JOURNAL OF COMPARATIVE RESEARCH IN
ANTHROPOLOGY AND SOCIOLOGY
Copyright © The Author(s), 2010
Volume 1(1): 57-72
ISSN 2068 – 0317
http://compaso.ro
Gender differences in impression formation
Bogdana Humă1
Abstract
This paper aims to highlight the differences between men and women regarding impression
formation. It is based on secondary analysis of the data gathered in two previous experiments with
similar conditions. However, the hypotheses formulated within this study have not been tested
before. The current analysis was conducted on 86 participants, 47 males and 39 females. Their ages
ranged between 15 and 32, as they were either high school or university students engaged in a
master’s program. Their task consisted of watching a 14 seconds long video of a female confederate
reading a neutral text and then evaluating her using a semantic differential with four dimensions:
sociability, ethics, power and activity. Based on previous studies, it was hypothesized that men and
women will form different first impressions of the actor employed in the movie. More precisely, the
majority of the studies undertaken in this area compare men and women’s accuracy scores of facial
expressions decoding, yielding mostly significant differences, with women achieving higher
accuracy. A small percentage has addressed other aspects of social perception like: personality traits
or socio-demographic characteristics, yielding similar results. However, the current experiment
failed to reveal any differences between men’s and women’s evaluations. Accuracy assessments
were disregarded in this study, since establishing unequivocal criteria for personality traits
evaluation is yet to be achieved. The results are consistent with a small percentage of the studies
conducted on gender differences in social perception and allow multiple interpretations.
Keywords
Gender differences, impression formation, social perception, nonverbal behavior, semantic
differential
1
Faculty of Sociology and Social Work, University of Bucharest, Romania, bogdana_huma@hotmail.com
57
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
Introduction
It is a common sociological practice to resort to gender differences when trying to
explain social phenomena. Whether it is religiosity, cultural consumption or political
opinions and preferences, social scientists draw on innate or acquired differences
between men and women’s actions or cognitions, revealing dissimilar or even opposed
patterns of behavior in regard to the concepts under investigation. Therefore, even if the
assertion that men are from Mars and women are from Venus is rather farfetched, not
admitting that gender differences exist and that they account for at least a small part of
the individual behavior variance would be a mistake.
This paper’s aim is to investigate such variations in the process of impression
formation. Until now, many studies have addressed this issue, most of them yielding
similar results, which revealed significant differences between the two sexes. However,
there are a few papers which contradict these findings, consequently rendering this topic
to uncertainty and speculations. Moreover, it must be mentioned that the largest part of
the studies in this area addressed the issue of gender differences in impression accuracy.
However, the accuracy of person perception is currently a debatable subject, research on
this topic having afforded sometimes convergent and other times divergent results, none
of which having been explained by existing theories. Therefore, comparing men and
women’s accuracy might not be the best method for pointing out gender differences in
impression formation.
Firstly, the terms which will be further used throughout this paper will be defined
and explained. Regarding the sex – gender controversy, I agree with the authors of the
anthology Sex Differences. Summarizing more than a Century of Scientific Research (Ellis et.
al., 2008) who point out that, in order to establish whether a characteristic appertains
the sex or the gender trait, it is necessary to establish if it is an inborn or an acquired
attribute. Therefore, since it is rather difficult to ascertain this last distinction, using the
terms as synonyms seems like an acceptable practice. Moreover, as recent studies have
shown, many if not all behavioral differences between men and women, might be caused
by biological as well as social factors (ibidem).
Social perception accuracy could be generally defined as the correspondence
between observers’ assessments of a characteristic of the target person and an external
criterion measuring the same concept. Self-ratings, evaluations of family or friends or
judgments provided by experts are usually employed. However, some attributes, such as
sex, age, ethnicity or profession can be objectively appraised.
According to Zebrowitz (1995/1999, 309-310) impression formation represents ‘the
process of forming descriptive and evaluative judgments about a target person’, while an
impression is defined by Hamilton, Katz and Leirer (1980, 1050-1051) as ‘a perceiver’s
cognitive representation of another person’.
58
Bogdana Humă / Gender differences in impression formation
Gender differences in empathic accuracy
As mentioned before, most studies regarding gender differences in social perception
addressed the subject from the viewpoint of male and female empathic accuracy, the
majority of them focusing on facial expressions of emotions.
In their study Kirouac and Dore (1985) focused on the issue of emotion recognition
accuracy as a function of gender and education. They employed high school, college and
university students which were shown pictures of the six fundamental emotions. The
statistical analysis of the results indicated a significant difference between men and
women in regard to decoding facial expressions of emotions. Nevertheless, further
analysis showed that gender accounted for only 2.83% of the data variance, whereas a
factor based on the depicted emotion explained 62.44% of the total variance.
A study conducted by Hall and Matsumoto (2004) employed seven point scales for
recording the participants’ answers. The stimuli consisted of facial expressions of
emotions shown for a very short period of time (10 s or 20 s). The results indicated that
women were more accurate than men in indentifying the depicted emotions. Using a
multiple-point scale, the authors were able to spot other differences between male and
female subjects. It seems that women’s answers had a wider range of variance than
men’s. Similar results were obtained by Katsikitis, Pilowsky and Innes (1997). The
researchers’ explanations drew upon the mechanisms of nonverbal communication
decoding, suggesting that, whereas men use a prototypical image in order to identify
emotions, women pay attention to different cues, analyzing them one by one. A second
explanation puts forward the fact that female participants might have been more
confident in their evaluations, thus venturing in selecting the scale’s extreme values as
well as the middle ones.
Thayer and Johnsen (2000) investigated perception accuracy using photographs
selected form Ekman and Friesen’s standardized facial expression set. Instead of asking
the participants to name the depicted emotion, they inquired about the subjects’ own
affective state, knowing that emotions are highly contagious. The results showed that
their assumptions were correct, participants being influenced by the facial expressions
shown in the pictures. Moreover, men as well as women were above chance accurate in
recognizing the depicted emotions. However, while female participants had no problems
with distinguishing all facial expressions, males experienced difficulties with photographs
depicting anger and fear.
Similarly, Montagne et. al. (2005, 136) asked if ‘men really lack emotional
sensitivity?’. Their study employed two measures of facial expression decoding: accuracy
and sensitivity. Their stimuli consisted of neutral faces which gradually morphed towards
expressing an emotion. Thus, they could record the participants’ accuracy, as well as
their sensitivity (how soon they were able to correctly recognize the depicted emotion).
The results yielded significant differences between female and male subjects’ accuracy as
well as sensitivity, with women obtaining higher scores. This study offers valuable insight
into the processes underlying the decoding of facial expressions, suggesting that there
may be quantitative, rather than qualitative differences between men and women.
59
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
Klein and Hodges (2001) offer an interesting interpretation of gender differences in
empathic accuracy. Their experiments revealed that, manipulating the subjects’
motivation influenced their ability to accurately assess the target person. Women
performed better than men only when they thought that the task demanded
interpersonal involvement. When financial stimulation was included, the differences in
accuracy disappeared.
Gender differences in personality traits assessments
Few studies concerning gender differences in social perception accuracy focused on the
assessment of personality traits, most of them employing the Big Five structure. Their
results are somewhat divergent, but there are also studies which yielded similar
conclusions.
Marcus and Lehman (2002) addressed the subject of gender differences in first
impression formation. Their results showed no significant discrepancy between men and
women’s rating consistencies. Nonetheless, it was noted that female participants
evaluated the actors as significantly more sociable and cheerful. This outcome is
contradicted by a study conducted by Johnson, Nagasawa and Peters (1977) who
concluded that the male and female subjects employed in their experiment yielded
similar evaluations of the target persons’ sociability. These divergent outcomes could be
explained by the fact that the mentioned studies show minor methodological
differences, using different stimuli and instruments.
Carney, Colvin and Hall (2007) conducted a similar study, employing the Big Five
personality questionnaire in assessing first impression accuracy. They concluded that
women were significantly more accurate than men in regard to openness, intelligence
and negative affect, whereas what neuroticism, extraversion and positive affect was
concerned, their evaluations didn’t differ from those of male subjects. In contrast, Lippa
and Dietz (2000) observed that men and women employed in a similar task – the
evaluation of a target person using the Big Five structure – rendered similar assessments.
The only discrepancy observed, involved the evaluation of neuroticism, where female
subjects were more accurate than males. Furthermore, Ambady, Hallahan and Rosenthal
(1995) examined the role of gender, among other aspects, in social perception accuracy.
The results suggested that women were more accurate in the evaluation of extraversion
and positive affect. Other variables bore an influence on perception fidelity, as well. It
was observed that shy and inexpressive women, as well as not shy men were better at
judging extraversion. Sociable and self-monitoring women were more accurate in
evaluating emotional stability. Furthermore, sociable men were significantly better
judges of a persons’ conscientiousness than were sociable women. Concluding, although
female subjects had, overall, a higher fidelity rate, participants who were more accurate,
either male or female, exhibited similar personality traits.
Other studies also revealed the influence of different personality characteristics on
social perception accuracy. Lippa and Dietz (2000) concluded that the intelligence and
openness of the observer correlated with higher accuracy scores. Moreover, Hall and
60
Bogdana Humă / Gender differences in impression formation
Halberstadt (1981) investigated the impact of masculinity/femininity, androginy
(masculinity plus femininity) and sex typing (masculinity minus femininity) on the fidelity
on impression formation. It was hypothesized that subjects with higher femininity scores
would be better judges. However the results showed no significant differences between
participants with high/low scores, although males with high masculinity scores tended to
be slightly more accurate. Moreover, men with high androginy and women with high sextyping scores proved to be better judges of video stimuli. Furthermore, partialling out
masculinity/femininity bore no result on accuracy scores of men and women, suggesting
that these characteristics are not the ones behind sex differences in person perception.
In a study conducted by Vogt and Colvin (2003) subjects were asked to rate the
personality of a target person shown in a 12 min video. Their evaluations were compared
with evaluations from self, family and friends of the confederate, thus obtaining accuracy
scores. Subjects also filled in a personality questionnaire which, among others assessed
their communion, which was defined as ‘the need to become one with the group of
others’ (Bakan apud Vogt and Colvin, 2003, 269). As expected, women showed higher
communion scores, as well as significantly higher accuracy ratings. However, when
partialling out gender, differences in accuracy scores due to communion still remained
significant.
Sex differences in impression formation fidelity are sometimes influenced by stimuli
employed in the study. Murphy, Hall and Colvin (2003) noticed that female participants
were better judges only when the target person was presented by means of a video
sequence with sound. Conversely, silent movies or transcripts did not yield significant
differences among sexes. Moreover, the meta-analysis conducted by Hall (1978) based
on 75 studies on the topic of nonverbal decoding skills rendered similar results. However,
only papers concerning emotions and states and not personality traits assessments were
included. Differences in accuracy between men and women were more likely to occur
when experiments employed stimuli with video and audio information. Nonetheless, the
author draws attention to the fact that studies yielding sex dissimilarities were more
likely to be published than the ones which failed to establish them. Additionally, since
gender cannot be experimentally manipulated, it is possible that other variables which
covariate with sex underlie differences in accuracy, thus yielding a spurious correlation.
The earlier mentioned study conducted Vogt and Covin (2003) identified communion as
playing an important role in social perception fidelity, notwithstanding gender.
An alternative explanation for differences in social perception accuracy is offered
by Hoffman (1977). His meta-analysis on studies involving children revealed that girls, as
compared to boys, are more inclined towards prosocial behavior, which includes, among
others, empathy and interpersonal sensitivity. Therefore, it is in their nature to put
themselves in somebody else’s place, thus being able to imagine what they feel or think.
Nevertheless, the meta-analysis conducted by Eisenberg and Lennon (1983) revealed that
gender differences in empathy occur more often in studies which employ self-ratings or
ratings by others and less when physiological measurements are undertaken.
Furthermore, Garner and Estep (2002) drew attention to the fact that other variables like
social context and age must be taken into consideration. Moreover, in real life situations
61
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
people might often wish to hide their feelings, employing a large variety of strategies,
which render new obstacles for emotional expressions decoding.
McClure’s (2000) meta-analysis of studies regarding facial expression processing
aimed to reveal the causes underlying females’ higher accuracy rates. Hence, the author
revised experiments which employed infants, children and adolescents as subjects. The
two main theoretical frameworks explaining gender differences discussed in the article
were: the neurobehavioral and the social constructivist model. According to the first,
girls might perform better on facial expression processing tasks only if their neurological
structures underlying facial recognition (the amygdala and the temporal cortex) undergo
an early maturation process. By contrast, the social constructivist approach suggests that
sex differences occur due to dissimilarities in emotion socialization. Gender stereotypes
play an important role in shaping decoding abilities, yielding different expectations for
men and women. Considering the studies undertaken on this topic, an integrated model
might be appropriate for explaining most of their results. It was observed that sex
differences were more likely to occur at an early age, thus sustaining the neurobehavioral
approach. However, since they remain constant over time, and do not decrease with age,
the assumptions of the social constructivism cannot be ruled out.
McAninch et. al. (1999) looked at children impression formation process
manipulating their expectations towards the target’s behavior. They were told that the
actor would either be shy or outgoing. Then, the children watched a tape of the
confederate (boy or girl) confirming or disconfirming their expectations. Finally they
were asked to rate the actor on several dimensions and to express their liking towards
him or her. The results showed that girls noticed more shy items when judging girls and
overall more neutral information than boys. Impression formation was influenced by the
targets’ observed behavior, whereas liking ratings were based on expectancies
generated by initial information. The authors’ explanations drew upon differences in sex
schemes which yield both observer and target impression formation effects. Therefore,
processes underlying social perception are to be influenced by own-sex schemes, which
consist of behavior prescriptions for the observers, as well as by superordinate schemes,
which contain information about what activities and characteristics are gender specific
and thus to be expected from the target person (Martin and Dinella, 2002).
In conclusion, the revised literature on gender differences in impression formation
suggests that women, more often than men, turn out to be better judges, yielding more
accurate assessments of targets’ internal states or personality traits. However, their
results were not always convergent, permitting alternative, if not contradicting
explanations. Moreover, the nature versus nurture controversy is still far from being
settled, thus rendering more possibilities for accounting for gender differences in social
perception.
The current study intends to reveal discrepancies in the first impression men and
women form of a confederate videotaped while reading a neutral text. It is hypothesized
that male and female subject will form different impressions of the target person,
underestimating or overestimating their sociability, morality, power and activity.
62
Bogdana Humă / Gender differences in impression formation
Method
Participants
Participants were 86 students, 47 males and 39 females, with ages ranging from 15 to 32,
forming two groups: high school students, with ages between 15 and 19, and university
students, with ages between 21 and 32. The 49 adolescents were recruited from the
Theodor Pietraru high school in Brăneşti and attended the experiment during school
time. The 37 graduate students were enrolled in a master’s program within the Faculty of
Sociology and Social Work of the Bucharest University. They took part in the experiment
during the first ten minutes of a course they were attending (Table 1).
Table 1 Sample structure
Male
Female
High school
43
6
University
4
33
Instruments
The experiment involved the use of electronic devices for audio-video recording and
playback. The participants’ evaluations of the target person were recorded using a
semantic differential, which was adapted and pretested for measuring impression
formation.
The 13 seconds long video employed in this study, showed a young female reading a
neutral text. Similar studies (Warner and Sugarman, 1986) have successfully used this
technique, selecting materials on scientific or artistic topics, in order to not reveal
anything about the actor’s personality through their content. For this purpose, two
paragraphs stemming from an article from the 8th issue of the Romanian magazine
Descoperă (2008) were randomly selected and read out loud by the confederate while
being filmed (Appendix 1). In the end, only 14 seconds, containing one sentence from the
middle of the passage, were chosen to be shown to the participants.
Since this study employs data from two previous experiments, few details
regarding instruments and procedure will be slightly different. However, it is supposed
that these small discrepancies will not interfere with the impression formation
measurements. The subjects watched the short movie either on a TV set connected to a
DVD-player or on a Smartboard connected to a laptop. Furthermore, a small number of
participants viewed the video sequence projected on a wall, which implied the use of a
projector, connected to a laptop.
In choosing a video instead of pictures or audio recordings as stimuli, I sought to
increase the ecological validity of the experiment, knowing it will also bring about a loss
in variables control. Nevertheless, in accordance with the ecological perspective on social
perception (McArthur and Baron, 1981) video sequences constitute better stimuli for
impression formation tasks, while static images or sounds are considered artificial,
lacking the necessary genuineness.
63
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
A semantic differential was employed in the measurement of the dependent
variable. It was adapted and pretested for impression formation assessment, thus
acquiring a few special features. It is based on Osgood’s (1969) semantic differential,
which consists of three dimensions: evaluation, power and activity. The author credits
the instruments with the ability to capture the affective meaning of the evaluated
concepts, which is supposed to stem from the reptilian brain, the oldest part of our
brains, common to both men and animals. Although this explanation sounds appealing it
was never scientifically confirmed.
The development of the semantic differential started with consulting studies which
employed similar instruments, which could be used or adapted for this experiment.
However, most of the research done on this subject employed personality
questionnaires, which focused only on certain aspects of impression formation, while
neglecting others ((Penton-Voak, Pound, Littke and Perrett, 2006; Johnson, Nagasawa
and Peters, 1977; Conner, Kathleen and Nagasawa, 1975). Therefore, I turned to studies
on different topics, using a semantic differential (Osgood 1969; Hartley, 1968; Hay 1970,
Osgood, 1971; Bush, 1973; Fagot, Leinbach, Hort and Strayer, 1997) and selected those
adjectives which could be employed in an impression formation task and grouped them
according to the dimension they belonged to, obtaining 33 items for evaluation, 15 for
power and 12 for activity. In the end, I created a semantic differential by randomizing the
order and polarity of the adjectives.
The validity and fidelity of the instrument was tested prior to the experiment on 141
students of the Faculty of Sociology and Social Work. They were shown a short movie,
similar to the one employed in this study and then were asked to evaluate the person
they had just seen in the video. The statistical analysis of the results yielded a four factors
structure, formed of 23 out of the 60 initial items (Table 2). The dimensions were labeled:
sociability, morality, power and activity. The validity of this outcome stems from the high
internal consistency of the four scales (above 0.775), as well as from the ecological
approach to person perception (McArthur and Baron, 1983). This theoretical framework
suggests that the characteristics of a target person perceived by an observer yield
certain affordances meaning that they are relevant for their interaction. The four
dimensions seem to respect this criterion, since they reveal essential information about
the observed person. Moreover, the newly emerged scales, sociability and morality,
couldn’t have been encountered in Osgood’s initial structure, since it was designed for
non-living objects and concepts.
64
Bogdana Humă / Gender differences in impression formation
Sociability
cold – warm
friendly – unfriendly
pleasant – unpleasant
close – distant
optimistic – pessimistic
Table 2 The structure of the semantic differential
Ethics
Power
fair – unfair
obedient – independent
sincere – insincere
cowardly – courageous
honest – dishonest
weak – powerful
correct – incorrect
bold – shy
determined –
undetermined
Activity
passive – active
apathetic – energetic
static - dynamic
slow – fast
slow – quick
funny – serious
interesting – not
interesting
likeable – not likeable
happy - unhappy
Unlike the semantic differentials’ initial form, which employed a seven point scale
assessment, the instrument used in this study consists of nine point scale evaluations.
The following arguments are supporting this choice: Whereas, until now, the semantic
differential was employed in the measurement of objects, abstract concepts or people, in
this study, the stimulus consists of a real person, presented to the participants by means
of a short movie. Since this task has an elevated level of difficulty, it was thought that a
nine point scale would be more adequate, offering a wider range of choices. Moreover,
since this experiment has an exploratory goal, it was considered better for the subjects
to have more freedom in forming an impression. In addition to that, when choosing a
nine point scale, I considered the nature of the dependent variable. The first impression
is both subjective and semi-conscientious, which makes it difficult to evaluate. Therefore,
it wouldn’t have been helpful to force the observers into selecting certain answer, which
didn’t entirely correspond with their opinion.
Design
This study aimed to reveal the effect of gender, the independent variable, on the first
impression, the dependent variable. However, since it is a secondary analysis, other
variables had to be taken into consideration. Therefore, the effect of age and education
were also taken into account when conducting the statistical analysis. Moreover, it must
be mentioned that data collection took place in two sessions yielding a few differences
between the two groups, which will be described in the next section. Nonetheless, it is
expected that this minor discrepancies will not significantly influence the outcome of the
experiment.
In order to forestall the subjects’ suppositions about the purpose of the study, a
false report technique was employed. Thus, participants were either told they were
attending an experiment on nonverbal communication or an experiment which aimed to
reveal the influence of information presentation on impression formation. Moreover, in
order to prevent a list effect in the semantic differential, two parallel versions of the
instrument were used in recording the subjects’ answers.
65
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
Procedure
Data collection sessions were conducted on the 18th, 20th of November and the 4th of
December 2008, for the high school group and on the 20th of November and the 4th of
December 2009 for the university students.
In the high school session, participants were informed by the vice-principal that
they will be asked to take part in a short experiment, which will take place in the
psychology lab. This location was chosen due to its technological facilities (TV set and
DVD-player) and also because of its limited seats. Only six students could participate in
the experiment at a time, which permitted a better control of their behavior.
In the beginning, the experimenter presented himself and mentioned the subject of
the investigation. He particularly informed the subjects that their answers were not
going to be evaluated in terms of right or wrong. Furthermore, since every individual is
unique, their personal opinion would be very important and it was therefore not
recommended that they copied their neighbors’ responses. After these instructions the
participants watched the video sequence for the first time and then were told how to fill
in the semantic differential. Then, they were shown the movie one more time before
evaluating the actor. After everybody had finished, the subjects were encouraged to ask
questions and they were thanked for their participation.
In the university session, the experiment was conducted either in a large classroom
with above 100 seats or in a small classroom with about 30 seats. The participants were
informed by their professor that they will be attending an experiment in the first part of
the course. Further, the procedure copied the one used in the high school session,
including instructions, instruments and debriefing sessions.
Results
Semantic differential
Before testing the hypotheses, it is useful to take a look at the semantic differentials’
fidelity and validity. Initially, it consisted of 23 adjectives grouped in four dimensions.
Nevertheless, the factor analysis – using Principal axis factoring extraction and Varimax
rotation – yielded a six factors solution and a KMO of 0.595. By successively eliminating
the adjectives: funny – serious, slow – fast, slow – quick, interesting – not interesting,
likeable – not likeable and happy – unhappy a four factor solution emerged which
explains 59.87% of the data variance and yielding a KMO of 0.745 (Table 3). This structure
was also validated by the individual internal consistency of the four scales: 0.648 for
sociability, 0.770 for ethics, 0.822 for power and 0.810 for activity.
66
Bogdana Humă / Gender differences in impression formation
Sociability
cold – warm
friendly – unfriendly
pleasant – unpleasant
close – distant
optimistic – pessimistic
Table 3 The final structure of the semantic differential
Ethics
Power
fair – unfair
obedient – independent
sincere – insincere
cowardly – courageous
honest – dishonest
weak – powerful
correct – incorrect
bold – shy
determined –
undetermined
Activity
passive – active
apathetic – energetic
static - dynamic
Hypotheses
It was hypothesized that men and women will form different impressions of an unknown
target person shown in a short movie. These impressions, measured using a semantic
differential with four dimensions – sociability, ethics, power and activity – are shown in
Table 4. It seems that the power scale recorded the largest discrepancy, women
underestimating this characteristic by 1.28 scale points. Next, the activity dimension
yielded a difference of 0.96 points, with female subjects underrating the actors’
dynamism. Sociability and ethics ranked third and forth, affording rather similar results,
with men slightly underestimating these traits.
Table 4 The mean scores of the four dimensions across gender
Sociability
Ethics
Power Activity
Male
4.46
5.59
4.93
4.11
Female
4.61
5.80
3.65
3.15
These results suggest that male and female participants might form different
impressions regarding the confederate’s power and activity. However, two issues have
to be taken into consideration. Firstly, in order to conclude that these results are valid
outside the considered sample, the evaluations should be significantly different. Since it
is known that semantic differential scores permit metric statistical computations, the
appropriate procedure is the independent sample t-test. The use of inferential statistics is
justified by the number of subjects in each sample (over 30). The results of the two
independent samples t-tests showed that there is a significant difference between the
power (t= -3.348, p=0.001) and activity (t=-2.208, p=0.03) scores of the two groups.
Secondly, when interpreting these results the educational level of the subjects has to be
taken into consideration. Therefore, a statistical analysis of the differences between high
school and university students’ evaluations is in order. The results of the independent
sample t-tests revealed significant differences between male and female subjects for the
power (t= 3.541, p=0.001) and activity dimensions (t=2.977, p=0.004).
Thus, both the gender of the observer as well as his educational level exert an
influence on the assessment of the power and activity of the confederate. Consequently,
a statistical analysis, using both variables, has to be conducted in order to establish which
bears an influence on the evaluation of the target person. Moreover, other factors, like
the age of the participants need to be taken into consideration.
67
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
In addition, in could be argued that, since most of the high school students were
males and most of the university students females (Table 1), the influence on impression
formation of the two variables cannot be clearly told apart.
Therefore, two linear regressions were conducted, using first the power and
second the activity score as dependent variables. The predictors inserted in the analysis
were: age, gender and education. The results of the ANOVA test showed a significant
improvement in the prediction of both power (F=4.042, p=0.01) and activity (F=2.845,
p=0.043) scores. The adjusted R square equaled 0.105 for power and 0.68 for activity.
However, none of the independent variables turned out to have a significant effect on
the assessment of the actor. Education came close to influencing the evaluation of
activity (beta= 0.611, p=0.056), suggesting that high school students were more likely to
overrate the target persons’ dynamism, all other factors being held constant. This
analysis clearly distinguishes between the effects of gender, age and educational level
and points out that neither have an influence on impression formation.
In conclusion, the statistical analysis of the collected data was not able to sustain
the study’s hypotheses. Possible explanations and future research directions are going to
be presented in the next section.
Discussion
The current study did not reveal an effect of the gender of the observer on the
impression formed of a target person shown by means of short video sequences. It was
hypothesized that evolutions rendered by male and female participants will yield
significant differences in the judgment of the confederate’s sociability, ethics, power and
activity. However, taking age and education into consideration, gender didn’t turn out to
be a significant factor for predicting the participants’ assessments of the actor’s
personality.
Firstly, it must be mentioned that, since this study is based on a secondary analysis
of two different experiments, several procedural discrepancies may have arisen, which
could account for the lack of evidence in support of the hypotheses. Although it is true
that every aspect of the environment might bear an influence on the studied
phenomenon, the minor differences between the two experimental settings cannot be
held accountable for the obtained results. It is hardly possible that gender differences in
impression formation did exist, but were annulled by the experimental conditions. On
the contrary, it is more likely that the differences between the two settings might have
augmented results discrepancies.
68
Bogdana Humă / Gender differences in impression formation
The influence of age on person perception
It might be useful, at this point, to take a closer look at the existing literature on the
influence of age on sex differences in person perception. McClure’s (2000) meta-analysis,
which was already mentioned in the first part of this paper, reviews studies on gender
differences in facial expression processing in the attempt to validate either the
neurobehavioral or the social constructivist model. Looking for evidence in support of
these opposing views, McClure describes emerging sex differences throughout infancy,
childhood and adolescence. The results, based on effect size analysis, revealed highest
scores during infancy. Differences could be observed in children and adolescents as well,
with no significant changes. Therefore, it can be assumed that age bears an influence on
some aspects of person perception, rendering differences between infancy and the two
other periods.
Although useful, McClure’s (ibidem) study sheds little light on this matter, since it is
limited to infants, children and adolescents. Moreover, the paper never intended to
address the influence of age on person perception. By contrast, Hall’s (1979) metaanalysis focused on studies using a large array of subjects of different ages, with the main
purpose of highlighting the influence of gender, among other variables, on nonverbal
communication. Based on effects size analysis, she concluded that age bears no influence
on decoding nonverbal cues.
Taking into account that the analysis conducted in this paper employed adolescents
as well as young adults, the fact that it failed to reveal the influence of age on impression
formation, falls in line with both studies mentioned above.
The hypotheses
Only one of the four characteristic investigated in this study was employed in two other
papers concerning gender differences. Marcus and Lehman (2002) and Johnson,
Nagasawa and Peters (1977) looked at male and female assessments of sociability,
obtaining divergent results. While the first study showed significant differences between
men and women in regard to sociability evaluations, the second one failed to reveal such
differences. Thus, the results obtained in the current study are consistent with the ones
yielded by Johnson, Nagasawa and Peters (1977).
Nevertheless, the lack of dissimilarities in ethics, power and activity assessments
need to be accounted for as well. The following explanation, based on the results of
similar studies already reviewed in the first part of this article, accounts for the lack of
discrepancies among all of the investigated dimensions. Most of them proposed
different interpretations of gender differences, or their absence. Table 5 summarizes
these explanations. It must be mentioned that these studies look at dissimilarities either
in assessment accuracy or just in the evaluation of a target person.
69
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
Table 5. Alternative explanations for gender differences
Study Gender differences
Explanation
Klein and Hodges (2001)
No
Motivation
Hall and Matsumoto (2004)
Yes
Different information processing
Hall and Matsumoto (2004)
Yes
Women are more confident
Montage et. al. (2005)
Yes
Women are more sensitive
Vogt and Colvin
No
Communion
McAnincg et. al. (1999)
Yes
Sex schemes
Since most of the variables underlying the identified explanations were not
employed when designing the current study, only their post factum assessment can be
undertaken. However, the following assumptions must be interpreted bearing in mind
the possibility of a hindsight bias. Since the subjects were not given any incentives or
specific information about the impression formation task, it can be assumed that their
motivation was to some degree similar. Furthermore, since the false report didn’t
mention the true purpose of the experiment – impression formation – instead distracted
the participants’ attention from this task, it can be assumed that their motivation to
accurately assess the target person was very low. Therefore, women, even though they
might have had an advantage in correctly evaluating the confederate, didn’t resort to
their abilities when judging the actor shown in the video.
This interpretation of the results needs to be further empirically tested. An
experiment assessing subjects’ communion and sensitivity, as well as their confidence in
their evolutions together with the manipulation of participants’ motivation might shed
light into this matter establishing once and for all the presence or absence of gender
differences in impression formation.
REFERENCES
Ambady, N. Hallahan M. and Rosenthal R. (1995). On judging and being judged in zeroacquaintance situations. Journal of Personality and Social Psychology, 69 (3), 518-529.
Bush, L. E. II. (1973). Individual differences multidimensional scaling of adjectives denoting
feelings. Journal of Personality and Social Psychology, 1, 50-57.
Carney, D. R., Colvin, C. R. and Hall, J. A. (2007). A thin slice perspective on the accuracy of first
impressions. Journal of Research in Personality, 41, 1054-1072.
Connor, B. H., Peters, K. and Nagasawa, R. H. (1975). Person and costume: Effects on the
formation of first impressions. Home Economics Research Journal, 4 (1), 32-41.
Eisenberg, N. and Lennon, R. (1983). Sex difference in empathy and related behavior.
Psychological Bulletin, 94(1), 100-131.
Ellis, L., Hershberger, S., Field, E., Wersinger, S., Pellis, S., Greary, D., Palmer, C., Hoyenga, K.,
Hetsroni, A. and Karadi, K. (2008). Sex Differences. Summarizing more than a Century of
Scientific Research. New York: Taylor and Francis Group.
Fagot, B. I., Leinbach, M. D., Hort, B. E. and Strayer, J. (1997). Qualities underlying the definitions
of gender. Sex Roles: A Journal of Research, 37 (1-2), 1-18.
70
Bogdana Humă / Gender differences in impression formation
Garner, P. W. and Estep, K. M (2002). Empathy and emotion expressivity. In J. Worell (ed.)
Encyclopedia of Women and Gender. San Diego, California. pp. 391-402
Hamilton, D. L., Katz, L. B. and Leirer, Von O. (1980). Cognitive representation of personality
impressions: organizational processes in first impression formation. Journal of Personality
and Social Psychology, 39 (6), 1050-1063.
Hall, J. A. (1978). Gender effects in decoding nonverbal cues. Psychological Bulletin, 85 (4), 845857.
Hall, J. A. and Halberstadt, A. G. (1981). Sex roles and nonverbal communication skill. Sex Roles, 7
(3), 273-287.
Hall, J. A. and Matsumoto, D. (2004). Gender difference in judgments of multiple emotions from
facial expressions. Emotion, 4 (2), 201-206.
Hartley, J. A. (1968). A semantic differential scale for assessing group process changes. Journal of
Clinical Psychology, 24 (1), 74.
Hay, R. D. (1970). Modified semantic differentials to evaluate formal organizational structures.
Journal of Business Communication, 7 (3), 13-23.
Hoffman, M. L. (1977). Sex differences in empathy and related behaviors. Psychological Bulletin,
84 (4), 712-722.
Johnson, B. H., Nagsawa, R. H. and Peters, K. (1977). Clothing style differences: Their impression
of sociability. Home Economics Journal, 6 (1), 58-63.
Katsikitis, M., Pilowsky, I. and Innes, J. M. (1997). Encoding and decoding of facial expression.
Journal of General Psychology,124(4), 357-370.
Kirouac, G. şi Dore, F. Y. (1985). Accuracy of the judgment of facial expression of emotion as a
function of sex and level of education. Journal of Nonverbal Behavior, 9 (1), 3-7.
Klein, K. J. K. and Hodges, Sara, D. (2001). Gender differences, motivation, and empathic
accuracy: When it pays to understand. Personality and Social Psychology Bulletin, 27 (6), 720730.
Lippa, R. A. and Dietz, J. K. (2000). The relation of gender, personality and intelligence to judges’
accuracy strangers’ personality form brief video segments. Journal of Nonverbal Behavior,
24 (1), 25-43.
Marcus, D. K. and Lehman, S. J. (2002). Are there sex differences in interpersonal perception at
zero acquaintance? A social relations analysis. Journal of Research in Personality, 36, 190-207.
Martin, C. L. and Dinella, L. M. (2002). Gender development. Gender scheme theory. In J. Worell
(ed.) Encyclopedia of Women and Gender. San Diego, California. pp. 507-521.
McAninch, C. B., Manolis, M. B., Milich, R. and Harris, M. J. (1993). Impression formation in
children: Influence of gender and expectancy. Child Development, 64, 1492-1506.
McClure, E. B. (2000). A meta-analytic review of sex differences in facial expression processing
and their development in infants, children, and adolescents. Psychological Bulletin, 126 (3),
424-435.
Montagne, B., Kessels, R. P. C., Frigerio, E., de Haan, E. H. F. and Perrett, D. I. (2005). Sex
differences in the perception of affective facial expressions: Do men really lack emotional
sensitivity? Cognitive Process, 6, 136-141.
Murphy, N. A., Hall, J. A. and Colvin, C. R. (2003). Accurate intelligence assessments in social
interactions: Mediators of gender effects. Journal of Personality, 7 (3), 466-493.
Osgood, Ch. E. (1969). Semantic differential technique in the comparative study of cultures.
American Anthropologist, 66 (3), 171-200.
Osgood, Ch. E. (1971). Exploration in semantic space: A personal diary. Journal of Social Issues, 27
(4), 5-64.
71
Journal of Comparative Research in Anthropology and Sociology, Volume 1, Number 1, Spring 2010
Penton-Voak, I. S., Pound, N., Little, A. C. and Perrett, D. I. (2006). Personality judgements form
natural and composite facial images: More evidence for a “Kernel of truth” in social
perception. Social Cognition, 24 (5), 607-640.
Perlini, A. H. and Lippe, C. D. E. (2006). Resourcefulness and social desirability trait ascriptions of
males: An evaluation of age and sex. Social Behavior and Personality, 34 (7), 827-836.
Stănescu, M.. (2008). Esenţa vieţii. Descoperă, 6 (8), 54-63.
Thayer, J. F. and Johnsen, B. H. (2000). Sex differences in judgement of facial affect: A
multivariate analysis of recognition errors. Scandinavian Journal of Psychology, 41, 243-246.
Vogt, D. S. and Colvin, C. R. (2003). Interpersonal orientation and the accuracy of personality
judgments. Journal of Personality, 71 (2), 267-295.
Warner, R. M. and Sugarman, D. B. (1986). Attributions of personality based on appearance,
speech and handwriting. Journal of Personality and Social Psychology, 50 (4), 792-799.
Zebrowitz, L. [1995] (1999). Impression Formation. In A. S. R. Manstead and M. Hewstone (eds.).
The Blackwell Encyclopedia of Social Psychology (pp. 309-314). Oxford: Blackwell Publishers
Ltd.
Bogdana Humă has graduated from the University of Bucharest in 2008, where she has
studied psychology and sociology and has earned a Masters degree in sociology in 2010.
Her current research interests include person perception and impression formation.
Among others, she has researched the role of factors such as odors and communication
(nonverbal and verbal) in impression formation.
72