2012. Vol.3, No.12A, 1183-1188
Published Online December 2012 in SciRes (
Copyright © 2012 SciRes. 1183
Individual Differences in Recognizing Spontaneous Emotional
Expressions: Their Implications for Positive Interpersonal
Sun-Mee Kang
Department of Psychology, California State University, Northridge, USA
Received September 29th, 2012; revised October 22nd, 2012; accepted November 17th, 2012
The main purpose of the current study was to introduce the Spontaneous Expressions Recognition Test
(SERT), a new thin-slice measure of emotion recognition for normative adults, and demonstrate its rela-
tive strengths for predicting positive interpersonal relationships. To explore this question, a semester-long
longitudinal study was conducted. In this study, college students were randomly assigned to small re-
search teams and worked together throughout the semester to conduct group research projects. Peer rat-
ings of interpersonal relationships were collected at the end of the semester. The results provided pre-
liminary support for the SERT, by demonstrating its relative strength for predicting positive interpersonal
Keywords: Emotion Recognition; Spontaneous Expressions; Interpersonal Relationships; Thin-Slice
Measure; Peer Ratings
Knowing others’ feelings is one of the core abilities that con-
tributes to maintaining positive interpersonal relationships
(Ambady, LaPlante, & Johnson, 2001; Hall, Andrzejewski, &
Yopchick, 2009; Hall, Bernieri, & Carney, 2005; Pickett,
Gardner, & Knowles, 2004). Individual differences in this core
ability have been assessed using a wide range of tests devel-
oped during the past several decades. A literature review re-
veals that there are quite a few self-reported measures, includ-
ing the Perceived Decoding Ability scale (Zuckerman & Lar-
rance, 1979), the Questionnaire Measure of Emotional Empathy
(Mehrabian & Epstein, 1972), and the Social Skills Inventory
(Riggio, 1986). Although they have been widely recognized
and used in psychological research, these self-reported mea-
sures have fundamental limitations, because they ask individu-
als to report their own level of emotion recognition ability. Since
people are not particularly good at judging their own abilities
(Ickes, 1993; Ickes et al., 2000), performance-based measures
thrive as a viable alternative to self-reported measures.
The long list of performance-based tests includes, although
not comprehensive, the Brief Affect Recognition Test (BART;
Ekman & Friesen, 1974), the Pictures of Facial Affects (POFA;
Ekman & Friesen, 1976), the Japanese and Caucasian Brief
Affect Recognition Test (JACBART; Masumoto et al., 2000),
the Diagnostic Analysis of Nonverbal Accuracy (DANVA;
Nowicki & Duke, 1994), the Profile of Nonverbal Sensitivity
(PONS; Rosenthal, Hall, DiMatteo, Rogers, & Archer, 1979)
test, the Child and Adolescent Social Perception measure
(CASP; Magill-Evans, Koning, Cameron-Sadava, & Manyk,
1995), The Awareness of Social Inference Test (TASIT;
McDonald, Flanagan, & Rollings, 2003), and the Multimodal
Emotion Recognition Test (MERT; Bänziger, Grandjean, &
Scherer, 2009).
An intensive review of the existing tests revealed that sur-
prisingly, few measures adopted spontaneous expressions as
test stimuli. In most cases, the emotional expressions of the
target people were determined by the researchers who devel-
oped the tests, rather than by the targets who expressed the
emotions. Thus, the targets were instructed either to move their
facial muscles to create certain facial expressions (e.g., the
BART, the POFA, and the JACBART) or to express emotions
based on given scripts (e.g., the PONS, the DANVA, the CASP,
the TASIT, and th e MERT).
Using posed expressions in emotion recognition tests has a
number of merits. This approach gives researchers control over
certain features of their test, including the range of the emotions
assessed and the degree of intensity of emotions expressed
(Bänziger, Grandjean, & Scherer, 2009). The downside is that
the ecological validity of the tests developed under this ap-
proach is somewhat compromised, because posed expressions
tend to be less natural than spontaneous expressions (Russell,
1994; Trimboli & Walker, 1993).
In sum, the literature review shows that very few measures of
emotion recognition that employ spontaneous emotional ex-
pressions are currently available for normative adult popula-
tions. The main purpose of the current study was to developing
a new test to fill this gap and to demonstrate the relative strengths
of the new test compared with several of existing emotion rec-
ognition tests .
The Spontaneous Expressions Recognition Test:
A New Measure for Normative Adults
A critical issue in the development of an emotion recognition
test is determining what the target person in the test question is
feeling (Mayer & Geher, 1996; Trimboli & Walker, 1993).
Mayer and Geher (1996) listed three criteria for determining the
target’s emotions: target agreement, consensus agreement, and
expert judgment. Since only the individual experiencing an
emotion has direct access to it (Mayer & Geher, 1996; Nisbett
& Wilson, 1977), it seems justified to use the target’s self-re-
ported emotions as the correct answer.
In this vein, the empathy accuracy paradigm developed by
Ickes and his colleagues (Ickes, Bissonnette, Garcia, & Stinson,
1990) provides one way to obtain the correct answers for emo-
tional experiences of the target person. Ickes and his colleagues
created an experimental setting in which two strangers interact
with each other for a short time (e.g., 6 minutes) and this ses-
sion is videotaped by a hidden camera. Following the taping,
the two participants are instructed to watch the videotape alone
in separate rooms, stopping the tape at points where they re-
member having had a specific thought or feeling. They are, then,
asked to review the videotape a second time, and this time the
video is stopped at the points at which their interaction partner
reported a thought or feeling. The participants’ task is to infer
the thoughts or feelings of their interaction partner. These in-
ferences are compared with the actual thoughts and feelings to
assess empathic accuracy.
Although this paradigm was originally developed to measure
empathic accuracy between two strangers involved in a social
scene, it has been widely applied to other settings beyond dy-
adic interaction (Barone, Hutchings, Kimmel, Traub, Cooper, &
Marshall, 2005; Klein & Hodge, 2001; Marangoni, Garcia,
Ickes, & Teng, 1995; Pickett, Gardner, & Knowles, 2004). A
new measure of emotion recognition, the Spontaneous Ex-
pressions Recognition Test (SERT), was developed by adopting
this empathy accuracy paradigm in the current study.
Possible Advantages of Using Spontaneous
Expressions as Study Stimuli
Given the plethora of existing measures of emotion recogni-
tion, it is essential to explain why another emotion recognition
test needed to be developed. As discussed before, the SERT is
one of the few tests for normative adults that uses spontaneous
emotional expressions. One advantage of using spontaneous
emotional expressions as study stimuli, instead of posed emo-
tional expressions, is that doing so promotes the ecological
validity of the test. Furthermore, the spontaneous emotional ex-
pressions in the SERT are presented as “thinslices” (Ambady,
LaPlante, & Johnson, 2001). A thin-slice is a short excerpt,
usually less than 5 minutes long, extracted from a behavioral
stream (Ambady, LaPlante, & Johnson, 2001).
A merit of the thin-slice measures, compared to still photo
tests, is that they provide richer information on expressed emo-
tions. They not only provide some context for the presented
emotional expressions, but also allow changes in expression to
be charted through a stream of behavioral sequences over time.
Since the format (short video clips) and the content (spontane-
ous expressions) of the SERT more closely resemble natural
interactions in which people read others’ emotions, these fea-
tures should contribute to the greater ecological validity of the
Previous studies have shown that individual differences in
emotion recognition are associated with positive interpersonal
adjustment (Ambady, LaPlante, & Johnson, 2001; Hall, An-
drzejewski, & Yopchick, 2009; Pickett, Gardner, & Knowles,
2004). If an individual recognizes what an interaction partner is
feeling during social interactions, that individual is more likely
to respond appropriately to the interaction partner (Ambady,
LaPlante, & Johnson, 2001). For this reason, someone better at
emotion recognition is more likely to maintain positive inter-
personal relationships with others. If the SERT has greater
ecological validity than other tests, it should also display rela-
tive strength in predicting positive interpersonal relationships
when compared with those emotion recognition measures.
To test this question, a semester-long longitudinal study was
conducted. In this study, college students were randomly as-
signed to small research teams that worked together through-
out the semester conducting group research projects. This en-
abled the team members to serve as qualified peer raters of
interpersonal relationships of each team member at the end of
the semester. The present study was conducted to investigate
the relative strength of the SERT, compared with several exist-
ing measures, in predicting positive interpersonal relationships.
Ninety-six undergraduate students enrolled in the “Research
Methods in Psychology” class at a large state university on the
West Coast served as the participants in this project. All par-
ticipants ranged in age from 18 to 29 years, with a mean age of
22.66 years (SD = 2.06). The majority of the students (82.3%)
were women.
The SERT. In the first phase, an interview study was con-
ducted to develop a pool of stimulus video clips. Twenty-six
European American college students (18 women and 8 men)
were interviewed in individual 25-minute sessions in which
they discussed their meaningful life experiences in a free-for-
mat interview. The interview sessions were videotaped with the
participants’ consent. Immediately after the interview, each
interviewee was asked to review her or his own tape alone and
was instructed to stop the tape at any point during which the
interviewee remembered having had a specific emotion and
record how she or he felt at that moment of the interview, al ong
with the time stamp corresponding to the segment. After the
review, interviewees were asked for their consent for using
portions of the tape to develop a stimulus video for future stud-
ies. The interviewees were not told about this request until the
review sessions were done in order to allow the interviewees to
behave more naturally during the interview. Three interviewees
refused to give their permission, and their interview tapes were
immediately erased.
To extract the marked segments from the interview tapes, the
interview tapes were edited using Adobe Premier 6.0. The re-
ported time stamp served as the ending point of each clip, and
the starting point was chosen on the basis of two considerations:
1) providing decoders with the smallest amount of information
sufficient for identifying the emotional experience of the inter-
viewee and/or 2) respecting the natural transition points of the
dialogues. In order to avoid later complications in calculating
recognition accuracy scores, only segments associated with a
single reported emotion were selected.
All the extracted segments were initially reviewed by the au-
thor. An issue that emerged during this initial review was that it
was too challenging, in many cases, for examinees to identify
the emotions reported by the interviewees, because there were
Copyright © 2012 SciRes.
insufficient clues available for the examinees to read the emo-
tions. In other words, in some clips, only the experiencing indi-
viduals (i.e., the interviewees) were able to identify the emo-
tions, because only they had direct access to the internal feel-
ings. Thus, it was necessary to check whether viewers would be
able to read the reported feelings based on the verbal and non-
verbal information given in the clip. Two trained undergraduate
students independently reviewed the extracted clips to evaluate
them. The judges verified that each selected segment matched
the interviewee’s written comments concerning what emotion
she or he experienced at the end point of the clip.
A series of small-scale pilot studies were conducted with a
pool of clips verified by both judges. Clips that were too diffi-
cult (below 20% accuracy scores) or too easy (above 80%)
were eliminated from the test to give the final version of the
SERT an optimal range of test difficulty (Trimboli & Walker,
1993). The exceptions for this selection criterion were any clips
describing a happy emotion. All the clips describing happy
emotions were quite easy to read (i.e., they had accuracy scores
above 80%). Despite this, happy clips were included in the final
version of the SERT in order to make the SERT cover both
positive and negative emotions. Nine clips from 9 senders (5
women and 4 men) were included in the final version of the
stimulus video tape.
Table 1 displays the item properties of the SERT for each
clip: the emotion described, the gender of the sender who de-
scribed it, the length of the clip (ranging from 12.10 to 30.05
seconds, with a mean length of 18.79 seconds, and a standard
deviation of 5.43 seconds), and the mean and standard devia-
tion of the accuracy rate for each item. The accuracy rates were
obtained from a large-scale pilot study with 260 college stu-
dents (68% women). The mean accuracies for the 9 clips
ranged from 50% to 91%, with an overall mean accuracy of
71%. Although the internal consistency of the SERT was not
high (.58) in the pilot study, it was acceptable given that the
SERT consists of only 9 items. In fact, a low internal consis-
tency is a hallmark of successful nonverbal decoding tests in
which emotions are expressed through multiple nonverbal
channels (Hall, 2001).
The selected clips were presented to test takers through a
computerized test platform that was developed by the author
using Visual Basic. The final version of the SERT followed
each clip with the question, “What emotion is this person ex-
periencing at this moment?” The examinees were asked to re-
spond by choosing one of five options—angry, anxious, frus-
trated, happy, or sad. They were given 5 seconds to do so.
Anxiety was included among the response choices because it
was the answer to one of the three practice questions. The
SERT takes about 15 minutes to complete.
The Japanese and Caucasian Facial Expressions of Emo-
tion (JACFEE) Test. A computerized emotion recognition test
based on the Japanese and Caucasian Facial Expressions of
Emotion (JACFEE; Matsumoto & Ekman, 1988) was devel-
oped by the author using Visual Basic. The test presented static
photos from the JACFEE to the participants and recorded their
responses to those stimulus photos (called the “JACFEE test,”
hereafter). The original JACFEE consisted of 28 Caucasian and
28 Asian faces, but, in order to avoid any confounds created by
a mismatch between the SERT and the JACFEE test in the
ethnic background of encoders, only Caucasian photos were
selected for the current study. The selected photos covered five
basic emotions—anger, disgust, happiness, sadness, and sur-
Table 1.
Item properties of the spontaneous expressions recognitio n t es t.
Emotion Gender of
Senderb Clip Length
1 FrustratedM 30.05 68 47
2 Happy M 24.02 87 34
3 Sad W 15.25 74 44
4 Happy M 15.12 91 19
5 Angry W 16.27 67 47
6 Sad W 12.10 72 45
7 Angry W 19.09 57 50
8 FrustratedM 20.14 76 43
9 FrustratedW 17.03 50 41
Note: aNo stands for the test item number of stimulus clips in the SERT; bSender
refers a person who expresses her/his emotion in each clip; cThe mean accuracies
and standard deviations for the nine clips were obtained through a large-scale
pilot study with 260 college stud ents.
prise1. Twenty pictures (5 emotions × 4 encoders per emotion)
were presented in a random order. No face is repeated in the
JACFEE test. Each was displayed on a computer screen for 1
second, with a 4-second interval between pictures.2 It took 5
minutes to complete the JACFEE test. The alpha coefficient of
the JACFEE test was .57 in the current study.
The Profile of Nonverbal Sensiti vity (PONS) Test. This te st
was included because it is a widely-used thin-slice measures of
emotion recognition. The original PONS test (Rosenthal, Hall,
DiMatteo, Rogers, & Archer, 1979) consists of a film with 220
two-second audio and/or visual segments that represent par-
ticular emotional responses. In each segment, a young Cauca-
sian woman portrays 1 of 20 different situations representing
four different affective domains including positive-submissive
(e.g., helping a customer), positive-dominant (e.g, talking about
one’s wedding), negative-submissive (e.g., asking for forgive-
ness), and negative-dominant (e.g., nagging a child) quadrants.
Each 2-second segment is followed by a 5-second interval,
during which the examinee selects which of two choices better
describes the scene. For example, in one scene (number #11),
the choices offered to the examinee are: 1) talking to a lost
child, and 2) helping a customer. Due to the time constraints of
this study, only the first half of the PONS test was administered
1Since both the SERT and the JACFEE test share the same test platform
developed with Visual Basic, the two tests were administered back to back.
Under this testin g c onditi on, the n um ber of c hoic es pe r question was se t to be
the same across the two tests in order to avoid any confounds that the changes
in the number of alternative s may bring. Thus, alt hough the original JACFEE
covers the seven basic emotions—including anger, disgust, happiness, sad-
ness, surprise, fear, and contempt—in the current study only the first five
emotions were selected, i n order to meet this constraint.
2The exposure time of the stimulus pic tures (1 second) w ith 4 second interval
was determ ined base d on the de scriptio ns on the developm ent of Japa nese an d
Caucasian Brief Affect Recognition Test (JACBART). In the JACBART,
each stimulus video clip was created by em
edding a JACF EE expressio n (for
1/5 second) in the middle of a one-second presentation of that poser’s neural
expression. Due to technical difficulties with programming the JACFEE test
in Visual Basic, the same presentation mode was not implemented in the
current study. Instead, all the JACFEE expressions were presented for one
second with a 4-second interval, because the results of a small-scale pilot
study showed that the 1/5 second exposure of a JACFEE expression alone in
the Visual Basic platform was too fast for examinees to identify the emotions.
Thus, all the JACFEE expressio ns were presente d for one second.
Copyright © 2012 SciRes. 1185
(i.e., 110 items). This first-half version of the PONS test has
been successfully used in a previous study (Horan, Kern, Sho-
kat-Fadai, Sergi, Wynn, & Green, 2009). The alpha coefficient
of this version of the PONS test was .68 in the current study.3
The participants took the PONS test using an individual
computer. The participants were instructed to mark one of the
two choices for each question on a response sheet provided at
the beginning of the test. It took about 30 minutes to complete
this short version of the test on average.
Perceived Social Support. The 7-item Social Support sub-
scale taken from the Quality of Relationships Inventory (QRI-
SS; Pierce, Sarason, & Saraon, 1991) was selected as a measure
of positive interpersonal relationships. The QRI-SS subscale
measures the perceived availability of social support from a
particular person (e.g., “To what extent can you count on this
person to listen to you when you are very angry at someone
else?” or “To what extent can you count on this person for help
with a problem?”). This particular scale was chosen to fit the
way in which peer ratings were collected in this study. Partici-
pants were asked to rate their teammates at the end of the se-
mester. If a student maintained good interpersonal relationships
with teammates throughout the semester, those teammates
would evaluate that student as someone on whom they could
rely for social support. Participants were asked to rate each
team member using a 5-point scale, from 1 (not at all) to 5
(very), on each the 7 items of the QRI-SS. The alpha coefficient
of the QRI-SS test was .95 in the current study.
During the first week of the semester, students were in-
formed that they were expected to participate in a series of
studies as part of their course credit. Students were also offered
alternatives to participating in this study. During the following
week, the students took the SERT, the JACFEE test, and the
PONS test in small-group settings (3 people per session) at a
research laboratory. Each student used an individual computer.
In the third week, research teams were formed by random as-
signment. Each team consisted of a minimum of 4 students
because peer ratings on perceived social support from this
round-robin design needed to be analyzed with the Social Rela-
tionship Model (SRM; Kenny, 1994). To estimate the parame-
ters of the SRM (target effect, perceiver effect, and relationship
effect), it is recommended that all groups have at least 4 people
(Kenny, 1994). In the 12th week, the students took the SERT
again to check its test-retest reliability. At the end of the se-
mester (14th week), the 7-item QRI-SS subscale was distrib-
uted to the students in sealed envelopes. Each student was
asked to rate all other team members individually on the en-
closed questionnaires and to drop the sealed envelope to a
locked drop box outside of the author’s laboratory within 1
A goal of this study was to examine how well the SERT, the
JACFEE test, and the PONS test predict the maintenance of
positive interpersonal relationships. To this end, the indicator of
the positive interpersonal relationships was computed first. The
peer-rating scores from the QRI-SS subscale were analyzed
with the computer program SOREMO (Kenny, 1998) to esti-
mate the target effect score of each participant.
There is compelling reason for applying Social Relations
Model (SRM) analysis to the peer-rating scores (Kenny, 1994):
The rating scores from the round-robin design are not statistic-
cally independent, since all the team members were asked to
rate one another (Kenny, 1994). Thus, rating scores could have
been influenced by the unique relationships between targets and
raters (the “relationship effect”). The rating scores could also
have been influenced by the raters’ response styles (the “per-
ceiver effect”). By using SRM analysis, researchers can esti-
mate the target effect while controlling for both the perceiver
and the relationship effects. Thus, the target effect scores esti-
mated by the SOREMO (Kenny, 1998), were used as the major
indicator of peer-rated interpersonal relationships.
Next, the intercorrelations among the four major variables
were examined. As shown in Table 2, the SERT was positively
and significantly associated with peer-rated interpersonal rela-
tionships (r = .33). Like the SERT, the PONS test had a posi-
tive association with the peer-rated interpersonal relationships
(r = .19). Interestingly, the JACFEE test has a negative, al-
though weak, association with the peer-rated interpersonal rela-
tionships (r = –.09). The SERT was modestly associated with
the PONS test (r = .20), but not with the JACFEE test (r = .02).
The PONS test and the JACFEE test were modestly associated
with each other (r = .20).
Multiple regression analyses were performed in order to
explore the relative contributions of the SERT, the JACFEE
test, and PONS test to the outcome measure. The measure of
perceived social support by peers (estimated target score) served
as the outcome variable, and the SERT, the JACFEE test, and
the PONS test were entered into the regression analyses as the
main predict o rs.
Table 2 presents the results of the regression analyses, in-
cluding standardized regression coefficients of those predictors,
their associated t values, squared semi partial correlations (sr2)
as an indicator of the effect size for each predictor, and the
squared R to evaluate the effect size for the overall regression
model. The results of the regression analyses revealed that only
Table 2.
Regression coefficients of predictors accounting for variance in per-
ceived social support by peers.
Perceived Social Support by Peers (Target Effect)
Predictors r t sr2
SERT .33 .31 3.11 .09
JACFEE –.09 –.13 –1.32 .02
PONS .19 .15 1.52 .02
R2 .14
3The internal consistencies of the 110-item PONS test (.68) and the 20-item
JACFEE test (.57) were not high in this study, mainly because only portions
of the original tests were employed due to time constraints. When the inter-
nal consistencies of the full-length PONS and JACFEE tests were estimated
using the Spearman Brown prophecy formula, the estimated reliabilities
were .81 (PONS test) and .79 (JACFEE test), which were similar to the
reported reliabilities for the 220-item PONS (about .86; Hall, 2001) and for
the 56-item Japanese and Caucasian Brief Affect Recognition Test
(JACBART; from .82 to .92; Matsumoto et al., 2000). Because the
JACBART test is not identical to the JACFEE test, the internal consistency
of the 56-item JACBART should be considered as a proxy estimate for the
JACFEE test. Note: Coefficients with an absolute t-value greater than 2.0 are considered sig-
nificant at the .05 level according to a two-tailed test (n = 96).
Copyright © 2012 SciRes.
the SERT emerged as a statistically significant predictor for
peer-rated interpersonal relationships ( = .31, t = 3.11, sr2
= .09). Both the JACFEE test ( = –.13, t = –1.32, sr2 = .02)
and the PONS test ( = .15, t = 1.52, sr2 = .02) were not sig-
nificant predictors for the outcome variable.
The total proportions of variance in perceived social support
explained by the current regression model were .14. Finally, the
test-retest reliability of the SERT for the 10-week period
was .72 in this study.
The main purpose of the current study was to introduce an
ecologically valid measure of emotion recognition for norma-
tive adults and to demonstrate its relative strengths for predict-
ing positive interpersonal relationships. Overall, the results of
the current study provide preliminary support for the construct
validity of the SERT. The SERT scores correlated weakly to
moderately with the JACFEE and PONS test scores and were
stable over a 10-week period. The SERT also demonstrated its
relative strength in predicting the peer ratings of interpersonal
relationships over the JACFEE and PONS tests.
One of the strengths of the current study is that peer ratings
of interpersonal relationships, which served as the main out-
come variable, were collected in a longitudinal study using a
round-robin design. Using this variable, the current study
showed that the SERT was meaningfully associated with peer-
rated interpersonal relationships. Compared to the SERT, the
JACFEE and PONS tests did not account for meaningful por-
tions of the variance in the outcome. This finding was some-
what unexpected, because the PONS test in particular has been
associated with the identification of interpersonal sensitivity
and positive interpersonal relationships (Bernieri, 1991; Di-
Matteo, Friedman, & Taranta, 1979; Rose n t hal et al., 1979).
Why Does the SERT Better Predict Peer
The relative predictive power of the SERT over the JACFEE
and PONS tests can be partially accounted for by its ecological
validity (Ambady, LaPlante, & Johnson, 2001). One of the
main features of the SERT is that it employs spontaneous ex-
pressions, whereas the JACFEE and the PONS tests (along with
a majority of existing emotion recognition tests) employ posed
emotions. Furthermore, the spontaneous expression clips of the
SERT are presented through a full video/audio channel and
average 18.79 seconds in length. These unique features of the
SERT contribute considerably to its ecological validity, be-
cause inferring the emotions presented in the SERT clips bears
a strong resemblance to inferring others’ feelings in everyday
life. Thus, it can be predicted that students with high scores on
the SERT will be comparatively good at reading their team-
mates’ emotions in daily life.
The relative predictive power of the SERT can be further ex-
plained by the close match between the skills required for the
SERT and the outcome measure of this study. The outcome
measure was the QRI-SS subscale, which assesses the per-
ceived availability of social support from a particular person.
The skills measured by the SERT (i.e., skills for reading emo-
tions accurately from a stream of on-going verbal and nonver-
bal behaviors) are the same skills that assisted the students in
assessing the needs of their teammates. If they judged the needs
of their teammates accurately, they were more likely to adjust
their behaviors in a desirable way, which subsequently led their
teammates to perceive them as available for social support and
to rate them more highly on this measure.
The relatively weak performance of the JACFEE and the
PONS tests in accounting for peer-rated interpersonal relation-
ships implies that the skills measured by the JACFEE test (i.e.,
skills for reading posed emotions accurately within 1-second
exposure) and the PONS test (i.e., skills for reading social
situations correctly with limited verbal/nonverbal expressions)
do not correlate with the particular outcome variable used in
this study (i.e., perceived availability of social support). The
JACFEE and PONS tests may still emerge as primary predic-
tors for other types of outcome variables that have a close
match with the particular skills measured by the JACFEE and
PONS tests.
In sum, the enhanced ecological validity of the SERT and the
close match between its required skills and the specific out-
come measure may explain the greater predictive power of the
SERT over the other measures of emotion recognition used in
this study.
Although the results of the current study highlight the im-
portance of the SERT as a significant predictor of peer-rated
interpersonal relationships, they should be interpreted with
caution due to a number of limitations. First, the participants of
the current study were predominantly women. Second, the peer
ratings of interpersonal relationships were collected under the
unique conditions of the current study, in which the participants
formed research teams and worked together on a group project
throughout an entire semester. In this setting, an individual’s
performance on the group project could unduly impact peer
ratings. The significance of the peer ratings obtained may there-
fore be limite d to work-re lat ed se tt ings. Due t o these li mit a tions,
the apparently strong contribution of the SERT to peer-rated
interpersonal relationships needs to be interpreted with caution
and should be replicated with a large gender-balanced sample
in a non work-related setting.
Closing Remarks
The results from the current study show sufficient promise to
warrant further development of the SERT. Even though it is
brief (only 9 items), it seems to be comparable with existing
measures of emotion recognition. Whether this measure can be
applied to other settings and to other samples beyond college
students needs to be further investigated.
Preparation of this article was supported by a grant from the
NIMH and NIGMS (no. SC2MH087466).
Ambady, N., LaPlante, D., & Johnson, E. (2001). Thin-slice judgments
as a measure of interpersonal sensitivity. In J. A. Hall, & F. J. Berni-
eri (Eds.), Interpersonal sensitivity: Theory and measurement (pp.
89-101). Mahwah , NJ: Lawrence Erlbaum Associates, Inc.
Bänziger, T., Grandjean, D., & Scherer, K. R. (2009). Emotion recogni-
tion from expressions in face, voice, and body: The Multimodal
Copyright © 2012 SciRes. 1187
Copyright © 2012 SciRes.
Emotion Recognition Test (MERT). Emotion, 9, 691-704.
Bernieri, F. J. (1991). Interpersonal sensitivity in teaching interactions.
Personality and Social Psychology Bulletin, 17, 98-103.
Bernieri, F. J. (2001). Toward a taxonomy of interpersonal sensitivity.
In J. A. Hall, & F. J. Bernieri (Eds.), Interpersonal sensitivity: The-
ory and measurement (pp. 3-20). Mahwah, NJ: Lawrence Erlbaum
Buck, R. (1984). The communication of emotion. New York: Guildford
Cohen, J., & Cohen, P. (1975). Applied multiple regression/correlation
analysis for the behavioral sciences. Oxford, England: Lawrence
Crocker, L., & Algina, J. (1986). Introduction to classical and modern
test theory. Orland, FL: Harcourt B race Jovanovich, Inc.
Davis, M. H. (1983). Measuring individual differences in empathy:
Evidence for a multidimensional approach. Journal of Personality
and Social Psychology, 44, 113-126.
DiMatteo, M. R., Friedman, H. S., & Taranta, A. (1979). Sensitivity to
bodily nonverbal communication as a factor in practitioner-patient
rapport. Journal of Nonverbal Behavior, 4, 18-26.
Ekman, P., & Friesen, W. V. (1974). Detecting deception from the
body or face. Journal of Personality and Social Psychology, 29, 288-
298. doi:10.1037/h0036006
Ekman, P., & Friesen, W. V. (1976) . Pictures of facial affect. Palo Alto,
CA: Consulting Psych o l og i st s ’ Press.
Hall, J. A. (2001). The PONS test and the psychometric approach to
measuring interpersonal sensitivity. In J. A. Hall, & F. J. Bernieri
(Eds.), Interpersonal sensitivity: Theory and measurement (pp. 143-
161). Mahwah, NJ: Lawrence Erlbaum Associates.
Hall, J. A., Andrzejewski, S. A., Yopchick, J. E. (2009). Psychosocial
correlates of interpersonal sensitivity: A meta-analysis. Journal of
Nonverbal Behavior, 33, 149-180. doi:10.1007/s10919-009-0070-5
Hall, J. A., Bernieri, F. J., & Carney, D. R. (2005). Nonverbal behavior
and interpersonal sensitivity. In J. A. Harrigan, R. Rosenthal, & K. R.
Scherer (Eds.), The new handbook of methods in nonverbal behavior
research (pp. 237-281). Oxford: Oxford University Press.
Horan, W. P., Kern, R. S., Shokat-Fadai, K., Sergi, M. J., Wynn, J. K.,
& Green, M. F. (2009). Social cognitive skills training in schizo-
phrenia: An initial efficacy study of stabilized outpatients. Schizo-
phrenia Research, 107, 47-54. doi:10.1016/j.schres.2008.09.006
Hunsley, J., & Meyer, G. J. (2003). The incremental validity of psy-
chological testing and assessment: Conceptual, methodological, and
statistical issues. Psychological Assessment, 15, 446-455.
Ickes, W. (1993) Empathic accuracy. Journal of Personality, 61, 587-
610. doi:10.1111/j.1467-6494.1993.tb00783.x
Ickes, W., Bissonnette, V., Garcia, S., & Stinson, L. (1990). Imple-
menting and using the dyadic interaction paradigm. In C. Hendrick,
& M. Clark (Eds.), Review of personality and social psychology:
Volume 11, research methods in personality and social psychology
(pp. 16-44). Newbury Park, CA: Sage.
Ickes, W., Buysse, A., Pham, H., Rivers, K., Erickson, J. R., Hancock,
M., Kelleher, J., & Gesn, P. (2000). On the difficulty of distinguish-
ing “good” and “poor” perceivers: A social relations analysis of em-
pathic accuracy data. Personal Relationships, 7, 219-234.
Kenny, D. A. (1994). Interpersonal perception: A social relations
analysis. New York: Guilford Press.
Kenney, D. A. (1998). SOREMO version V. 2; A FORTRAN program
for the analysis of round-robin data structures. Unpublished manu-
script, Storrs-Mansfield, CT: University of Connecticut.
Magill-Evans, J., Koning, C., Cameron-Sadava, A., & Manyk, K.
(1995). The child and adolescent social perception measure. Journal
of Nonverbal Behavior, 19, 151-169. doi:10.1007/BF02175502
Marangoni, C., Garcia, S., Ickes, W., & Teng, G. (1995). Empathic
accuracy in a clinically relevant setting. Journal of Personality and
Social Psychology, 68, 854-869. doi:10.1037/0022-3514.68.5.854
Matsumoto, D., & Ekman, P. (1988). Japanese and Caucasian Facial
Expressions of Emotion (JACFEE). San Francisco, CA: Intercultural
and Emotion Research Laboratory, Department of Psychology, San
Francisco State University.
Matsumoto, D., LeRoux, J., Wilson-Cohn, C., Raroque, J., Kooken, K.,
Ekman, P., Yrizarry, N., Loewinger, S., Uchida, H., Yee, A., Amo,
L., & Goti, A. (2000). A new test to measure emotion recognition
ability: Matsumoto and Ekman’s Japanese and Caucasian Brief Af-
fect Recognition Test (JACBART). Journal of Nonverbal Behavior,
24, 179-209. doi:10.1023/A:1006668120583
Mayer, J. D., & Geher, G. (1996). Emotional intelligence and the iden-
tification of emotion. Intelligence, 22, 89-113.
Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In
P. Salovey, & D. J. Sluyter (Eds.), Emotional development and emo-
tional intelligence: Educational implications (pp. 3-31). New York,
NY. Basic Books.
McDonald, S., Flanagan, S, Rollins, J., & Kinch, J. (2003). A new
clinical tool for assessing social perception after traumatic brain
injury. Journal of Head Trauma Rehabilitation, 18, 219-238.
Mehrabian, A., & Epstein, N. (1972). A measure of emotional empathy.
Journal of Personalit y, 40, 525-543.
Nowicki, S., & Duke, M. P. (1994). Individual differences in the non-
verbal communication of affect: The Diagnostic Analysis of Non-
verbal Accuracy Scale. Journal of Nonverbal Behavior, 18, 9-25.
Paulhus, D. L. (1984). Two-component models of socially desirable
responding. Journal of Personality and Social Psychology, 46, 598-
609. doi:10.1037/0022-3514.46.3.598
Pickett, C. L., Gardner, W. L., & Knowles, M. (2004). Getting a cue:
The need to belong and enhanced sensitivity to social cues. Person-
ality and Social Psychology Bulletin, 30, 1095-1107.
Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1991). General and
relationship-based perceptions of social support: Are two constructs
better than one? Journal of Personality and Social Psychology, 61,
1028-1039. doi:10.1037/0022-3514.61.6.1028
Riggio, R. E. (1986). Assessment of basic social skills. Journal of Per-
sonality and Social P sychology, 51, 649-660.
Rosenthal, R., Hall, J. A., Di Matteo, M. R ., Rog ers, P. L., & Arch er, D.
(1979). Sensitivity to NONVERBAL COMmunication: The PONS
Test. Baltimore, MD: Johns Hopkins University Press.