ArticlePDF Available

Dimorphous Expressions of Positive Emotion: Displays of Both Care and Aggression in Response to Cute Stimuli

Authors:

Abstract and Figures

Extremely positive experiences, and positive appraisals thereof, produce intense positive emotions that often generate both positive expressions (e.g., smiles) and expressions normatively reserved for negative emotions (e.g., tears). We developed a definition of these dimorphous expressions and tested the proposal that their function is to regulate emotions. We showed that individuals who express emotions in this dimorphous manner do so as a general response across a variety of emotionally provoking situations, which suggests that these expressions are responses to intense positive emotion rather than unique to one particular situation. We used cute stimuli (an elicitor of positive emotion) to demonstrate both the existence of these dimorphous expressions and to provide preliminary evidence of their function as regulators of emotion.
Content may be subject to copyright.
Psychological Science
1 –15
© The Author(s) 2015
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0956797614561044
pss.sagepub.com
Research Article
Some people cry at graduations, at the birth of their chil-
dren, when a hero returns from war, when they reach
their goals, and when someone gives to another person
unselfishly. Some concertgoers scream as if in horror in
the presence of their teen idol, and some people playfully
growl and express their desire to pinch a baby’s cheeks.
What these diverse situations have in common is that
these positive experiences have elicited dimorphous
expressions—not only positive expressions, but also
expressions normatively associated with negative emo-
tions (e.g., anger, sadness, and fear; Ekman & Friesen,
1971, 1986). During these dimorphous displays, both pos-
itive and negative expressions occur simultaneously in a
disorganized manner, which leaves witnesses to rely on
the context of the situation to interpret them (Carroll &
Russell, 1996; Zaki, Hennigan, Weber, & Ochsner, 2010).
The Structure of the Dimorphous
Expression of Emotion
Highlighting points of consensus among emotion
researchers, Gross, John, and Richards (2000) proposed a
process model of emotion that begins with a stimulus
event, followed by an appraisal of the event, an emo-
tional experience, and then an expressive behavior.
Dimorphous expressions of emotion resemble this pro-
cess model but feature a distinct pattern1 of one stimulus
event, one appraisal, one emotional experience, and two
expressive behaviors.
To give an example, a person who has won $100 mil-
lion in a lottery and appraised this event as an incredibly
good thing might feel overwhelmed with happiness and
express this feeling by both smiling and crying. Crying,
which normatively expresses sadness, would seem to
contradict the situation, the appraisal, and the positive
emotions. The negative expression might be merely a
facial display, or it might reflect the onset of an actual
negative emotion.2 At this point, we are not making a
distinction on this matter, but we would note that one’s
561044PSSXXX10.1177/0956797614561044Aragón et al.Dimorphous Expressions
research-article2015
Corresponding Author:
Oriana R. Aragón, Department of Psychology, Yale University,
2Hillhouse Ave., New Haven, CT 06520
E-mail: oriana.aragon@yale.edu
Dimorphous Expressions of Positive
Emotion: Displays of Both Care and
Aggression in Response to Cute Stimuli
Oriana R. Aragón, Margaret S. Clark, Rebecca L. Dyer, and
John A. Bargh
Yale University
Abstract
Extremely positive experiences, and positive appraisals thereof, produce intense positive emotions that often generate
both positive expressions (e.g., smiles) and expressions normatively reserved for negative emotions (e.g., tears). We
developed a definition of these dimorphous expressions and tested the proposal that their function is to regulate
emotions. We showed that individuals who express emotions in this dimorphous manner do so as a general response
across a variety of emotionally provoking situations, which suggests that these expressions are responses to intense
positive emotion rather than unique to one particular situation. We used cute stimuli (an elicitor of positive emotion) to
demonstrate both the existence of these dimorphous expressions and to provide preliminary evidence of their function
as regulators of emotion.
Keywords
emotion, emotion expression, emotion regulation, cute
Received 8/14/14; Revision accepted 11/1/14
Psychological Science OnlineFirst, published on January 27, 2015 as doi:10.1177/0956797614561044
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
2 Aragón et al.
expression does not necessarily correspond with one’s
emotional experience (Gross et al., 2000; Kappas, 2003).
The Function of Dimorphous
Expressions
We presume that dimorphous expressions of emotion
occur during situations in which people feel over-
whelmed with emotion, when they perceive that a point
has been reached at which their emotions have become
unmanageable. These perceptions of feeling over-
whelmed (Folkman & Moskowitz, 2000) may be dictated
by physiological limits and may deter people from sus-
taining high levels of emotion that can be deleterious for
the body (e.g., Colom et al., 2000).
Dimorphous expressions of emotion may help regu-
late emotions (see Gross, 2013, for a review), possibly
through balancing one emotion with the expression of
another. If the expression of one emotion regulates
another emotion (Fredrickson, Mancuso, Branigan, &
Tugade, 2000; Kappas, 2011; Samson & Gross, 2012;
Schimmack, 2001), one might expect to see negative
emotion expression when positive emotions run too high
and see positive emotion expression when negative emo-
tions run high. In fact, Fredrickson and Levenson (1998)
reported such a dimorphic response to negative emotion,
when over half of their participants spontaneously smiled
while watching the most intense moments of a sad movie
scene. Those who displayed their sadness in this dimor-
phic manner reported feeling sad but had faster cardio-
vascular recovery from the sad event than those who did
not smile.
Preliminary Studies
We initially tested the proposed dimorphous expression
of emotion with stimuli that are considered “cute” because
the mere presentations of photographs of infants pro-
duce strong positive emotional responses and activate
the reward system in the brain (Glocker, Langleben,
Ruparel, Loughead, Valdez, et al., 2009). When people
see characteristics such as large, wide-set eyes; round
cheeks; and small chins (known as baby schema; Lorenz,
1943), they get the impulse to approach and provide pro-
tection and care (Glocker, Langleben, Ruparel, Loughead,
Gur, & Sachser, 2009; Lorenz, 1971; Sherman, Haidt, &
Coan, 2009; Sherman, Haidt, Iyer, & Coan, 2013).
Yet our observations yielded distinctly different
responses to cute stimuli, including playful growling,
squeezing, biting, and pinching. Considering the context
in which these aggressive expressions occur, we assume
that these responses are not generated from negative
appraisals of cute beings, the intent to harm, or true
aggression (Anderson & Bushman, 2002). Here, we
explored such expressions as apt and testable illustrations
of dimorphous expressions of emotion and the function
of dimorphous expressions as emotion regulators.
We first wanted to establish that cute stimuli are elici-
tors of dimorphous expressions. To do this, we con-
ducted a preliminary study in which participants (N =
105; 57 female, 48 male; mean age = 36.10 years) reported
whether, within the explicit boundary of not wanting to
harm cute beings, they had ever pinched (30%) and
squeezed (52%) a cute baby or child. Although there is
no word in English to describe these behaviors, we con-
ducted a survey that identified such words in other lan-
guages (e.g., in Filipino, the word gigil refers to the
gritting of teeth and the urge to pinch or squeeze some-
thing that is unbearably cute; Rubino & Llenado, 2002).
(See Sections S2 and S3 in the Supplemental Material
available online.) To test the generality of these responses
as dimorphous expressions of emotions, not specific
expressions in response to “cuteness,” and to test the
model of one stimulus event, one appraisal, one emo-
tional experience, and two expressive behaviors, we con-
ducted Study 1.
Study 1: Displays of Both Care and
Aggression in Response to Cute
Stimuli
We reasoned that dimorphous expressions could be a
form of emotion regulation, because they appear to occur
when people feel overwhelmed with intense feelings. We
further hypothesized that there could be stable individual
differences in people’s tendencies toward dimorphous
displays across a variety of situations that produce intense
emotions, because other emotion-regulation mechanisms
generalize in this way. For instance, an individual might
evoke cognitive reappraisal (Gross & John, 2003) both to
cope with anger from being mistreated and to cope with
sadness from a loss.
We further predicted that our data would fit the model
of the proposed emotional cascade (Gross et al., 2000), in
which babies with higher infantile characteristics would
induce higher positive appraisals; arouse higher reports of
being overwhelmed with positive emotions; provoke
higher expressions of care, as other researchers have dem-
onstrated (e.g., Sherman et al., 2013); and also provoke
higher expressions of aggression, as we hypothesized.
Method
Test of generality of dimorphous expressions.
Participants. Participants were recruited online through
Amazon’s Mechanical Turk (N = 143; 69 female, 74 male;
mean age = 34.22 years, range = 19–73, SD = 12.31) and
were compensated 25¢ to complete a survey that was
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 3
advertised as follows: “Short Survey (5–10 minutes to
complete). In this survey we ask you to answer a short
questionnaire.” Of the 158 participants who logged in to
the survey, 15 (9%) abandoned the survey before com-
pleting it, which left data from 143 participants in our
final analyses.
Materials, procedure, and analysis. We created 30
items that described dimorphous expressions across
a variety of situations. Thirteen items asked about the
dimorphous expression of various positive emotions
across situations that did not include responses to cute
stimuli (e.g., crying at the happiest moment of a movie).
The intercorrelation among these items was high (Cron-
bach’s α = .92). Another 13 items (α = .90) asked about
the dimorphous expression of negative emotions within
various situations,3 and 4 items asked about situations
involving cute stimuli and specific dimorphous expres-
sions of aggression (α = .89). We provided a forced-
choice response scale with no neutral point, as a neutral
point would not be interpretable when asking whether
someone does or does not behave in a certain manner
(1= strongly disagree, 2 = disagree, 3 = somewhat dis-
agree, 4 = somewhat agree, 5 = agree, 6 = strongly agree).
After participants provided informed consent, an intro-
duction to the survey explained that we were interested
in how people express emotion. Participants then
answered the questionnaire, provided basic demograph-
ics, and were thanked for their time. Participants were
allowed to only move forward through the survey; no
back button was provided.
Even though we were not creating a scale per se, we
thought it would be helpful to use a factor analysis to
determine how items related to one another. A simple
correlations array of all the items would not account for
all intercorrelations simultaneously. Therefore, all items
were subjected to an exploratory factor analysis. Our one
a priori prediction was that items involving cute stimuli
would load on the same factor as various situations that
elicit other dimorphous expressions of positive emotions.
We had no a priori prediction for the items focused on
the dimorphous expression of negative emotions. Two
factors emerged, explaining 50% of the variance of the
scale. As predicted, items concerning responses to cute-
ness clearly loaded on the same factor as other dimor-
phous expressions of positive emotions (see Table 1).
Our results indicated that the dimorphous expression of
positive and negative emotions does cross situations and
different emotions. This factor analysis differentiated the
latent constructs of intense positive and negative emo-
tions. As predicted, the dimorphous expression of positive
emotion in response to cute stimuli loaded on the same
factor as other situations that evoked dimorphous
responses of positive emotions, but not dimorphous
responses of negative emotions. This underscores the idea
that it is not a negative emotional response to cute stimuli
that is being expressed along with positive expressions,
but rather a positive emotional response to cute stimuli
that is being expressed with negative expressions.
The dimorphous expression of emotions in response
to cute stimuli was strongly correlated with the dimor-
phous expression of positive emotion in response to
other types of stimuli (r = .79, p < .001), but the dimor-
phous expressions of negative emotions in response to
other types of stimuli showed a lower correlation (r = .21,
p = .01; z-scored difference between correlations = 7.18,
p < .001). The dimorphous expression of positive emo-
tion was correlated with the dimorphous expression of
negative emotion as well (r = .38, p< .001).
Test of the dimorphous-expression model.
Participants. Participants were recruited online through
Amazon’s Mechanical Turk (N = 299; 127 female, 172
male; mean age = 29.79 years, range = 18–63, SD = 9.70)
and were compensated 35¢ for answering a survey that
took approximately 5 min. The study was advertised as
follows: “Short Survey (5–15 minutes to complete). In this
survey we ask you to answer some questions about pho-
tos, and answer a short questionnaire.
Using G*Power software (Faul, Erdfelder, Lang, &
Buchner, 2007) and the effect size from the pretest (see
the Stimuli section below) for feelings of being over-
whelmed (as we thought these feelings were essential to
expressions of aggression), we estimated (.80 power,
alpha error probability = .05, two-tailed) that we would
need approximately 300 participants. Therefore, we set a
stopping point for data collection when 300 surveys were
logged in as complete. Of the 314 participants who
logged in to the survey, 14 (5%) abandoned the survey
before completing the dependent variables, and 1 simply
clicked through the survey without responding. Of the
participants who did not complete the survey, 9 were
originally assigned to the more-infantile condition and 6
were assigned to the less-infantile condition, which left
data from 299 participants in the final analysis.
Materials. For stimuli, we used photographs of babies
that had been prepared and validated by Sherman and
colleagues (2013). Eight photographs of infants and
toddlers (two female, six male) were morphed so that
babies had more-infantile characteristics (larger eyes,
cheeks, and forehead; smaller noses, lips, and chins) and
less-infantile characteristics (the reverse); these morphs
were based on extensive prior research documenting
characteristics of cuteness (e.g., Alley, 1981; Glocker,
Langleben, Ruparel, Loughead, Gur, & Sachser, 2009; Hil-
debrandt & Fitzgerald, 1979). We pretested these pho-
tographs utilizing an independent online sample from
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
4 Aragón et al.
Amazon’s Mechanical Turk (N = 212; 103 female, 109
male; mean age = 33.08 years, age range 18–70, SD =
10.95). Participants were assigned randomly to view the
more-infantile or the less-infantile photographs and to
endorse statements designed to capture an overall posi-
tive appraisal of the stimuli and how those stimuli made
the participant feel. We captured a positive appraisal of
each baby with the items “I think that baby is cute” and
“That baby is good” (α = .87). The statement, “When I
look at this baby I feel overwhelmed with very strong
positive feelings” captured the overwhelming positive
emotional response toward the baby. Sliding scales were
provided for response (1–20 = not at all true, 21–40 = a
little bit true, 41–60 = true, 61–80 = very true, 81–100 =
completely true).
As expected, the photographs presented in the more-
infantile condition (M = 66.77, SD = 20.60) were appraised
more positively than the photographs presented in the
Table 1. Factor Loadings for Each Item on the Test of Generality of Dimorphous Expressions
Item Factor 1 Factor 2
Proposed dimorphous expression of positive emotions
I can be so happy to see someone that I cry. .74
I can get so excited when something great happens that I scream. .51
I can imagine myself crying (or I have cried) at the birth of my children. .66
I cry at weddings when the vows are exchanged. .79
I do not cry when I am overwhelmed with happiness. (reverse-coded) .79
I am not the type of person who would scream (as if in horror) if I came close to my favorite musician at
their concert. (reverse-coded)
.50 —
I cry when I see a stranger give unselfishly to another. .70 .31
I cry when I see loved ones reunite. .80
I can cry when I achieve something that I worked long and hard to get. .76 .22
When I am feeling strong positive emotions, I do not express them with negative expressions. (reverse-
coded)
.65 .28
I cry while watching the happiest moments of movies. .75
I laugh so hard that I cry when I think that something is hysterically funny. .53
When I am feeling a strong positive emotion (e.g., extreme happiness, strong sense of relief, strong feeling
of connection to others), my expression can look like I am feeling a negative emotion (e.g., I might cry,
or scream as though in fear even though I am happy or excited).
.71 —
Proposed dimorphous expression of negative emotions
I can be so angry that I laugh. .74
I never laugh when I am frustrated with a situation. (reverse-coded) .56
If I am anxious enough I will actually smile. .66
I never get so sad that I laugh (laughter through tears). (reverse-coded) .33 .66
I never smile when I am devastated about a bad thing that happened. (reverse-coded) .72
I can be so nervous that I chuckle. .63
I can laugh when I am in a situation that seems utterly hopeless. .72
A situation can be so sad that I find myself laughing. .76
I never smile when I am angry. (reverse-coded) .62
If I am very sad, I might raise the corners of my mouth like a smile, even though there is nothing to smile
about.
— .74
I sometimes smile while watching the saddest moments of movies. .21 .46
When I am feeling a strong negative emotion, I display positive expressions. .73
When I am feeling a strong negative emotion (e.g., deep sadness, strong anxiety, strong anger), my
expression can look like I am feeling a positive emotion (e.g., I might smile or chuckle even though I
am sad, anxious, or angry).
— .82
Proposed dimorphous expressions in response to cute stimuli
If I am holding an extremely cute baby, I have the urge to squeeze his or her little fat legs. .67
If I look at an extremely cute baby, I want to pinch those cheeks. .75
When I see something I think is so cute, I clench my hands into fists. .73
I am the type of person that will tell a cute child “I could just eat you up!” through gritted teeth. .78
Note: Eigenvalues below .20 are indicated by a dash. Factor 1 is proposed to be the dimorphous expression of positive emotion. Factor 2 is
proposed to be the dimorphous expression of negative emotion.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 5
less-infantile condition (M = 59.84, SD = 24.60), t(210) =
2.23, p = .03, d = 0.31. Also as expected, the photographs
presented in the more-infantile condition (M = 50.04,
SD = 27.03) provoked higher reports of being over-
whelmed with positive feelings than the photographs
presented in the less-infantile condition (M = 41.42, SD =
28.76), t(210) = 2.25, p = .03, d = 0.31.
We captured in-the-moment responses with prompts
situated below each photograph, which allowed partici-
pants to respond with the stimulus remaining in view.
Slider bars for all responses had values between 1 and
100 (1–20 = not at all true, 21–40 = a little bit true, 41–
60 = true, 61–80 = very true, 81–100 = completely true).
For the trials that measured appraisal of the babies, we
used the same items as in the validation of the stimuli.
On each trial, a statement appeared below each photo-
graph that read either “When I look at this baby, I feel
like this baby is cute” or “When I look at this baby, I feel
like this baby is good.” These two items (α = .87) were
each averaged across the eight trials.
For the trials that measured emotional experience, we
also used the same items as in the validation of the stim-
uli. Situated below each photograph was the statement,
“When I look at this baby, I feel like I am overwhelmed
by very strong positive feelings.” For trials that measured
care expressions, one of three statements appeared
below each photograph. The statements always began
with the phrase “When I look at this baby, I feel like,” but
they ended with “I want to take care of it!” “I want to hold
it!” or “I want to protect it!” These three items (α = .95)
were each averaged across the eight trials.
For each trial that measured aggressive expressions,
the phrase “When I look at this baby, I feel like” was
completed by one of three phrases: “pinching those
cheeks!” “saying ‘I want to eat you up!’ through gritted
teeth,” or “being playfully aggressive!” Participants were
asked to respond to all three items (α = .93), which were
averaged across the eight trials. The term “playful aggres-
sion” was described to participants in the following way:
We also ask about “playful aggression.” Playful
aggression is in reference to the expressions that
people show sometimes when interacting with
babies. Sometimes we say things and appear to be
more angry than happy, even though we are happy.
For example some people grit their teeth, clench
their hands, pinch cheeks, or say things like “I want
to eat you up!” It would be difficult to ask about
every possible behavior of playful aggression, so
we ask generally about things of this kind—calling
them playful aggressions.
Our demographics questionnaire asked for age, eth-
nicity, the participants’ number of children, desire for
children (or for more children if participants had any
already), and whether participants who did not have chil-
dren regularly spent time with children.
Procedure. After providing informed consent, par-
ticipants read a short introduction to the survey that
informed them that they would be asked to respond to
some photographs. Participants were allowed to move
only forward through the survey; no back button was
provided. We told participants that we wished to mea-
sure positive experiences with photographs of babies,
not negative experiences such as doing actual harm to or
disliking babies. We further instructed them that if they
did not experience such feelings as desiring to pinch a
baby’s cheeks within these boundaries that they could
indicate that by choosing “not at all true” ratings. Tri-
als measuring expressions of aggression, expressions of
care, and appraisals of the stimuli were presented sepa-
rately and counterbalanced randomly. Directly following
this, we collected demographic information.
Results
We found that more-infantile babies (M = 66.88, SD =
18.10) were appraised more positively than less-infantile
babies (M = 56.68, SD = 21.28); an independent-samples
t test revealed that this difference was significant, t(297)=
4.47, p < .001, d = 0.52. As expected, the photographs
presented in the more-infantile condition (M = 52.48,
SD = 23.85) provoked higher reports of being over-
whelmed with very strong positive feelings than the
photographs presented in the less-infantile condition
(M= 42.74, SD = 23.81), t(297) = 3.54, p < .001, d = 0.41.
Participants reported higher expressions of care for
more-infantile babies (M = 55.81, SD = 27.07) than for
less-infantile babies, (M = 47.47, SD = 27.30), t(297) =
2.65, p < .01, d = 0.31. Participants also reported higher
expressions of aggression for more-infantile babies (M =
39.63, SD = 23.69) than for less-infantile babies (M =
33.35, SD = 21.68), t(297) = 2.39, p = .02, d = 0.28.
Next, we tested whether expressions of care in
response to more- (vs. less-) infantile characteristics were
mediated in serial fashion (progressing through one stage
to the next). Serial mediation allows tests of mediation
pathways with more than one mediator working sequen-
tially rather than working in parallel. This analysis allowed
us to test the entire hypothesized emotion-process frame-
work (stimuli appraisal emotional experience
emotional expression; see Gross et al., 2000) in a single
model.
In a bootstrapped serial mediation model with 5,000
samples using Process Model 6 (Hayes, 2013), we pre-
dicted the influence of infant-characteristic conditions on
expressions of care, as mediated by positive appraisals
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
6 Aragón et al.
and being overwhelmed with very strong positive feel-
ings, while controlling for expressions of aggression.
There was a significant indirect path (95% confidence
interval, or CI = [0.03, 0.13]) from viewing more-infantile
babies (vs. less-infantile babies) through the participants’
positive appraisals of such babies (b = 0.36,4 SE = 0.10),
t= 3.74, p < .001, next through the evoked overwhelming
positive emotion (b = 0.55, SE = 0.04), t = 13.49, p < .01,
to the care expressions made toward those babies (b =
0.35, SE = 0.06), t = 4.64, p < .001. Also, as expected, the
manipulation of infancy no longer predicted care
responses with positive appraisals and feelings of being
overwhelmed with positive emotions in the model (c
path; b = −0.05, SE = 0.07), t = −0.65, p = .52. This analysis
tested all possible pathway combinations with the pro-
posed mediators. There was another significant indirect
pathway from infantile characteristics through positive
appraisals predicting care responses (without over-
whelming positive emotions; 95% CI = [0.07, 0.28]). This
suggests that care responses can be mediated by both
positive appraisals and feelings of being overwhelmed
with very strong positive feelings toward the baby, but
also that the feeling of being overwhelmed is not essen-
tial to an outcome of care responses, whereas having a
positive appraisal of the baby is.
We ran an analogous serial mediation model with the
same structure and factors as the previous one, except
that it predicted aggressive expressions by infant-charac-
teristic condition while controlling for expressions of
care. There was a significant indirect path (95% CI =
[0.02, 0.10]) from viewing more-infantile babies (vs. less-
infantile babies) through the participants’ positive
appraisals of such babies (b = 0.30, SE = 0.08), t = 3.56,
p < .001, next through being overwhelmed with very
strong positive feelings (b = 0.44, SE = 0.05), t = 9.55, p <
.001, to the aggressive expressions made in reaction to
those babies (b = 0.39, SE = 0.07), t = 5.49, p < .001.
Further, the manipulation of infancy no longer predicted
aggressive expressions with positive appraisals and feel-
ings of being overwhelmed with positive emotions in the
model (c path; b = 0.03, SE = 0.02), t = −0.31, p = .75.
Again this analysis tested all possible pathway combina-
tions with the proposed mediators. True to our hypoth-
esis that the function of these expressions is to regulate
emotions, there were no other significant pathways,
which indicated that it was solely through the experience
of being overwhelmed by very strong positive feelings
that aggression was expressed. One might wonder if we
were actually able to capture feelings of being over-
whelmed. We remind the reader that we directly asked if
participants were overwhelmed with strong positive feel-
ings, and they responded that they were.
Expressions of care did not require being over-
whelmed by very strong positive feelings for the indirect
pathway to be significant, and expressions of care most
likely have a function of caring for the baby, which leads
to the baby’s well-being. Unlike expressions of care,
expressions of aggression were specifically linked to
overwhelming emotional experience, which suggests
that they may serve the function of coping with those
high emotions and lead to the expresser’s well-being. It
should be noted that, ultimately, the baby’s well-being is
served by cuteness eliciting both expressions of care and
of aggression, because if the expresser is no longer inca-
pacitated with overwhelming positive affect, that person
may be better able to care for the baby (see Fig. 1).
Discussion
We found support for the idea that individuals’ self-
reports of dimorphous expressions correlate across situa-
tions and across the precise emotion expressed (e.g.,
happiness and excitement). Furthermore, responses to
cute stimuli appear to be of the same kind as other
dimorphous expressions of positive emotions, such as
crying when reuniting with a loved one. We next illus-
trated our model of the dimorphous expression of emo-
tion. As we hypothesized, people reported that they
would make more caring and aggressive expressions
after making higher positive appraisals and higher reports
of feeling overwhelmed with positive feelings toward the
stimuli that featured more characteristics of infancy.
One limitation to the studies presented in this article is
that all measures were self-reports. In Section S7 of the
Supplemental Material, we provide the results of a behav-
ioral investigation into the dimorphous expression of
emotion that corroborate what we report here. Another
limitation is the use of online samples, for which there is
little experimental control. In the Supplemental Material,
we report replications of Study 1 (see Sections S4, S5, and
S6) and experiments with university students in con-
trolled laboratory settings (Sections S7 and S8), in which
we found results similar to those reported here.
Study 2: Test of a Mechanism
Underlying Dimorphous Expressions
of Emotion
Next, we tested whether dimorphous expressions in reac-
tions to infantile stimuli functioned to regulate emotion. If
the dimorphous expression of emotion aids in emotion
regulation, we expected participants who spontaneously
express emotions in a dimorphous manner to return closer
to prearousal levels after viewing cute stimuli, relative to
those who do not show dimorphous responding. In other
words, we expected that in a manner analogous to that
reported by Fredrickson and Levenson (1998), negative
expressions would help to regulate positive emotions.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 7
We collected evidence of positive affect being regu-
lated by such dimorphic displays by measuring partici-
pants’ affective states before, directly after (peak of
experience), and 5 min after (end of recovery period)
exposure to more- and less-infantile stimuli. We predicted
that, overall, participants would report increased positive
affect directly after viewing the stimuli and decreased
positive affect following the recovery period. We further
predicted that participants who had reported wanting to
make aggressive expressions, relative to those who had
not, would show greater recovery from the high positive
affect by the end of the recovery period.
Here, we again tested the hypothesis that responses to
infantile stimuli are an example of the general dimor-
phous expression of positive emotions. We predicted that
participants’ questionnaire responses about dimorphous
displays in other domains (e.g., “I cry while watching the
happiest moments of movies”), but not the tendency to
express emotion in congruent ways (e.g., “I smile while
watching the happiest moments of movies”), would pre-
dict aggressive expressions captured while viewing cute
stimuli. We also collected additional measures of the
expression, the strength, and the dysregulation of emo-
tions to test the prediction that the dimorphic expression
of positive emotion would specifically explain aggressive
displays during viewing of cute stimuli.
Method
Participants. Participants were recruited online
through Amazon’s Mechanical Turk (N = 679; 390 female,
289 male; mean age = 37.88 years, range = 18–79, SD =
12.87) and compensated 75¢ for the approximately
20-min survey. The survey was advertised as follows:
“Short Survey (20–30 minutes to complete). In this survey
we ask you to answer a questionnaire, answer some
questions about photos, work on a puzzle, and answer
another questionnaire.”
To determine the number of participants needed for
this study, we used G*Power software and the effect size
(d = 0.24) from Sherman and colleague’s (2013) study
asking about participants’ current mood after viewing
these stimuli (1 = extremely negative, 9 = extremely posi-
tive). We felt that these mood ratings were the closest
equivalent to the scale we used. We estimated (.80 power,
alpha error probability = .05, two-tailed) that we would
need approximately 550 participants. However, because
we were collecting positive and negative affect with a
Positive
Appraisal
Care
Expression of
r = .69***
Overwhelmed
by Strong
Positive
Feelings
β = 0.30***
β = 0.55***
β = 0.44***
β = 0.25***
Positive
Appraisal
Expression of
Care
Expression of
Aggression
Overwhelmed
by Strong
Positive
Feelings
β = 0.30***
β = 0.36*** β = 0.55***
β = 0.44***
c path: β = 0.15**, cpath: β = –0.05, p = .52
c path: β = 0.14*, cpath: β = 0.03, p = .75
β = 0.25***
β = 0.35***
β = 0.39***
β = 0.36***
β = –0.08
β = 0.30***
More-
(vs. Less-)
Infantile
Stimuli
Fig. 1. Serial mediation models from Study 1 showing the influence of more-infantile (vs. less-infantile) stimuli on
both expressions of care and expressions of aggression, as mediated by positive appraisals of the stimuli and reports of
being overwhelmed with very strong positive feelings while viewing the stimuli. The model for each outcome variable
controlled for the other. Black lines show results for models in which both mediators were included. Gray lines show
results for models in which only positive appraisals were included. The one full path that is not significant is represented
by a dashed line. Asterisks indicate significant paths (*p < .05, **p < .01, ***p < .001). The c path in the model reflects the
total effect; the c path reflects the direct effect.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
8 Aragón et al.
different measure, we set a conservative stopping point
for data collection, ceasing when 700 surveys were
logged in as complete. Of the 735 participants who
logged in to the survey, 21 clicked through but did not
answer the survey, and 35 abandoned the survey before
completing the dependent variables (total attrition: N =
56 (8%); cute condition = 31, less-cute condition = 25),
which left data from 679 participants in the final
analysis.
Materials and stimuli. We administered measures of
emotional expression, dimorphous expression, strength
of emotion, and dysregulation of emotion, as well as the
Positive and Negative Affect Schedule (PANAS; Watson,
Clark, & Tellegen, 1988) and a prosocial measure. A
demographic questionnaire was also given. We used the
same stimuli as in Study 1.
To capture respondents’ experience and how they
express their emotions, we used the 16-item Berkeley
Expressivity Questionnaire (α = .87; Gross & John, 1997).
We added 32 adapted items (total = 48 items) that were
changed in two ways. First, some original items asked
about emotions but did not specify the emotion’s valence
(e.g., “My body reacts very strongly to emotional situa-
tions”). We rephrased such questions to ask both about
positive emotions (i.e., “My body reacts very strongly to
emotional, positive situations”) and about negative emo-
tions (i.e., “My body reacts very strongly to emotional,
negative situations”). Second, we added items to capture
the dimorphous expression of both positive and negative
emotions.
Expression of positive emotions was captured with
seven items (α = .90). Expression of negative emotions
was captured with seven analogous items (α = .80).
Dysregulation of positive emotions (α = .81) and dys-
regulation of negative emotions (α = .79) were captured
with six items each. Strength of positive emotions (α =
.82) and strength of negative emotions (α = .82) were
each captured with three items. Finally, dimorphous
expressions of positive emotions (α = .70) and dimor-
phous expressions of negative emotions (α = .66) were
each captured with three items. See Table 2 for all items.
These eight subscales were subjected to factor analy-
sis, which showed that 51% to 74% of the variance was
explained in each subscale by the first factor5 and that
loadings on the first factor were consistently high (all
items on all scales had Eigen values above .52 on the first
factor). (See Section S9 in the Supplemental Material for
intercorrelations of the eight subscales.)
The PANAS was administered before and after presen-
tation of the infantile stimuli, as well as after recovery.
The PANAS measures positive and negative affect by
presenting participants with 20 feelings and emotions
(e.g., “interested,” “irritable”) and asking them to rate “to
what extent you feel this way right now.” The 10 positive
items were averaged into a positive-affect score, and the 10
negative items were averaged into a negative-affect score.
The prompts during each trial and the instructions
were the same as in Study 1. They measured appraisals
of the babies (α = .82, two items), emotional experience
when looking at the baby, care expressions (α = .92,
three items), and aggressive expressions (α = .86, three
items).
To create a filler task, we designed a word-search puz-
zle on the Web site Discovery.com using words we
judged to be unrelated to the main task (e.g., “bay,
“brook,” “coastline”). Participants were told that the “puz-
zle may be difficult. Please do not worry about finding all
of the words. This page will automatically advance after
5 minutes.” The puzzle had 30 words, which is more than
would typically be found in a matrix this large (20 let-
ters× 20 letters) in the 5-min interval. We did this to keep
the experience similar for all participants. Because this
was an online study, it cannot be said whether all indi-
viduals attended to the puzzle equally. However, the
number of words found in the less-infantile (M = 5.12
words found, SD = 2.84) and more-infantile (M = 5.26
words found, SD = 3.10) conditions did not differ signifi-
cantly, p = .55. Therefore, differences in recovery from
the arousing event are not likely to have been accounted
for by performance on this filler task, as equal perfor-
mance should be an indication of equal attention given
to the task. This supports the idea that attention to things
other than the survey did not differ by condition.
The subscales of Universalism (rating values of equal-
ity, world peace, social justice, broadmindedness, and
wisdom; α = .75) and Benevolence (rating feelings of
helpfulness, honesty, forgiveness, loyalty, and responsi-
bility; α = .68) from the Schwartz Value Survey (Schwartz,
1992) were combined into one scale of prosocial care
(α= .79), following the protocol used by Sherman and
colleagues (2013). We used the standard Schwartz instruc-
tions and presentation of values. (See Section S10 in the
Supplemental Material for details on the use of this scale
in replicating Sherman et al.’s, 2013, study on care
responses to cute stimuli and prosocial values.)
Finally, our demographics questionnaire included
items regarding age, ethnicity, the participants’ number of
children, desire for children (or for more children if par-
ticipants had any already), and whether participants who
did not have children regularly spent time with children.
Procedure. Participants provided informed consent and
were given a short introduction to the survey that
explained that there would be questionnaires, pictures
with questions, a puzzle, and another questionnaire. We
first asked participants to fill out the emotional-expressiv-
ity measure. We then administered the preexperiment
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 9
Table 2. Survey Items Used in Study 2
Category and item Original itema
Expression of positive emotions (α = .90)
When I am feeling positive it is written all over my face with positive expressions. What I’m feeling is written all over my face.
Whenever I feel positive emotions, people can easily see exactly what I am
feeling.
When I am feeling strong positive emotions, I express with positive expressions.
I smile while watching the happiest moments of movies. I cry during sad movies.
When I am feeling a strong positive emotion my expression can look like I am
feeling a positive emotion.
When I’m happy, my feelings show.
I am an emotionally expressive person when it comes to positive emotions. I am an emotionally expressive person.
Expression of negative emotions (α = .80)
When I am feeling negative it is written all over my face with negative
expressions.
What I’m feeling is written all over my face.
Whenever I feel negative emotions, people can easily see exactly what I am
feeling.
When I am feeling strong negative emotions, I express with negative expressions.
I cry while watching the saddest moments of movies. I cry during sad movies.
When I am feeling a strong negative emotion, my expression can look like
I am feeling a negative emotion.
When I’m sad, my feelings show. When I’m happy, my feelings show.
I am an emotionally expressive person when it comes to negative emotions. I am an emotionally expressive person.
Dysregulation of positive emotions (α = .81)
It is difficult for me to hide my excitement. It is difficult for me to hide my fear.
It is difficult for me to hide my happiness. It is difficult for me to hide my fear.
It is difficult for me to hide my joy. It is difficult for me to hide my fear.
I am sometimes unable to hide my positive feelings, even though I would
like to.
I am sometimes unable to hide my feelings,
even though I would like to.
There have been times when I have not been able to stop laughing even
though I tried to stop.
There have been times when I have not been able to stop smiling even
though I tried to stop.
There have been times when I have not
been able to stop crying even though I
tried to stop.
Dysregulation of negative emotions (α = .79)
It is difficult for me to hide my anger. It is difficult for me to hide my fear.
It is difficult for me to hide my anxiety. It is difficult for me to hide my fear.
It is difficult for me to hide my fear. It is difficult for me to hide my fear.
I am sometimes unable to hide my negative feelings, even though I would
like to.
I am sometimes unable to hide my feelings,
even though I would like to.
There have been times when I have not been able to stop crying even
though I tried to stop.
No matter how nervous or upset I am I tend to keep a calm exterior.
Strength of positive emotions (α = .82)
My body reacts very strongly to emotional, positive situations. My body reacts very strongly to emotional
situations.
My positive emotions can be very strong. My emotions can be very strong.
I experience my positive emotions (for example happiness, relief, connected-
ness, or peacefulness) very strongly.
I experience my emotions very strongly.
Strength of negative emotions (α = .82)
My body reacts very strongly to emotional, negative situations. My body reacts very strongly to emotional
situations.
My negative emotions can be very strong. My emotions can be very strong.
I experience my negative emotions (for example sadness, anger, fear or
anxiety) very strongly.
I experience my emotions very strongly.
(continued)
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
10 Aragón et al.
PANAS. Next, we told participants that they would be
asked to respond to some photographs with the same
instructions as in Study 1 (i.e., “not in reference to doing
actual harm or disliking the baby”). Participants were
allowed to move only forward through the survey; no
back button was provided.
We presented trials in which we measured aggressive
expressions and care expressions in random order.
Directly after these in-the-moment trials, we administered
the postexperimental PANAS. Participants were then
asked to work on the puzzle for 5 min. The online page
was designed not to advance during this period. Directly
following this filler task, we administered the recovery
PANAS. We then collected appraisals of the babies after a
second exposure to the photographs, followed by a pro-
social measure (the Schwartz Value Survey). Finally, we
collected demographic information. See Figure 2 for an
illustration of the experimental paradigm.
Results
Replication of Study 1. A comparison of means
showed that more-infantile babies (M = 71.14, SD =
19.58) were appraised more positively than less-infantile
babies (M = 67.08, SD = 20.79); independent-samples
t tests confirmed that this difference was significant,
t(677) = 2.62, p < .01, d = 0.20. As expected, participants
reported being more overwhelmed with positive feelings
when viewing photographs in the more-infantile condi-
tion (M = 53.10, SD = 28.46) than when viewing photo-
graphs in the less-infantile condition (M = 47.75, SD =
28.16), t(677) = 2.46, p = .01, d = 0.19. Participants
reported marginally higher care expressions for more-
infantile babies (M = 58.64, SD = 26.70) than for less-
infantile babies (M = 51.15, SD = 25.90), t(677) = 1.73, p=
.08, d = 0.29. Participants also reported higher aggressive
expressions for more-infantile babies (M = 37.62, SD =
22.82) than for less-infantile babies (M = 33.29, SD =
23.12), t(677) = 2.46, p = .01, d = 0.19.
Next, we ran a bootstrapped serial mediation model
with 5,000 samples using PROCESS Model 6 (Hayes,
2013). As in Study 1, this model predicted the influence of
viewing more-infantile (vs. less-infantile) babies on care
expressions, through positive appraisals and being over-
whelmed with very strong positive feelings, while control-
ling for expressions of aggression. There was a significant
indirect path (95% CI = [0.01, 0.10]) from viewing more-
infantile babies through the participants’ positive apprais-
als of such babies (b = 0.20,6 SE = 0.07), t = 2.62, p < .01,
and then through being overwhelmed with very strong
positive feelings (b = 0.65, SE = 0.03), t = 21.93, p < .001,
to the expressions of care made toward those babies (b =
0.43, SE = 0.03), t = 14.05, p < .001. The manipulation of
infant characteristics no longer marginally predicted care
responses when positive appraisals and being over-
whelmed with very strong positive feelings were included
in the model (c path; b = −0.04, SE = 0.05), t = −0.85, p =
.40. Again there was a second significant indirect pathway
from infant characteristics through positive appraisals pre-
dicting care responses (95% CI = [0.02, 0.16]). This sug-
gests that being overwhelmed with very strong positive
feelings is not essential to experiencing care responses
but having a positive appraisal of the baby is.
We ran another serial mediation model to predict
aggressive expressions from infant characteristics,
through the pathways of positive appraisals and being
overwhelmed with very strong positive feelings, while
controlling for expressions of care. There was a signifi-
cant indirect path (95% CI = [0.02, 0.12]) from viewing
more-infantile babies (vs. less-infantile babies) through
Category and item Original itema
Dimorphous expressions of positive emotions (α = .70)
I cry while watching the happiest moments of movies.
When I am feeling strong positive emotions, I express with negative expressions.
When I am feeling a strong positive emotion (for example extreme happiness,
strong sense of relief, strong feeling of connection to others etc.), my expression
can look like I am feeling a negative emotion (for example I might cry, or scream
as though in fear even though I am happy or excited).
Dimorphous expressions of negative emotions (α = .66)
I sometimes smile while watching the saddest moments of movies.
When I am feeling strong negative emotions, I express with positive expressions.
When I am feeling a strong negative emotion (for example deep sadness, strong
anxiety, strong anger, etc.), my expression can look like I am feeling a positive
emotion (for example I might smile or chuckle even though I am sad, anxious,
or angry).
aOriginal items from which the current items were adapted are from the Berkeley Expressivity Questionnaire (Gross & John, 1997).
Table 2. (continued)
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 11
positive appraisals of such babies (b = 0.20, SE = 0.07),
t= 2.62, p < .01, and then through being overwhelmed
with very strong positive feelings (b = 0.65, SE = 0.03),
t= 21.93, p < .001, to the aggressive expressions made
toward those babies (b = 0.54, SE = .04), t = 13.36, p <
.001. In confirmation of mediation, the manipulation of
infant characteristics no longer predicted aggressive
expressions when positive appraisals and being over-
whelmed with very strong positive feelings were
included in the model (c path; b = 0.07, SE = 0.06), t =
1.136, p = .25. There were no other significant pathways,
which indicated that it was solely through the experience
of being overwhelmed with very strong positive feelings
that the expression of aggression occurred.
Testing emotion regulation through the dimor-
phous expression of emotion. We collected partici-
pants’ affective states before, directly after (our assumed
peak of experience), and 5 min after exposure to more-
and less-cute stimuli. We predicted that participants
would report increased positive affect immediately after
viewing the stimuli and decreased affect after the recov-
ery period, but we also predicted that those who had
reported aggressive displays at the moment of viewing
the stimuli would show greater recovery from the peak of
the postexperimental measurement to the 5-min postre-
covery measurement than those who reported fewer
aggressive displays of emotion.
We created a linear mixed model to account for the
repeated measurements of positive affect. We tested our
predictions that positive affect would increase from
before to after exposure to our stimuli (postexperiment
score = postexperiment positive affect – preexperiment
positive affect) and that positive affect would decrease
from after the exposure to after the recovery period
(postrecovery score = postrecovery positive affect – post-
experiment positive affect) with measurement (postex-
periment score, postrecovery score), stimulus condition
(more infantile, less infantile), and the interaction
between measurement and stimulus condition as fixed
factors. Participants’ preexperiment positive-affect scores
were entered as a covariate. As expected, positive affect
increased in the postexperiment score (b = 0.07, SE =
0.02) and decreased in the postrecovery score (b = −0.24,
SE = 0.02), t(666.01) = 6.91, p < .001. No other factors or
interactions were significant.
Next, we tested whether people who expressed ag -
gression had systematically different postrecovery scores
than those who did not. We ran a linear regression model
with postrecovery scores predicted by expressions of
aggression, and peak-of-experience positive-affect scores
as a covariate. As one might expect, participants who had
the higher peaks of positive affect after viewing babies
showed greater declines in positive affect (b = −0.10, SE =
0.03, β = −0.14), t = −3.38, p < .001. Even after controlling
for this main effect, there was an improvement to the
Emotion-
Expressivity
Measures
P
A
N
A
S
P
A
N
A
S
P
A
N
A
S
Time
Filler Task:
Word
Search
(5 min)
Prosocial
Measure
Demo-
graphics
Random Assignment to
More- and Less-Infantile
Stimuli: 2 Prompts ×
8 Photos = 16 Trials
Care and Aggression
Expressions
(Counterbalanced):
Responses Collected
During Viewing of Stimuli
Same More- and Less-
Infantile Stimuli as in
Main Task
Appraisal of Stimuli
and Endorsement of
Overwhelming Positive
Emotions
Fig. 2. Design of Study 2. In the main task, participants viewed photos of infants that were manipulated to look more and less infantile, and then
participants rated their own expressions of care and aggression. Participants completed the Positive and Negative Affect Schedule (PANAS; Watson,
Clark, & Tellegen, 1988) before viewing stimuli, after viewing stimuli, and after a filler task. After the final PANAS, participants viewed the stimuli
again and endorsed statements designed to capture how positive they felt about the stimuli and whether they had a positive overwhelming emo-
tional response to each one. Additional measures were completed before and after the experiment.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
12 Aragón et al.
model (p for ΔR2 = .01). When aggressive-expression
scores were entered into the model, the results showed
that as predicted, participants who expressed more
aggression while viewing the babies showed a greater
decline in positive affect during the recovery period (b =
−0.06, SE = 0.03, β = −0.10), t = −2.53, p = .01, R2 = .04.
Test of general mechanism: predicting aggressive
expressions to cute stimuli with individual-differ-
ence measures of dimorphous expression in other
domains. We constructed a linear regression model in
which we entered condition (more infantile = 1, less infan-
tile = 0), the dimorphous expression of positive emotion,
the congruent expression of positive emotion, the dysreg-
ulation of positive emotion, and the strength of positive
emotion as predictors of expressions of aggression, as well
as all interactions between condition and the four positive-
emotion variables. As would be expected from the prior
analyses of this sample, condition predicted expressions of
aggression in response to the babies, with more-infantile
babies eliciting higher aggressive expressions than less-
infantile babies (b = 3.85, SE = 1.70, β = 0.08), t = 2.62, p =
.02. We also found, as predicted, a main effect of individ-
ual-difference reports of dimorphous expressions of posi-
tive emotions, which predicted increased aggressive
expressions in response to the infant stimuli regardless of
condition (b = 5.42, SE= 1.32, β = 0.24), t = 4.10, p < .001.
The congruent expression of emotion, the dysregulation
of positive emotion, and the strength of positive emotion
did not predict expressions of aggression in this experi-
ment, all ps > .18, which emphasizes the specificity of the
dimorphous predictor. That is, it is not just any expression
of emotion that is related to these responses, but specifi-
cally the dimorphous expression of emotion. No interac-
tions were significant.
Given that average ratings of being overwhelmed with
very positive feelings in Study 1 (M = 48.18, SD = 25.27)
and Study 2 (M = 50.44, SD = 28.42) were only moderate,
some readers may wonder why moderate positivity
would create these dimorphous expressions that we
claim take place when an individual is overwhelmed.
However, we remind the reader that we did not ask par-
ticipants how positive they were feeling, but rather we
asked them how true the statement that they were “over-
whelmed with very strong positive feelings” was for
them, and the average of participants’ overall responses
in both conditions was equivalent to “true.
More important, in our studies, we randomly assigned
participants to infantile-characteristics conditions, with
the understanding that people have varying emotional
responses to babies. Indeed, the variation that we saw
among participants in such feelings of being overwhelmed
should have been randomly distributed between condi-
tions. We randomly assigned participants to condition to
experimentally manipulate such feelings of being over-
whelmed with carefully controlled stimuli, and there were
indeed meaningful differences between conditions.
We assert—and our statistics indicate—that feelings of
being overwhelmed are an essential component in the
dimorphous expression of emotions. Even so, this asser-
tion may seem less than intuitive, particularly if the indi-
vidual is not likely one to be overwhelmed by our stimuli
or one who expresses emotion in a dimorphous manner.
Therefore, in Figure 3, we provide a distribution of par-
ticipants’ scores of being overwhelmed with very strong
positive feelings (the average of each participant’s ratings
across the eight stimuli) cross-tabulated with distributions
of aggressive displays (depicted here by a split at the
score of 50 on the scale from 0 to 100) for all participants
in both Studies 1 and 2 combined (N = 978).
Discussion
Our hypothesis that there are two distinct expressions
arising from a single stimulus, a singular positive appraisal
and a singular emotional response, was supported. It has
long been established that infantile characteristics spur
caretaking behaviors in adults (Lorenz, 1943). We found
such care expressions in our investigation, and it follows
that caretakers’ expressions of caring serve the well-being
of infants (e.g., Lorenz, 1971). We also found expressions
of aggression, and we presume that these expressions
may help regulate emotion and support the immediate
well-being of the caretaker.
General Discussion
In the present research, we illustrated the pathways through
which dimorphous expressions arise and the possible func-
tion of such expressions (Levenson, 1994) as regulators of
emotion. Dimorphous expressions may arise to regulate
positive emotions because (a) people have less experience
suppressing positive than negative emotions (Oishi, Diener,
& Lucas, 2007; Wegener & Petty, 1994), (b) people might
not be motivated to cognitively reappraise positive events
(i.e., rethink them in negative ways), and (c) even though
removing themselves from a positive situation is an effec-
tive strategy to regulate strong positive emotions (Nezlek &
Kuppens, 2008), people may not be able to do so at all
times, such as when taking care of a child. Exactly how
these emotions are regulated through the expression of a
second expression will need to be explored in future
research. Perhaps dimorphous displays of emotion reflect
the onset of a second emotion that arises to tamp down the
original overwhelming emotion, or perhaps dimorphous
expressions elicit physiological shifts away from intense
positive emotions through afferent facial or postural feed-
back (e.g., Strack, Martin, & Stepper, 1988).
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 13
Additionally, perhaps people who feel positive express
negativity to give important events the appropriate gravi-
tas (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001),
or maybe a dimorphous expression serves as a social
signal to other people that the expresser is overwhelmed
or incapacitated (Clark, Pataki, & Carver, 1996; Kappas,
2013; Smith, Mahler, Peciña, & Berridgem, 2010).
We defined and tested a model of dimorphous emo-
tional expression—that is, the expression of negativity
when one feels overwhelming positivity. We found that
individuals who express positive emotions in this dimor-
phous manner do so across a variety of emotionally pro-
voking situations. We used cute stimuli (an elicitor of
positive emotion) to illustrate the existence of these dimor-
phous expressions, as well as to provide preliminary evi-
dence of their possible function as regulators of emotion.
Author Contributions
O. R. Aragón originated the idea and theoretical framework,
developing it further with all authors. O. R. Aragón designed all
studies, with assistance from M. S. Clark and R. L. Dyer on the
studies reported in Sections S2 and S4 through S7 in the
Supplemental Material available online. O. R. Aragón collected
all data, with assistance from R. L. Dyer on the studies reported
in Sections S2 and S4 through S7 in the Supplemental Material.
O. R. Aragón analyzed all data and drafted the manuscript. J. A.
Bargh and M. S. Clark provided critical revisions. All authors
approved the final version of the manuscript for submission.
Acknowledgments
We thank Paul Bloom for feedback on the early theoretical
framework of dimorphous expressions. We also thank Lindsay
Davis and Chelcie Piasio for their assistance in the studies
reported in Sections S2, S7, and S8 in the Supplemental Material
available online.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Supplemental Material
Additional supporting information can be found at http://pss
.sagepub.com/content/by/supplemental-data
Notes
1. See Section S1 in the Supplemental Material for distinctions
among theories of dimorphous expression in other models of
emotional responses.
2. The possibility that two emotions can be experienced simul-
taneously, particularly one of negative and one of positive
0
50
100
150
200
250
300
Not at All True A Little Bit True True Very True Completely True
Number of Participants
Response
High Aggression
Low Aggression
Fig. 3. Distribution of responses to the question, “When I look at this baby, I feel like
I am overwhelmed by very strong positive feelings,” among participants in Studies 1
and 2 who reported high and low levels of aggression in response to infant stimuli.
Scores above 50.000 were classified as high; scores at or below 50.000 were classified
as low.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
14 Aragón et al.
valence, has been suggested previously (see Andrade & Cohen,
2007; Schimmack, 2001, for discussions).
3. This article focuses primarily on the dimorphous expression
of intense positive emotions, but we thought it important to
explore the dimorphous expression of intense negative emo-
tions as well.
4. The reported coefficients are from models in which all vari-
ables were standardized.
5. All items loaded on one factor except for dimorphic expres-
sion of negative emotion, in which the question “I sometimes
smile while watching the saddest moments of movies” appeared
to indicate a second factor.
6. The reported coefficients are from models in which all vari-
ables were standardized.
References
Alley, T. R. (1981). Head shape and the perception of cuteness.
Developmental Psychology, 17, 650–654.
Anderson, C. A., & Bushman, B. J. (2002). Human aggression.
Annual Review of Psychology, 53, 27–51.
Andrade, E. B., & Cohen, J. B. (2007). On the consumption
of negative feelings. Journal of Consumer Research, 34,
283–300.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs,
K.D. (2001). Bad is stronger than good. Review of General
Psychology, 5, 323–370.
Carroll, J. M., & Russell, J. A. (1996). Do facial expressions sig-
nal specific emotions? Judging emotion from the face in
context. Journal of Personality and Social Psychology, 70,
205–218.
Clark, M. S., Pataki, S. P., & Carver, V. H. (1996). Some thoughts
and findings on self-presentation of emotions in relation-
ships. In G. J. O. Fletcher & J. Fitness (Eds.), Knowledge
structures in close relationships: A social psychological
approach (pp. 247–274). New York, NY: Psychology Press.
Colom, F., Vieta, E., Martínez-Arán, A., Reinares, M., Benabarre,
A., & Gastó, C. (2000). Clinical factors associated with treat-
ment noncompliance in euthymic bipolar patients. The
Journal of Clinical Psychiatry, 61, 549–555.
Ekman, P., & Friesen, W. V. (1971). Constants across cultures
in the face and emotion. Journal of Personality and Social
Psychology, 17, 124–129.
Ekman, P., & Friesen, W. V. (1986). A new pan-cultural facial
expression of emotion. Motivation and Emotion, 10, 159–
168.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007).
G*Power 3: A flexible statistical power analysis program for
the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39, 175–191.
Folkman, S., & Moskowitz, J. T. (2000). Positive affect and
the other side of coping. American Psychologist, 55,
647–654.
Fredrickson, B. L., & Levenson, R. W. (1998). Positive emotions
speed recovery from cardiovascular sequelae of negative
emotion. Cognition & Emotion, 12, 191–220.
Fredrickson, B. L., Mancuso, R. A., Branigan, C., & Tugade,
M. M. (2000). The undoing effect of positive emotions.
Motivation and Emotion, 24, 237–258.
Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W.,
Gur, R. C., & Sachser, N. (2009). Baby schema in infant
faces induces cuteness perception and motivation for care-
taking in adults. Ethology, 115, 257–263. doi:10.1111/j.1439-
0310.2008.01603.x
Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W.,
Valdez, J. N., Griffin, M. D., . . . Gur, R. C. (2009). Baby
schema modulates the brain reward system in nulliparous
women. Proceedings of the National Academy of Sciences,
USA, 106, 9115–9119.
Gross, J. J. (2013). Emotion regulation: Taking stock and mov-
ing forward. Emotion, 13, 359–365.
Gross, J. J., & John, O. P. (1997). Revealing feelings: Facets
of emotional expressivity in self-reports, peer ratings, and
behavior. Journal of Personality and Social Psychology, 72,
435–448.
Gross, J. J., & John, O. P. (2003). Individual differences in two
emotion regulation processes: Implications for affect, rela-
tionships, and well-being. Journal of Personality and Social
Psychology, 85, 348–362.
Gross, J. J., John, O. P., & Richards, J. M. (2000). The disso-
ciation of emotion expression from emotion experience: A
personality perspective. Personality and Social Psychology
Bulletin, 26, 712–726.
Hayes, A. F. (2013). Introduction to mediation, moderation,
and conditional process analysis. New York, NY: Guilford
Press.
Hildebrandt, K. A., & Fitzgerald, H. E. (1979). Facial feature
determinants of perceived infant attractiveness. Infant
Behavioral Development, 2, 329–339.
Kappas, A. (2003). What facial activity can and cannot tell us
about emotions. In M. Katsikitis (Ed.), The human face:
Measurement and meaning (pp. 215–234). Dordrecht, The
Netherlands: Kluwer.
Kappas, A. (2011). Emotion and regulation are one! Emotion
Review, 3, 17–25.
Kappas, A. (2013). Social regulation of emotion: Messy layers.
Frontiers in Psychology, 4, Article 51. Retrieved from http://
journal.frontiersin.org/Journal/10.3389/fpsyg.2013.00051/full
Levenson, R. W. (1994). What is the function of emotion?
In E. Paul & R. Davidson (Eds.), The nature of emotion:
Fundamental questions (pp. 97–177). New York, NY:
Oxford University Press.
Lorenz, K. (1943). Die angeborenen formen m̈glichen erfah-
rung [The innate forms of potential experience]. Zeitschrift
für Tierpsychologie, 5, 245–409.
Lorenz, K. (1971). Studies in animal behavior. Cambridge, MA:
Harvard University Press.
Nezlek, J. B., & Kuppens, P. (2008). Regulating positive and
negative emotions in daily life. Journal of Personality, 76,
561–580.
Oishi, S., Diener, E., & Lucas, R. E. (2007). The optimum level
of well-being: Can people be too happy? Perspectives on
Psychological Science, 2, 346–360.
Rubino, C. R. G., & Llenado, M. G. T. (2002). Tagalog-English,
English-Tagalog dictionary. New York, NY: Hippocrene.
Samson, A. C., & Gross, J. J. (2012). Humour as emotion regula-
tion: The differential consequences of negative versus posi-
tive humour. Cognition & Emotion, 26, 375–384.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
Dimorphous Expressions 15
Schimmack, U. (2001). Pleasure, displeasure, and mixed feel-
ings: Are semantic opposites mutually exclusive? Cognition
& Emotion, 15, 81–97.
Schwartz, S. H. (1992). Universals in the content and struc-
ture of values: Theory and empirical tests in 20 countries.
In M. P. Zanna (Ed.), Advances in experimental social
psychology (Vol. 25, pp. 1–65). New York, NY: Academic
Press.
Sherman, G. D., Haidt, J., & Coan, J. A. (2009). Viewing cute
images increases behavioral carefulness. Emotion, 9,
282–286.
Sherman, G. D., Haidt, J., Iyer, R., & Coan, J. A. (2013).
Individual differences in the physical embodiment of care:
Prosocial women respond to cuteness by becoming more
physically careful. Emotion, 13, 151–158.
Smith, K. S., Mahler, S. V., Peciña, S., & Berridge, K. C. (2010).
Hedonic hotspots: Generating sensory pleasure in the
brain. In K. C. Berridge (Ed.), Pleasures of the brain (pp.
27–49). New York, NY: Oxford University Press.
Strack, F., Martin, L. L., & Stepper, S. (1988). Inhibiting and
facilitating conditions of the human smile: A nonobtrusive
test of the facial feedback hypothesis. Journal of Personality
and Social Psychology, 54, 768–777.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development
and validation of brief measures of positive and negative
affect: The PANAS scales. Journal of Personality and Social
Psychology, 54, 1063–1070.
Wegener, D. T., & Petty, R. E. (1994). Mood management
across affective states: The hedonic contingency hypothesis.
Journal of Personality and Social Psychology, 66, 1034–1048.
Zaki, J., Hennigan, K., Weber, J., & Ochsner, K. N. (2010). Social
cognitive conflict resolution: Contributions of domain-gen-
eral and domain-specific neural systems. The Journal of
Neuroscience, 30, 8481–8488.
by Rebecca Dyer on February 2, 2015pss.sagepub.comDownloaded from
... Because the feeling of kawaii is associated with a motivation to approach a target, touch behaviors create a situation where the robot's body part (i.e., a hand) closely approaches the target. Other studies identified a phenomenon called cute aggression that captures the relationship between touch and feelings of kawaii [14,15]. These studies reported behaviors or attitudes toward cute things, e.g., participants felt like "when I look at this baby, I feel like pinching her cheeks or being playfully aggressive" [15]. ...
... Touch behaviors are related to expressing the feeling of kawaii because such feelings are associated with a motivation to approach a target and are related to aggressive touch behaviors [5,8,14]. Past studies reported the positive effects of kawaii on people's behavioral changes [4,5], providing positive feelings [1], and attractiveness [2,3]. ...
... We also investigated the effects of motion styles in the context of expressing a feeling of kawaii, based on the concept of cute aggression [14]. We prepared two motion styles: normal and emphasized. ...
... However, such an assumption of a 1:1 correspondence between facial expressions and experience is upended when considering how these expressions are actually used in real life. Our literature (for example, Fernández-Dols and Ruiz-Belda, 1995;Carroll and Russell, 1996;Fernández-Dols et al., 1997;Aviezer et al., , 2012Fernandez-Dols and Crivelli, 2013;Aragón et al., 2015;García-Higuera et al., 2015;Durán et al., 2017;Aragón and Bargh, 2018;Aragón and Clark, 2018) and lives abound with examples of violations of this supposed correspondence, e.g., the happy tears upon the birth of one's child, the seemingly violent rage across a soccer field upon winning, a smile when embarrassed, and the tooth baring growl at a cute little baby. There might be a definitional understanding of a correspondence between expression and discreet emotions, but that appears to be separate and apart from how expression and discrete emotions correspond in real life. ...
... Previous research has shown that when anger and sadness expressions are situated within positive contexts the majority of observers show consensus in their interpretation of the expressions as representing predominantly positive-not negative experiences (Fernández-Dols and Ruiz-Belda, 1995;Aviezer et al., 2012;Aragón et al., 2015;Aragón, 2016Aragón, , 2017Wenzler et al., 2016;Aragón and Bargh, 2018;Aragón and Clark, 2018). This finding that anger and sadness expressions can be associated with predominantly positive emotions has been consistent whether the expressions arose within participants themselves during emotionally evocative situations (Aragón et al., 2015;Aragón, 2017), the expressions were presented to participants to probe for reflection of their own past experiences (Aragón and Bargh, 2018), or when participants were asked to interpret what those expressions might represent Aragón, 2016Aragón, , 2017Aragón, , 2020Aragón and Bargh, 2018;Wenzler et al., 2016). ...
... Previous research has shown that when anger and sadness expressions are situated within positive contexts the majority of observers show consensus in their interpretation of the expressions as representing predominantly positive-not negative experiences (Fernández-Dols and Ruiz-Belda, 1995;Aviezer et al., 2012;Aragón et al., 2015;Aragón, 2016Aragón, , 2017Wenzler et al., 2016;Aragón and Bargh, 2018;Aragón and Clark, 2018). This finding that anger and sadness expressions can be associated with predominantly positive emotions has been consistent whether the expressions arose within participants themselves during emotionally evocative situations (Aragón et al., 2015;Aragón, 2017), the expressions were presented to participants to probe for reflection of their own past experiences (Aragón and Bargh, 2018), or when participants were asked to interpret what those expressions might represent Aragón, 2016Aragón, , 2017Aragón, , 2020Aragón and Bargh, 2018;Wenzler et al., 2016). These patterns were consistent whether anger and sadness expressions were pulled from photographs of real-life contexts (Aragón and Bargh, 2018), created through trained actors (Aragón, 2017;Aragón and Clark, 2018), were photographs of anger and sadness classified through facial action coding (Karolinska directed emotional faces; Lundqvist et al., 1998 as used in Aragón andBargh, 2018), and whether experiences had been presented through static photographs, narrative accounts, or dynamic video displays. ...
... Emotional tears are consistently associated with sadness, and both tears and sadness signal helplessness or the need for support. If contexts were found to reduce or nullify the sadness enhancement effect of tears but support or enhance the perception of other emotions (e.g., happiness) in tearful expressions, it might suggest that tears could be a form of dimorphous emotion expressions (i.e., two different expressions with one emotion origin, Aragón et al., 2015) where intense positive emotions could elicit emotional tears. Nonetheless, the lack of research on this topic makes it unclear whether this speculation would be empirically supported. ...
... These showed that our studies differed from Aragón and Clark's study in many ways, and the methodological differences as a whole could have led to inconsistent findings. Investigating whether the contexts remove the perceived helplessness of the tearing individual could help to address the uncertainties in the boundary conditions of the sadness enhancement effects of tears and the conditions in which tears become part of dimorphous emotional expressions that conveys intense emotions instead (Aragón et al., 2015). ...
Article
Emotional tears tend to increase perceived sadness in facial expressions. However, it is unclear whether tears would still be seen as an indicator of sadness when a tearful face is observed in an emotional context (e.g., a touching moment during a wedding ceremony). We examine the influence of context on the sadness enhancement effect of tears in three studies. In Study 1, participants evaluated tearful or tearless expressions presented without body postures, with emotionally neutral postures, or with emotionally congruent postures (i.e., postures indicating the same emotion as the face). The results show that the presence of tears increases the perceived sadness of faces regardless of context. Similar results are found in Studies 2 and 3, which used visual scenes and written scenarios as contexts, respectively. Our findings demonstrate that tears on faces reliably indicate sadness, even in the presence of contextual information that suggests non‐sadness emotions.
... In addition, past studies confirmed "cute aggression," which describes the relationship between aggressive behaviors and perceived strong kawaii feelings [23,24]. For example, a participant reported that "looking at this baby makes me want to pinch her cheeks and be playfully aggressive" [24]. ...
... In this study, we investigated the influences of viewing a touch behavior toward feelings of kawaii. Using non-static visual stimuli is one unique point compared to past related studies that used static visual stimuli (e.g., pictures) [8,23,24]; in this context, using different ...
Article
Full-text available
We investigated how a presenter’s touching behaviors of an object during its explanation affect the observer’s perceived feelings of kawaii , a Japanese word that means “cute,” toward the object and the presenter. We conducted a face-to-face experiment with a robot presenter as well as a web survey experiment with both robot and human presenters. Based on the phenomenon that people more firmly touch an object when their perceived kawaii feeling is overwhelmingly strong, we investigated the effects of touching behavior with emphasized styles. First, we conducted a face-to-face experiment with a robot presenter where participants observed their presentations about an object to explain its characteristics. The results showed that participants who observed the robot’s touch behaviors perceived the object to be more kawaii and thought that the robot also felt the object was more kawaii . On the other hand, the results did not effectively show any increase in the participant’s feelings of kawaii toward the robot or the emphasized touch style. Based on these results, we next conducted a web survey experiment to investigate whether such knowledge about touching effects is applicable for human presenters. The results resembled those obtained when the presenter was a robot, i.e., viewing a touch behavior increased both the presenter’s perceived feelings of kawaii toward the object and the participant’s feelings of kawaii toward it. These results suggest that viewing the touch behaviors of others influenced the perceived emotional feelings toward both presenters and objects.
... Moreover, the findings demonstrated the importance of soft body contact for subsequent psychological development, a finding that Harlow implied would most likely also apply to humans. Additionally, the psychological phenomenon known as dimorphous expressions [64] might also be relevant for soft robots. In [64] researchers found that when presented with an image of something cute that produced positive emotions, people also experienced stronger aggressive expressions, such as wanting to pinch a baby's cheeks. ...
... Additionally, the psychological phenomenon known as dimorphous expressions [64] might also be relevant for soft robots. In [64] researchers found that when presented with an image of something cute that produced positive emotions, people also experienced stronger aggressive expressions, such as wanting to pinch a baby's cheeks. Dimorphous expressions explain why people make both caring and aggressive gestures towards appealing stimuli: a soft body that is pleasing to touch could inadvertently also prompt aggressive touching or excessive force. ...
Article
Full-text available
Soft robotics technology has been proposed for a number of applications that involve human–robot interaction. It is commonly presumed that soft robots are perceived as more natural, and thus more appealing, than rigid robots, an assumption that has not hitherto been tested or validated. This study investigates human perception of and physical interaction with soft robots as compared with rigid robots. Using a mixed-methods approach, we conducted an observational study to explore whether soft robots are perceived as more natural, and what types of interactions soft robots encourage. In a between-subjects study, participants interacted with a soft robotic tentacle or a rigid robot of a similar shape. The interactions were video recorded, and data was also obtained from questionnaires (Nvideo = 123, Nquest = 94). Despite their drastically different appearances and materials, we found no significant differences in how appealing or natural the robots were rated to be. Appeal was positively associated with perceived naturalness in all cases, however we observed a wide variation in how participants define “natural”. Although participants showed no clear preference, qualitative analysis of video data indicated that soft robots and rigid robots elicit different interaction patterns and behaviors. The findings highlight the key role of physical embodiment and materiality in human–robot interaction, and challenge existing assumptions about what makes robots appear natural.
... The main studies in the literature that have focused on TOJ consider it as a dimorphous expression of positive emotions (cf. Aragón et al., 2015;Aragón, 2017;Aragón and Clark, 2017): the strong emotion felt (joy) is manifested through an expressive mode (crying) that refers to a different, or even opposite, emotional state (sadness, pain). An example of this is the tears shed by Olympic athletes after winning a gold medal: their joy is so enormous and uncontrollable that it triggers a crying reaction. ...
Article
Full-text available
This article describes a research project in which a qualitative research was carried out consisting of 24 semi-structured interviews and a subsequent data analysis using the MAXQDA software in order to investigate a particular dimorphic emotional expression: tears of joy (TOJ). The working hypothesis is that TOJ are not only an atypical expression due to a “super joy,” or that they are only an attempt by the organism to self-regulate the excess of joyful emotion through the expression of the opposite emotion (sadness), but that it is an emotional experience in its own right—not entirely overlapping with joy—with a specific adaptive function. Through the interviews, conducted in a cross-cultural context (mainly in India and Japan), we explored the following possibility: what if the adaptive function of crying for joy were to signal, to those experiencing it, the meaning of their life; the most important direction given to their existence? The material collected provided positive support for this interpretation.
... According to Aragόn and her colleagues' work on the dimorphous expression of emotion (Aragόn, Clark, Dyer, & Bargh, 2015;Aragόn, 2017;Aragόn & Bargh, 2018), the match or mismatch between emotions expressed on a face and an emotional undertone of contexts play a role in observers' evaluation (Aragόn and Clark, 2017). When the two are matching (e.g., sad words with tearful faces), participants reported the intention to upregulate the expresser's emotional experience. ...
Article
Full-text available
Visible tears have been shown to enhance the perception of sadness. Whether the sadness perception from visible tears can occur automatically, which is essential for the rapid identification of emotional cues in real-life social interactions, is still unclear. We employed the reaction-time-based Implicit Association Test (IAT) to assess the implicit association of tears and sadness in two studies. Study 1 (N = 58) used sadness/non-sadness or negative/positive affect words as attribute pairs and images of tearless or tearful neutral expressions as targeted concepts. In Study 2 (N = 54), the neutral expressions were replaced with anger, disgust, fear, surprise, and happiness expressions with or without tears. Both studies revealed a strong tendency among participants to implicitly associate tears with sadness and negative affect. The results complemented findings from self-report measures by showing that the perception of sadness from visible tears can occur efficiently with little control.
... Such questions are addressed by scientific research (e.g. [2], [3]). They are highly interesting not only for humans to understand their own emotions but also for machines that need a model of humans' emotions to interact with them in a socially appropriate way. ...
Conference Paper
Full-text available
Computational emotion recognition focuses on observable expressions. In the case of highly unpleasant emotions that are rarely displayed openly and mostly unconsciously regulated -- such as shame -- this approach can be difficult. In previous studies, we found participants to smile and laugh while experiencing shame. Most current approaches interpret smiles and laughter as signals of enjoyment. They neglect the internal emotional experience and the complexity of social signals. We present a planned mixed-methods study that will investigate underlying functions of smiles and laughter in shameful situations and how those reflect in the morphology of expression. Participants' smiles and laughter during shame-eliciting situations will be analyzed using behavioral observations. Semi-structured interviews will investigate their functions. The gained knowledge can improve computational emotion recognition and avoid misinterpretations of smiles and laughter. In the scope of the open science initiative, we describe the planned study in detail with its research questions, hypotheses, design, methods, and analyses.
... Another possibility is that high-intensity scenarios actually encourage facial movements and other expressive behaviors that stereotypically correspond with other emotion categories (e.g., nose wrinkle in pleasure rather than disgust, crying in joy rather than sadness). These dimorphous expressions may be an effective means of communicating the intensity of emotion 71,72 , while at the same time consisting of facial movements that cannot be discriminated without additional context. ...
Article
Full-text available
It is long hypothesized that there is a reliable, specific mapping between certain emotional states and the facial movements that express those states. This hypothesis is often tested by asking untrained participants to pose the facial movements they believe they use to express emotions during generic scenarios. Here, we test this hypothesis using, as stimuli, photographs of facial configurations posed by professional actors in response to contextually-rich scenarios. The scenarios portrayed in the photographs were rated by a convenience sample of participants for the extent to which they evoked an instance of 13 emotion categories, and actors' facial poses were coded for their specific movements. Both unsupervised and supervised machine learning find that in these photographs, the actors portrayed emotional states with variable facial configurations; instances of only three emotion categories (fear, happiness, and surprise) were portrayed with moderate reliability and specificity. The photographs were separately rated by another sample of participants for the extent to which they portrayed an instance of the 13 emotion categories; they were rated when presented alone and when presented with their associated scenarios, revealing that emotion inferences by participants also vary in a context-sensitive manner. Together, these findings suggest that facial movements and perceptions of emotion vary by situation and transcend stereotypes of emotional expressions. Future research may build on these findings by incorporating dynamic stimuli rather than photographs and studying a broader range of cultural contexts.
Conference Paper
Full-text available
Sometimes happiness is displayed with smiles, but sometimes it is also displayed with expressions that look like sadness or aggression, called dimorphous expressions. Dimorphous expressions effectively communicate unique signals about expressers' motivations. This study demonstrates that experimentally manipulated dimorphous expressions alter consumers’ expectations of product characteristics and experiences.
Chapter
Full-text available
The belief that facial activity is linked to emotional states has a long history in science and in popular belief. In particular, the 1872 publication of Charles Darwin’s book The expression of the emotions in man and animals (1872/1998) has had a major impact on the role of emotional expressions in modern emotion research and has largely reinforced the commonly held belief that there is a link between emotions and expressive behavior. Milestone research, particularly by Paul Ekman and his colleagues, has cemented the common notion that faces express emotions. However, all of the involved researchers, including Darwin and Ekman, have usually maintained a view that is far more differentiated than the distilled version that has resulted from repeated summarizing and synthesizing in the secondary and tertiary literature. In fact, no current researcher holds that all emotions are always reflected in facial or other nonverbal activity. Inversely, almost all theorists will concur that often times facial expressions are not associated with a concordant affective state, or for that matter, any affective state at all.
Article
Psychologists, self-help gurus, and parents all work to make their clients, friends, and children happier. Recent research indicates that happiness is functional and generally leads to success. However, most people are already above neutral in happiness, which raises the question of whether higher levels of happiness facilitate more effective functioning than do lower levels. Our analyses of large survey data and longitudinal data show that people who experience the highest levels of happiness are the most successful in terms of close relationships and volunteer work, but that those who experience slightly lower levels of happiness are the most successful in terms of income, education, and political participation. Once people are moderately happy, the most effective level of happiness appears to depend on the specific outcomes used to define success, as well as the resources that are available. © 2007, Association for Psychological Science. All rights reserved.
Article
G*Power (Erdfelder, Faul, & Buchner, 1996) was designed as a general stand-alone power analysis program for statistical tests commonly used in social and behavioral research. G*Power 3 is a major extension of, and improvement over, the previous versions. It runs on widely used computer platforms (i.e., Windows XP, Windows Vista, and Mac OS X 10.4) and covers many different statistical tests of the t, F, and chi2 test families. In addition, it includes power analyses for z tests and some exact tests. G*Power 3 provides improved effect size calculators and graphic options, supports both distribution-based and design-based input modes, and offers all types of power analyses in which users might be interested. Like its predecessors, G*Power 3 is free.
Article
When we want to know what others are feeling, we look to the face for clues. However, individual differences matter: Some faces are more expressive than others. Do both emotion experience and dispositional expressivity predict emotion expression? Based on an analysis of display rules, the authors hypothesized that expressivity would moderate the relation between experience and expression for negative, but not for positive, emotion. Study 1 examined the relation between habitual emotion experience and peer-rated expressive behavior and showed the predicted moderator effect for negative emotion: Experience was related to expression only for dispositionally high-expressivity participants, not for low-expressivity participants. For positive emotion, however, experience was related to expression for both groups. Study 2 replicated these findings using momentary emotion experience and objectively coded expressive behavior during films that elicited amusement and sadness. Results are interpreted in terms of low-expressivity individuals' propensity to dynamically regulate negative emotion-expressive behavior.