. https://doi.org/10.1080/00332747.1978.11023994
Gottman, J., Levenson, R., & Woodin, E. (2001). Facial expressions during
marital conflict. Journal of Family Communication, 1(1), 37–57. https://
doi.org/10.1207/S15327698JFC0101_06
Hall, J. A., Carter, S., Cody, M. J., & Albright, J. M. (2010). Individual
differences in the communication of romantic interest: Development of
the Flirting Styles Inventory. Communication Quarterly, 58(4),
365–393. https://doi.org/10.1080/01463373.2010.524874
Hall, J. A., Xing, C., & Brooks, S. (2015). Accurately detecting flirting:
Error management theory, the traditional sexual script, and flirting
base rate. Communication Research, 42(7), 939–958. https://doi.org/
10.1177/0093650214534972
Haselton, M. G., & Buss, D. M. (2000). Error management theory: A new
perspective on biases in cross-sex mind reading. Journal of Personality
and Social Psychology, 78(1), 81–91. https://doi.org/10.1037/0022-3514.
78.1.81
Henningsen, D. D. (2004). Flirting with meaning: An examination of
miscommunication in flirting interactions. Sex Roles, 50(7/8),
481–489. https://doi.org/10.1023/B:SERS.0000023068.49352.4b
Henningsen, D. D., Braz, M., & Davies, E. (2008). Why do we flirt? Flirting
motivations and sex differences in working and social contexts. The
Journal of Business Communication (1973), 45(4), 483–502. https://doi.
org/10.1177/0021943608319390
Keltner, D. (1995). Signs of appeasement: Evidence for the distinct dis-
plays of embarrassment, amusement, and shame. Journal of Personality
and Social Psychology, 68(3), 441–454. https://doi.org/10.1037/0022-
3514.68.3.441
Keltner, D. (2003). Expression and the course of life: Studies of emotion,
personality, and psychopathology from a social functional perspective.
Annals of the New York Academy of Sciences, 1000(1), 222–243. https://
doi.org/10.1196/annals.1280.011
Keltner, D., & Haidt, J. (1999). Social functions of emotions at four levels
of analysis. Cognition & Emotion, 13(5), 505–521. https://doi.org/10.
1080/026999399379168
Koeppel, L. B., Montagne-Miller, Y., O’Hair, D., & Cody, M. J. (1993).
Friendly? Flirting? Wrong. In P. J. Kalbfleisch (Ed.), Interpersonal
communication: Evolving interpersonal relationships (pp. 13–32).
Lawrence Erlbaum.
Moore, M. M. (2010). Human nonverbal courtship behavior—A brief
historical review. Journal of Sex Research, 47(2–3), 171–180. https://
doi.org/10.1080/00224490903402520
Muehlenhard, C. L., Koralewski, M. A., Andrews, S. L., & Burdick, C. A.
(1986). Verbal and nonverbal cues that convey interest in dating: Two
studies. Behavior Therapy, 17(4), 404–419. https://doi.org/10.1016/
S0005-7894(86)80071-5
Petrie, M., Tim, H., & Carolyn, S. (1991). Peahens prefer peacocks with
elaborate trains. Animal Behaviour, 41(2), 323–331. https://doi.org/10.
1016/S0003-3472(05)80484-1
Renninger, L. A., Wade, T. J., & Grammer, K. (2004). Getting that
female glance: Patterns and consequences of male nonverbal
behavior in courtship contexts. Evolution and Human Behavior,
25(6), 416–431. https://doi.org/10.1016/j.evolhumbehav.2004.08.
006
Schvaneveldt, R. W., & Meyer, D. E. (1973). Retrieval and comparison
processes in semantic memory. In S. Kornblum (Ed.), Attention and
Performance IV (pp. 395–409). New York: Academic Press.
Shotland, R. L., & Craig, J. M. (1988). Can men and women differentiate
between friendly and sexually interested behavior? Social Psychology
Quarterly, 51(1), 66–73. https://doi.org/10.2307/2786985
Simpson, J. A., Gangestad, S. W., & Biek, M. (1993). Personality and
nonverbal social behavior: An ethological perspective of relationship
initiation. Journal of Experimental Social Psychology, 29(5), 434–461.
https://doi.org/10.1006/jesp.1993.1020
Speer, S. A. (2017). Flirting: adesignedly ambiguous action? Research on
Language and Social Interaction, 50(2), 128–150. https://doi.org/10.
1080/08351813.2017.1301297
Tisdale, L., & Sheldon, P. (2018). Female flirting cues and male perception.
American Communication Journal, 19(2), 1–11.
Tracy, J. L., & Robins, R. W. (2004). Show your pride: Evidence for
a discrete emotion expression. Psychological Science, 15(3), 194–197.
https://doi.org/10.1111/j.0956-7976.2004.01503008.x
Trivers, R. L. (1972). Parental investment and sexual selection. In
B. Campbell (Ed.), Sexual selection and the descent of man (pp.
136–179). Aldine.
Wexler, D. (1972). Method for unitizing protocols of descriptions of
emotional states. Journal of Supplemental Abstracts Service, Catalogue
of Selected Documents in Psychology, Vol. 2, p. 116. Washington, DC:
American Psychological Association.
White, J., Lorenz, H., Perilloux, C., & Lee, A. (2018). Creative Casanovas:
Mating strategy predicts using—but not preferring—atypical flirting
tactics. Evolutionary Psychological Science, 4(4), 443–455. https://doi.
org/10.1007/s40806-018-0155-7
Whitty, M. T. (2003). Cyber-flirting: Playing at love on the Internet.
Theory & Psychology, 13(3), 339–357. https://doi.org/10.1177/
0959354303013003003
Whitty, M. T. (2004). Cyber-flirting: An examination of men’s and
women’s flirting behaviour both offline and on the internet.
Behaviour Change, 21(2), 115–126. https://doi.org/10.1375/bech.21.2.
115.55423
THE JOURNAL OF SEX RESEARCH 9
