ArticlePDF Available

Compensation between warmth and competence: Antecedents and consequences of a negative relation between the two fundamental dimensions of social perception

Authors:

Abstract and Figures

In the present chapter we first review research that has identified two fundamental dimensions of social perception. Having defined these two dimensions, we then present the results of a research program conducted to explore the relationship between them. In general, using both experimental and correlational data, we find evidence of a compensation effect between the two dimensions when two targets are compared. That is, when one target is judged more positively on one of the two fundamental dimensions, the second is judged more positively on the other dimension. We show that this compensation effect is confined to these two fundamental dimensions rather than something that more broadly characterises comparative judgements on any two judgemental dimensions. We then explore the importance of this compensation effect for the formation, maintenance, confirmation, and communication of mixed stereotypes of social groups.
Content may be subject to copyright.
Compensation between warmth and competence:
Antecedents and consequences of a negative relation
between the two fundamental dimensions of social
perception
Nicolas Kervyn
Universite
´catholique de Louvain, and FRS-FNRS (Fonds de la Recherche
Scientifique), Louvain-la-Neuve, Belgium
Vincent Yzerbyt
Universite
´catholique de Louvain, Louvain-la-Neuve, Belgium
Charles M. Judd
University of Colorado at Boulder, USA
In the present chapter we first review research that has identified two
fundamental dimensions of social perception. Having defined these two
dimensions, we then present the results of a research program conducted to
explore the relationship between them. In general, using both experimental
and correlational data, we find evidence of a compensation effect between the
two dimensions when two targets are compared. That is, when one target is
judged more positively on one of the two fundamental dimensions, the second
is judged more positively on the other dimension. We show that this
compensation effect is confined to these two fundamental dimensions rather
than something that more broadly characterises comparative judgements on
any two judgemental dimensions. We then explore the importance of this
compensation effect for the formation, maintenance, confirmation, and
communication of mixed stereotypes of social groups.
In college, students who work diligently and get straight As are seen as
nerds. Everyone tries to copy their notes and summaries but no one invites
them to parties. In sharp contrast, a girl who is on the cheerleading squad
will be invited to at least three different parties every Friday night, but she
Correspondence should be addressed to Nicolas Kervyn, Department of Psychology,
Catholic University Leuven, Louvain-la-Neuve, Belgium.
E-mail: nicolas.o.kervyn@uclouvain.be
EUROPEAN REVIEW OF SOCIAL PSYCHOLOGY
2010, 21, 155–187
Ó2010 European Association of Social Psychology
http://www.psypress.com/ersp DOI: 10.1080/13546805.2010.517997
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
will have a hard time finding a group to work with for her major assignment.
Later in life, stay-at-home mothers are seen as warm, caring, and sensitive.
People will turn to them if they don’t know how to soothe a baby, but not if
they need advice on an important career decision. Successful career women
are seen as cold and calculating. They will get the job done, albeit a bit
ruthlessly. Everyone can think of similar examples coming from their own
experience.
We believe that these different examples all have in common what we
have come to call the compensation effect between the two fundamental
dimensions of social perception: warmth and competence. As we show
below, the compensation effect is defined as social perceivers’ tendency to
differentiate between two social targets in a comparative context on the two
fundamental dimensions of social judgement by contrasting them in a
negative direction. So, when two groups or individuals are judged and
compared, the one judged more positively on one dimension is also judged
less positively on the other dimension. In the present chapter we start by
reviewing the research that has identified warmth and competence as the two
fundamental dimensions of social perception. We then present our
programme of research on the compensation effect.
THE TWO DIMENSIONS ACROSS TIME AND
RESEARCH DOMAINS
This first section reviews theoretical and empirical efforts showing that there
seem to be two fundamental dimensions underlying social perception. We
start with evidence from the person perception literature and then review
evidence from the intergroup literature. Finally we briefly discuss some
relevant research in other domains. In these literatures, while there is
considerable agreement about the existence of two dimensions, these have
been named and defined differently over time and across research domains.
But, as recent research reveals (Abele & Wojciszke, 2007) and as our
discussion shows, these different pairs of dimension all share a common
core.
Person perception
In the person perception domain, Rosenberg, Nelson, and Vivekananthan
(1968) were the first to provide evidence of the presence of two dimensions
that organise how we perceive others in terms of personality traits. These
authors asked their participants to describe people they knew by means of
64 commonly used personality traits. Specifically, participants were asked to
sort these traits into several piles according to whether they did or did not
co-occur in a given person. Rosenberg and colleagues (1968) used these trait
156 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
groupings to estimate pairwise similarities among the traits and these
similarities were then analysed with multidimensional scaling. Rosenberg
et al. (1968) found that a two-dimensional space adequately accounted for
the similarity among all pairs of the 64 personality traits. Further analyses of
the resulting two-dimensional space suggested dimensional labels of social
good–bad and intellectual goodbad. Figure 1 shows the resulting spatial
configuration.
The theoretical importance of this two-dimensional structure was nicely
demonstrated by Zanna and Hamilton (1972) who reinterpreted Asch’s
(1946) classic centrality effect in person perception. Asch had argued that
warm/cold were central traits based on the very different impressions that
were formed by describing someone as intelligent, skilful, industrious, warm
or cold, determined, practical, and cautious. Zanna and Hamilton argued that
what made warm/cold central traits in this list was that they were the only
traits to make reference to the social goodbad dimension, thus radically
affecting the overall impression of the person along that dimension. They
Figure 1. Personality traits on the two dimensions of ‘‘social good–bad’’ and ‘‘intellectual good–
bad’’ (Rosenberg et al., 1968).
WARMTH AND COMPETENCE 157
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
showed that if all adjectives in a list made reference to the social goodbad
dimension except for one, then that one, making reference to the intellectual
goodbad dimension, would emerge as central. Thus, impressions of
someone described with the following trait list: warm, sociable, industrious
or lazy, good-natured, humorous, were dramatically affected by the ‘‘central’’
industrious/lazy traits.
Wojciszke (1994) interpreted Rosenberg et al.’s (1968) two-dimensions
model through a discussion of the goals that underlie behaviours. Wojciszke
(1994, 2005) identified two categories of behavioural goals: the moral
category and the competence category. The moral category refers to an
actor’s intended goal: Does the actor have good or bad intentions? whereas
the competence category refers to the actor’s ability to attain the goal: Is the
actor able to carry out his/her intentions? (see also Read & Miller, 1989). The
combination of the two categories leads to four possible classifications of
actions that Wojciszke (1994) called virtuous success, virtuous failure, sinful
success, and sinful failure. While these descriptions of behavioural goals
engender different terms for the underlying dimensions, they map on very
nicely to Rosenberg et al.’s special configuration. Wojciszke, Abele, and
Baryla (2009) recently showed that those who virtuously succeed are liked
and respected, those who sinfully succeed are disliked and respected, those
who virtuously fail are liked and disrespected, and those who sinfully fail are
disliked and disrespected.
Intergroup perception
In 1996, Glick and Fiske introduced the concept of ambivalent sexism.
Questioning a simple view of prejudice in terms of one valenced dimension
ranging from positive to negative, these authors argued that prejudice
against women takes the form of benevolent sexism and/or of hostile sexism.
Benevolent sexists believe that women are high in warmth but low in
competence and therefore that men have to provide for them. This is the
traditional stereotypic perception of women (Eagly & Steffen, 1984). Hostile
sexists on the other hand see women, or a subgroup of women, as a
competent group that is in competition with men (i.e., low in warmth). In
Glick and Fiske’s (1996; Glick, Fiske, & Mladinic, 2000) ambivalent sexism
theory, the two constructs are distinct but are nevertheless positively
correlated. And, fundamentally, they rely on two underlying dimensions,
warmth and competence.
Turning to national stereotypes, a number of authors (Cuddy et al., 2009;
Phalet & Poppe, 1997; Poppe & Linssen, 1999) have shown a two-
dimensional structure underlying the perception of nations. In fact, national
stereotypes of the four different categories of sinful-loser, sinful-winner,
virtuous-loser, and virtuous-winner (Wojciszke, 1994) emerged in these
158 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
studies (Phalet & Poppe, 1997; Poppe & Linssen, 1999). Additionally, Phalet
and Poppe (1997) found that perceived conflict between the target country
and the participants’ own country predicted perceived warmth. They also
showed that the perceived power of a country was a positive predictor of its
perceived competence (see also Poppe & Linssen, 1999). Similarly, Poppe
and Linssen (1999) found that the size of the nation was a negative predictor
of its perceived warmth.
In order to study the applicability of a two-dimensional model of
stereotype content beyond gender stereotypes (Glick & Fiske, 1996), Fiske
Xu, Cuddy, and Glick (1999; Fiske, Cuddy, Glick, & Xu, 2002) asked their
participants to rate a series of groups on traits of warmth and competence.
The analysis of those ratings revealed that the different groups were
organised into four clusters defined by the two dimensions (see Figure 2).
First, there were a few commonly derogated groups (e.g., drug dealers;
welfare recipients . . .) that ended up with low ratings on both dimensions.
The ingroup or the culture’s main reference groups (aspirational groups)
were rated high on both dimensions (e.g., whites; students . . .). Other groups
were rated high on warmth but low on competence (e.g., elderly people,
blind people . . .) and still others were rated high on competence but low on
warmth (e.g., Jews, Asians . . .). Fiske et al. (2002) called these latter two
Figure 2. Distribution of social groups on the competence and warmth dimension in the
stereotype content model (Fiske et al., 2007).
WARMTH AND COMPETENCE 159
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
groups the paternalised and the envied groups, respectively. These authors’
stereotype content model (SCM) identifies two structural variables,
competition and status, that predict warmth and competence, respectively,
and argues that different configurations of warmth and competence lead to
different emotions towards social groups. Specifically, admiration is felt
towards ingroups and aspirational groups. Contempt is felt towards
derogated groups. Paternalised groups trigger pity. Finally, envied groups
cause envy and jealousy.
Other research in social sciences
Social psychology is not the only domain in which two dimensions have
been thought to underlie social perception. One example can be found in
political science, where Kinder and Sears (1985) argued that moral integrity
and competence constitute the two central and most important dimensions in
the overall evaluation of politicians. Each of these dimensions shows up
repeatedly when analysing voters’ open-ended descriptions of candidates
(Miller & Miller, 1976) and in factor analyses of trait descriptions of
political candidates (Kinder & Sears, 1985).
These same two dimensions also emerge in work on face perception. For
instance, Montepare and Dobish (2003) have shown that different facial
expressions lead to different dominance and affiliation trait inferences (see
also Knutson, 1996; Livingstone & Pearce, 2009). Happiness and surprise
expressions lead to high dominance and high affiliation trait inferences.
Angry expressions lead to high dominance and low affiliation trait
inferences. And sad expressions lead to low dominance trait inferences.
Integration of the different research traditions
Although the names are different in different research domains, commun-
ality (Abele, 2003; Bakan, 1966), social good–bad (Rosenberg et al., 1968),
morality (Wojciszke, 1994, 2005), affiliation (Montepare & Dobish, 2003),
and warmth (Fiske et al., 1999) are all very closely interrelated. Similarly,
dominance (Montepare & Dobish, 2003), agency (Abele, 2003; Bakan, 1966;
Wojciszke, 1994, 2005), intellectual good–bad (Rosenberg et al., 1968), and
competence (Fiske et al., 1999) all share considerable meaning similarity. In
research testing this intuition, Abele and Wojciszke (2007) selected five
prominent pairs of dimensions that have been proposed in social perception:
communion/agency; femininity/masculinity; collectivism/individualism;
morality/competence; other-interest/self-interest. Communion and agency
were first proposed by Bakan (1966) as the separate dimensions of human
existence, and have been used by Abele (2003) to study self-descriptions.
Femininity and masculinity have been identified in the gender role literature
160 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
as being stereotypically related to communion and agency respectively
(Bakan, 1966; Bem, 1974; Eagly & Steffen, 1984). Collectivism and
individualism were proposed by Triandis (1995) in the cultural difference
literature. Morality and competence have been used in person perception as
two separate types of behavioural information (Reeder & Brewer, 1979;
Wojciszke, 1994, 2005). And other-interest and self-interest are motivations
behind behaviours that are respectively other-profitable and self-profitable
(Peeters, 1992, 2005). Abele and Wojciszke (2007) asked their participants to
rate 300 traits names on one of these 12 dimensions in a between-
participants design. They then conducted a factor analysis on the data and
found a two-factor solution that explained 89% of the variance. Commu-
nion, interest in others, collectivism, morality, and femininity all loaded
strongly on the first factor. Agency, self-interest, masculinity, individualism,
and competence all loaded highly on the second factor (see Figure 3). This
study illustrates the fact that although the specific terms and the definition
behind those terms may vary somewhat, these various pairs of dimensions
also greatly overlap.
Furthermore, we note that Wojciszke’s (1994, 2005) oriented goal theory
on the one hand, and Fiske et al.’s (1999, 2002) stereotype content model on
the other, provide well-articulated and widely used models in the person and
the intergroup perception domains respectively. Comparing those two
models allows us to draw interesting conclusions. The four quadrants
identified by the oriented goal theory (virtuous-winner, sinful-winner, sinful-
loser, and virtuous-loser) correspond to the four clusters identified by the
stereotype content model (aspirational groups, envied groups, depreciated
groups, and paternalised groups). The predictors of the two dimensions
Figure 3. Clustering solution for the five pairs of dimensions (Abele & Wojciske, 2007).
WARMTH AND COMPETENCE 161
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
proposed in the two models are rather similar: competition and status for
the stereotype content model, intended goals and efficiency for the oriented
goal theory. There are also great similarities between these predictors and
the predictors of a country’s perceived warmth and competence identified by
Phalet and Poppe (1997) that they call perceived conflict (negative predictor)
and perceived power respectively.
At the same time, the oriented goals theory and the stereotype content
model differ in the way they predict the emotional consequences of the two
dimensions. The stereotype content model identifies four specific emotions:
admiration, pity, envy, and contempt. Each of these four emotions
corresponds to a specific combination of high/low warmth and high/low
competence. As for the oriented goal theory, Wojciszke (1994, 2005)
considers that communion leads to liking and agency to respect. There is
thus a 45-degree shift between the emotional predictions of the two models.
At the same time, this shift does not necessarily mean that the two models
contradict one another. As a matter of fact, the specific emotions of the
stereotype content model can be considered to be the result of a combination
of respect and liking. So, admiration would be felt towards a target that is
both liked and respected. Pity would be felt towards a target that is liked but
not respected. Envy would be felt towards a target that is respected but not
liked. And contempt would be felt towards a target that is disliked and
disrespected.
Given all this work, it seems clear that there really are two fundamental
dimensions of social judgement and perception. Additionally, it seems fairly
evident that although different terms for the two dimensions have been used
in these literatures, they all share a great similarity of meaning. Following
Fiske et al. (1999) we have chosen to use the terms warmth and competence
to refer to these two dimensions. Our reading of the literature leads us to
believe that there is immense overlap in meaning between these two
dimensions and other that have been used. Said differently, we are not aware
of any theory of social perception that proposes to use two dimensions that
are clearly not related to warmth and competence. Of course, it is possible to
further break down the two fundamental dimensions into a greater number
of dimensions in order to capture an even more nuanced description of
social objects, but these sub-dimensions definitely seem to be more domain-
specific (Alexander, Brewer, & Herrmann, 1999; Leach, Ellemers, & Barreto,
2007).
RELATIONSHIPS BETWEEN WARMTH AND
COMPETENCE
In this section we address the question of how these two fundamental
dimensions of social perception relate to each other. After reviewing past
162 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
work on this question, we introduce the compensation effect hypothesis and
the empirical work that supports it. Although the relationship between
warmth and competence has generally not been the primary focus of the
literature in this area, there are a number of suggestive clues that in fact turn
out to be rather contradictory. Some work points to a positive relationship
between the two dimensions, other research suggests that they are
orthogonal, and still other work suggests they are negatively related. We
review each in turn.
Evidence for a positive relationship
As can be observed in Figure 1, Rosenberg et al. (1968) found that the
social goodbad and the intellectual goodbad dimensions were not
orthogonal to each other. Indeed, the observed angle between them of
658corresponds to a positive correlation of .42. This means that a trait
or a target that is perceived positively on one of the dimensions also
tends to be perceived positively on the second. This is an example of
what is known as the halo effect. In 1920, E. L. Thorndike defined the
halo effect as the tendency to ‘‘think of a person in general as rather
good or rather inferior and to color the judgement of the separate
qualities by this feeling’’ (p. 25). Such an effect has often been reported in
person perception research (Anderson, 1965; Asch, 1946; Kelley, 1950;
Srull & Wyer, 1989).
In a classic halo effect experiment, Kelley (1950) manipulated expecta-
tions of students before they met a new instructor. Before the course,
students were given a vignette with information about their new instructor.
Participants read People who know him consider him to be a rather cold/very
warm person, industrious, critical, practical, and determined. The stimulus
person then came in and led the class in a 20-minute discussion. Students
who had read the very warm vignette rated the instructor as more
considerate to others, less formal, more sociable, more popular, less
irritable, more humorous, and more humane than the students that had read
the rather cold vignette. As far as behavioural measures were concerned,
56% of the students in the very warm conditions took part in the discussion,
whereas only 32% of the students in the rather cold condition did so.
Kelley’s (1950) experiment is generally taken as good evidence of the
prevalence and importance of the halo effect. But, having the two
dimensions of social perception in mind, we note that all the traits on
which Kelley finds an effect are related to warmth. No significant difference
was found on traits such as: self-assured, intelligent, will go far, knows his
stuff, etc. For these traits, the warmth manipulation proved to have no
impact. This leads us to the second possibility: the absence of a relation
between warmth and competence.
WARMTH AND COMPETENCE 163
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Evidence for an orthogonal relationship
Research in the field of intergroup relations also informs us on the nature of
the relationship between warmth and competence. As already mentioned,
studies by Fiske and colleagues (1999, 2002; for a review, see Fiske, Cuddy,
& Glick, 2007) showed that different social groups are perceived differently
on the warmth dimension and on the competence dimension, and that they
were roughly equally distributed around the resulting two dimensional
space. Ingroups and aspirational groups are perceived as warm and
competent; derogated groups are seen negatively on both dimensions;
envied groups are seen as competent and cold; and finally paternalised
groups are seen as warm and competent. Given the roughly equal
distributions of groups in these four clusters, the implication is that the
two dimensions of warmth and competence are orthogonal to each other.
Similarly, the oriented goal theory crosses the two dimensions of morality
and competence orthogonally (Wojciszke, 1994). According to this theory a
‘‘winner’’ can either be virtuous or sinful and a loser can either be virtuous
or sinful. A series of models proposed in other domains also support this
idea of an orthogonal relation. For instance, for the perception of political
candidates, Kinder and Sears (1985) do not assume any systematic
relationship between moral integrity and competence.
Evidence for a negative relationship
There exist various theoretical formulations that make strong assumptions
that the two dimensions of warmth and competence are negatively
correlated. In fact, some of these make the very strong assumption that
the negative correlation between the two dimensions is perfect, thus
reducing the two dimensions to a single one, with one end referring to
competent and cold targets and the other end referring to incompetent but
warm targets.
In documenting differences among cultures, Triandis (1995) proposed to
differentiate between collectivistic and individualistic cultures. Collectivistic
cultures are high on communality and individualistic cultures are high on
competence. Since cultures are either collectivistic or individualistic, it is
impossible with such a framework for a culture to be at the same time high
(or low) on both communality and competence.
Similarly, in political science, Funk (1997) argued that as political
expertise increases political perceivers value a candidate’s competence more
than their warmth. In the stimulus materials used to support her argument
she asked participants to judge two candidates, one who was competent but
cold and a second who was less competent but warmer. As hypothesised,
those who possessed great expertise preferred the former candidate to the
164 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
latter. Note, however, that the stimulus materials make the assumption of a
perfect negative relationship between the two dimensions. It is entirely
possible that more expertise is associated with valuing both competence and
warmth more in political candidates. However her stimulus materials could
not reveal such effects on separable dimensions.
Besides the work that simply assumes the existence of a negative relation
between the two fundamental dimensions, recent work has also started to
examine the content of stereotypes about groups that are characterised by
mixed stereotypes. As already mentioned (Glick & Fiske, 1996, 2001), there
are two very different stereotypic perceptions of women. On the one hand,
there is the stereotype of the traditional, warm, and caring woman who
suffers from a perceived lack of competence. On the other, professional
women and feminists suffer from a perceived lack of warmth. These two
stereotypes would suggest that beliefs about women are organised around
two negatively correlated dimensions. Similar results have been reported by
Cuddy and colleagues in examining stereotypic views of other groups that
manifest what are called mixed stereotypes. Beyond perception of women,
research on the stereotype content model (Fiske et al., 1999, 2002, 2007) has
shown that an important proportion of groups are characterised by mixed
stereotypes.
Building on these efforts, Cuddy and colleagues focused on two specific
groups that are associated with mixed stereotypes, namely working women
(Cuddy, Fiske, & Glick, 2004) and elderly people (Cuddy, Norton, & Fiske,
2005). In one of their experiments, Cuddy et al. (2004) presented one of two
curricula vitae to their participants. The target was a young working
woman: in one condition she had no children, whereas in the other condition
she was a mother. Results showed that, compared to the childless working
woman, the working mother was perceived as warmer but also as less
competent and less likely to be hired, promoted, and trained. What seems to
happen in these data is that participants’ stereotypic perception moves from
one mixed stereotype (competent and cold) to the other mixed stereotype
(warm and incompetent). A similar message emerged in the research by
Cuddy et al. (2005) on the stereotype of elderly people. Older people are
stereotypically perceived as warm but incompetent (Fiske et al., 2002).
Cuddy et al. (2005) found that when an elderly target was presented as more
competent than expected, i.e., an older person who has a sharp memory,
participants rated this target as more competent but also as less warm than
an elderly target with memory losses.
Last but not least, Yzerbyt, Provost, and Corneille (2005) proposed that,
when it comes to intergroup perception, the two fundamental dimensions of
social perception may in fact be related to one another negatively.
Specifically, these authors argued that, in the context of a comparison
involving two groups, perceiving one of the two groups as higher on one of
WARMTH AND COMPETENCE 165
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
the two fundamental dimensions (e.g., competence) should result in
perceiving the other group as higher on the other dimension (e.g., warmth),
Yzerbyt and colleagues (2005, Study 1) investigated what they called the
compensation effect in the context of a fully crossed design by focusing on
stereotypes held by French-speaking Belgians and by the French about each
other. French-speaking Belgian and French participants were asked to rate
both their own group and the other group on competence and warmth
(Figure 4). Results confirmed that both groups tended to agree that Belgians
were less competent but warmer than the French. This compensation
pattern in how these two national groups compare to each other was also
found in a follow-up study (Yzerbyt et al., 2005, Study 2) in which ratings
were collected from the members of a third group (French-speaking Swiss).
Conclusion
In sum, there are substantial differences in the literature about the nature of
the relationship between the two fundamental dimensions of social
judgement, warmth and competence. It is our belief that this disagreement
has arisen in part because no one has ever done the systematic work that is
necessary to directly address the question of how the two dimensions are
related. Rather, conflicting conclusions have emerged because of widely
diverse approaches in methods and measurement. The time has come, we
believe, for a systematic and experimental approach to assessing the
relationship that characterises these two fundamental dimensions in
perceivers’ perceptions of individuals and groups. This endeavour is all
the more important given the potential consequences of such perceptions for
interpersonal and intergroup relations.
Figure 4. Competence and warmth perception of French and Belgians in Yzerbyt et al. (2005).
166 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
SYSTEMATIC EXAMINATIONS OF THE
RELATIONSHIP BETWEEN WARMTH AND
COMPETENCE
In an effort to examine how the two dimensions of warmth and
competence relate to each other in the absence of pre-existing stereotypic
beliefs about particular individuals or groups, Judd, James-Hawkins,
Yzerbyt, and Kashima (2005) asked participants to form impressions of
two groups that they had never encountered before, the Blue and Green
groups. Individual members of these groups were described by behaviours
that they had supposedly engaged in. For instance, in the first study
reported by Judd et al. (2005, Expt. 1) the majority of the behaviours
attributed to one group were high or positive on competence while being
relatively undiagnostic of warmth (e.g., X published a short story in a
literary magazine while still in college.) and the majority of the behaviours
attributed to the other group were low or negative on competence and
undiagnostic of warmth (e.g., X failed to make his/her 9 am class because of
sleeping through the alarm.). A few positive and negative warmth
behaviours (but relatively undiagnostic of competence) were also
attributed to each group. These warmth behaviours were counterbalanced
between the two groups across participants, so that the same warm
behaviours and cold behaviours were seen with the Blue group and the
Green group on average. In sum, the two groups were clearly
differentiated on competence but both were equivalent and ambiguous
on warmth. After twice reading all the behaviours of both groups,
participants were asked to write down their impression of each group.
They then rated each group on a number of warmth and competence
traits.
Having outlined the specifics of the method, we can now spell out the
divergent predictions that would be made by the various traditions of
research. According to the supporters of the halo effect, the competent
group should be perceived as more competent and also as warmer than the
incompetent group. In contrast, for those researchers who conceptualise the
two dimensions as being essentially independent, the two groups would be
perceived as different on the dimension of competence but they would also
be seen as equally warm. Last but not least, the compensation hypothesis
predicts that the more-competent groups should in fact be perceived as less
warm than the less-competent group. This latter pattern is the one that
emerged: The group associated with a majority of positive competence
behaviours was not only rated as more competent than the other group but,
most importantly, the high-competence group was also rated as less warm
than the low-competence group (see Table 1). This negative difference on the
unmanipulated dimension constitutes clear experimental support for the
WARMTH AND COMPETENCE 167
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
compensation effect proposed by Yzerbyt and colleagues (2005): the group
that is portrayed by the experimental manipulations as higher on one of the
two dimensions is judged as lower on the other.
Importantly, the behaviours that were used as stimuli had been pretested,
by asking pretest participants to rate each behaviour individually on the two
dimensions. All of the behaviours included in this pretest were written so as
to be diagnostic of one of the two dimensions but relatively undiagnostic of
the second. However, when we looked at the resulting pretest ratings for the
behaviours, the mean ratings of the behaviours on the two dimensions were
positively related, thus replicating the positive relationship between the two
dimensions found by Rosenberg et al. (1968). The behaviours that were
actually used in the stimulus materials for the study were selected to
minimise this positive correlation. But even for these behaviours there was a
small positive correlation in how they had been rated on the two dimensions.
And yet, in spite of this, compensation in the rating of the two groups was
found, such that the high-competence group was judged as less warm than
the low-competence group.
In a follow-up experiment, Judd et al. (2005, Expt. 2) found a similar
compensation effect when warmth was the manipulated dimension.
Specifically, they presented participants with a high-warmth group and a
low-warmth group, both equivalent and ambiguous on competence. And
they found that in the impressions participants formed the high-warmth
group was rated as less competent than the low-warmth group.
Because warmth and competence have been shown to be fundamental
dimensions in both person and group perception, Judd et al. (2005) further
assessed whether the compensation effect could also be demonstrated with
individuals as targets of judgement. They replicated their experiment with
individuals rather than groups as targets of judgement. The results again
demonstrated compensation. Regardless of whether targets of judgements
were groups or individuals, the target to which high-competence behaviours
had been attributed was judged as less warm than the target to which low-
competence behaviours had been attributed.
TABLE 1
Means competence and warmth ratings for high-competence and low-
competence groups
Group
Dimension Low-competence High-competence
Competence 71.02 4.99
Warmth 2.34 0.06
(Judd et al., 2005; Expt. 1.)
168 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Necessary conditions
Whether involving national groups (Kervyn, Yzerbyt, Demoulin, & Judd,
2008; Yzerbyt et al., 2005), artificial groups (Judd et al., 2005; Kervyn, Judd,
& Yzerbyt, 2009a; Kervyn et al., 2008; Kervyn, Yzerbyt, Judd, & Nunes,
2009b) or individuals (Judd et al., 2005), all of the studies that report a
compensation effect included a comparison between two groups. In light of
this, Judd et al. (2005) further predicted that the compensation effect would
be weaker or even absent without a comparison context. To test this
hypothesis, Judd et al. (2005, Expt. 4) replicated their first experiment but
had participants learn about only one of the two groups instead of both of
them, i.e., the high-competence group for half of the participants and the
low-competence group for the other half. Results showed that no
compensation effect was found on the warmth dimension. The high-
competence group was rated more competent than the low-competence
group but both groups were rated similarly on the warmth dimension. Thus
it seems that a two-target comparative context is necessary in order to
observe the compensation effect. This perhaps partially explains why it is
that when individual traits and behaviours are rated, as in Rosenberg et al.
(1968) and in the pretest data reported by Judd et al. (2005), a positive
relationship is found, whereas these same stimuli yield compensation when
they are attributed to two targets and those targets are judged.
Another necessary condition was identified by Yzerbyt, Kervyn, and
Judd (2008, Expt. 2). Building on the apparently unique status of the two
dimensions in people’s perception of individuals and groups, these authors
hypothesised that the compensation effect would only hold for the two
fundamental dimensions of social perception and not for any pair of
dimensions. Replicating the compensation experiment by Judd et al. (2005),
these authors manipulated competence for half of the participants and
warmth for the other half, but this time they replaced the unmanipulated
dimension (warmth or competence) with a third dimension. That third
dimension was healthiness. The behaviours used for the healthiness had
been pretested to select those that were significantly positive or negative on
that third dimension while being neutral on warmth and on competence. For
instance, X hates vegetables and avoids eating them as much as possible was
used as a negative health behaviour and X does a few stretching exercises
every morning was used as a positive behaviour on the third dimension. So,
whereas one of the two groups was high on competence (warmth), the other
was low on competence (warmth) and both were equal and ambiguous on
the third dimension (healthiness). Results revealed the presence of a halo
effect. Specifically, the high-competence (warmth) group was rated as
healthier than the low-competence (warmth) group. In a follow-up
experiment Yzerbyt et al. (2008, Expt. 3) manipulated competence and
WARMTH AND COMPETENCE 169
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
presented equal and ambiguous information on warmth and on the third
dimension. They found a compensation effect on warmth and a halo effect
on the third dimension. So the high-competence group was simultaneously
perceived as healthier and colder than the incompetent group (see Table 2).
These experiments have thus allowed us to identify two necessary
conditions of the compensation effect. First, it seems that having a
comparative context is necessary for compensation to emerge. But we note
that Cuddy et al. (2004, 2005) found a negative relationship between warmth
and competence even though their experiment did not set up a comparative
context. Comparison might thus be a facilitating factor but not necessarily
an indispensable condition for compensation to occur. Second, we showed
that compensation does not occur on any pair of dimensions but seems to be
specific to the relation between warmth and competence. We interpret this
specificity as being due to the fact that, as we have reviewed above, warmth
and competence are the two fundamental dimensions of social perception.
They are thus the default dimensions that social perceivers use to assess their
social environment. This specific status is probably the reason why they
relate to one another in a different way (compensation) than any pair of
dimensions relate to one another. These boundaries of the compensation
effect also allow us to account for the fact that, as reviewed above, we found
contradictory evidence about the relation between the two fundamental
dimensions in past research. The compensation effect is not due to a
systematic negative relation between warmth and competence as assumed by
some authors (de Dreu, Beersma, Stroebe, & Euwema, 2006; Funk, 1997;
Triandis, 1995), as it is only found in some circumstances that can be
identified.
Mixed stereotype communication and maintenance
Judd et al. (2005) have shown how people bias their perception in a
compensatory direction when they form first impressions of two groups or
TABLE 2
Mean competence, healthiness, and warmth ratings of the low-compe-
tence and high-competence group as a function of dimension
Group
Dimension Low-competence High-competence
Competence 72.34 5.66
Healthiness 70.49 2.59
Warmth 2.65 0.16
(Yzerbyt et al., 2008; Expt. 3.)
170 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
individuals. Three series of studies have sought to identify ways in which
compensatory processes influence the perception of targets beyond mere first
impressions. This work has explored how social perceivers protect, confirm,
and communicate compensatory impressions they have formed. Kervyn
et al. (2009b, Expt. 1) replicated Judd et al.’s (2005) experiment with a full
design, involving manipulations of both dimensions. Half of the participants
encountered two groups that differed on competence but were ambiguous on
warmth. The other half encountered groups that differed on warmth but
were ambiguous on competence. Compensation on the unmanipulated
dimension was found in both conditions. And there was no manipulation by
dimension interaction. So the competent group was seen as colder than the
incompetent group to the same extent than the warm group was seen as less
competent than the cold group.
After participants had rated the two groups, Kervyn et al. (2009b, Expt.
1) gave them a list of questions. The questions had been written so that they
implied either warmth, lack of warmth, competence, or lack of competence.
For instance What kind of things are you likely to do to cheer up a friend who
is depressed or having personal problems? is a question that implies warmth,
and When you decide to cut class or skip a lecture, what kinds of things are
you likely to do instead? is a question that implies lack of competence.
Participants’ task was to indicate for each question whether they would
prefer to ask it of the high-competence (warmth) or the low-competence
(warmth) group (Dumont et al., 2003; Snyder, 1984; Snyder & Swann, 1978;
Snyder, Tanke, & Bersheid, 1977). As illustrated in Table 3, showing the
category of the 10 questions most strongly associated with the low-warmth
and the high-warmth group, questions implying low warmth and high
competence were selected for the low-warmth group and questions implying
warmth and lack of competence were selected for the high-warmth group.
And the reverse was true for the competence manipulation.
In a subsequent experiment, Kervyn et al. (2009b, Expt. 2) collected
answers to the questions of each of the four categories (i.e., questions
TABLE 3
Questions selected for the low-warmth and high-warmth group
Group
Questions implying Low-warmth High-warmth
Negative warmth 6 0
Positive warmth 0 9
Negative Competence 0 1
Positive Competence 4 0
(Kervyn et al., 2009.)
WARMTH AND COMPETENCE 171
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
implying high warmth, low warmth, high competence, and low compe-
tence) from naı
¨ve respondents, with every respondent answering all of the
questions. Then they showed these answers to a new group of participants
who thought they were getting interviews with two groups of individuals.
Specifically, some participants read the answers to the high-warmth
questions, attributed to one group, and the answers to the low-warmth
questions, attributed to a second group. Other participants read the
answers to the high-competence questions, attributed to one group, and
the answers to the low-competence questions, attributed to a second
group. Importantly, these participants did not know that all of the answers
they read were in fact generated by the same naı
¨ve participants. After
reading the interviews, participants wrote a few lines and then rated their
impression of each group. Results showed that even though the answers
had been collected from the exact same respondents, the different sets of
questions led to very different impressions that were in line with the
compensation effect. The interview using the set of questions selected for
the high-competence group led to an impression of a group more
competent and less warm than the interview using the set of questions
selected for the low-competence group.
In a final experiment, Kervyn et al. (2009b; Expt. 3) took this effect one
step further, involving live responses to the questions that had been
generated in the first study. They set up real interactions between triads of
participants. One participant was randomly assigned to the role of
interviewer. And supposedly through their answer to a minimal group
paradigm task, one of the remaining participants was told that he was a
typical member of the blue group while the other remaining participant was
told that he was a typical member of the green group. The interviewer then
went on to interview each of the other two participants using, for one
interviewee, the 10 questions selected for the group high on warmth or high
on competence, depending on the condition and, for the other interviewee,
the 10 questions selected for the group low on the same dimension. So, for
instance, in the warmth manipulation condition, one interviewee got six
questions implying low warmth and four implying high competence,
whereas the other interviewee answered nine questions implying high
warmth and one implying low competence.
After the interview, each of the three participants rated their impression
of the two groups that the interviewees supposedly belonged to, as well as
their impressions of the individual interviewees. With the exception of
ratings given by participants of themselves, all ratings were in line with
compensation. The interviewee asked the high-competence questions was
seen as less warm than the interviewee asked the low-competence questions,
and the reciprocal was true for interviewees asked the high- and low-warmth
questions.
172 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
In another series of experiments, Kervyn, Yzerbyt, and Judd (in press)
showed that not only the acquisition but also the interpretation of new
information is influenced by the compensation effect. In the first experiment,
as in Judd et al. (2005), Kervyn et al. (2009b), and Yzerbyt et al. (2008, Expt.
1), two groups were presented by their behaviours, one high and the other
low on one fundamental dimension and both ambiguous on the other
fundamental dimension. After the impression-writing and trait-ratings task,
participants were presented with a set of new behaviours that had allegedly
been performed by members of the two groups. For each of these new
behaviours, participants had to choose whether they would attribute the
behaviour to internal dispositions of the group member or not. Past research
(for a review, see Gilbert, 1998) has shown that dispositional attributions
mean that the behaviour is seen as relatively diagnostic of the actor. Results
supported the compensation effect for the trait ratings variable as well as for
the causal attribution measure. For instance, a warm behaviour was
attributed more dispositionally if it was associated with a low-competence
group member than with a high-competence group member.
In a follow-up experiment, Kervyn et al. (in press, Expt. 2) replicated
their first experiment, but they dropped the trait-rating task and instead of
new behaviours they showed participants a series of drawings that each
represented a behaviour (see Figure 5). For each drawing, participants were
asked to choose among four sentences the one that they thought best
described the drawing (Geeraert, Yzerbyt, Corneille, & Wigboldus, 2004).
The four sentences varied in language abstraction (Semin & Fiedler, 1988).
Results were in line with a compensation pattern and the semantic
expectancy model (Wigboldus, Spears, & Semin, 2005). For instance, a
drawing showing a warm behaviour (see Figure 5) was described in more
abstract terms if it was performed by a low-competence group member than
by a high-competence group member.
Kervyn et al. (2009a) identified how mixed stereotypes are maintained
through the greater differentiation of the two fundamental dimensions when
they are in a compensatory pattern. In this work, the authors described two
groups to participants, again providing behaviours that group members had
allegedly performed. But this time the groups were described with both
competence and warmth behaviours. Between participants, Kervyn et al.
(2009a) manipulated how the two dimensions were related between the two
groups. For one group of participants, the two dimensions manifested a
positive or halo relationship: one group was high in both warmth and
competence while the other was low in both. For the second group of
participants, the two dimensions manifested a compensatory relationship:
one group was high in competence but low in warmth while the other was
low in warmth but high in competence. Participants were then asked to write
down their impression of both groups and to rate them on a number of
WARMTH AND COMPETENCE 173
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
warmth and competence traits. Results showed that the ratings of the two
groups in the compensation condition were more extreme than those in the
halo condition. More precisely, the cold and competent group was rated as
more competent than the warm and competent group. And the warm and
incompetent group was rated as warmer than the warm and competent
group (see Figure 6).
These results on question selection, causal attribution, language
abstraction, and group differentiation (Kervyn et al., 2008, 2009a, in press)
not only highlight important ways through which compensated impressions
are maintained and communicated to others, they also provide evidence for
the robustness of the compensation effect. It is important to note that these
are very indirect ways of measuring the compensation effect. With traits
ratings of the two targets, it might be argued that participants were just
using a response strategy or even a logic-of-conversation rule to avoid giving
only negative ratings of one target and positive ratings of the other and that
this response strategy caused the observed compensation effect. This
Figure 5. Examples of drawings of high warmth and low warmth used in Kervyn et al. (in press).
174 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
interpretation does not hold for the more indirect measures such as
differentiation (Kervyn et al., 2009a), causal attribution, and language
abstraction (Kervyn et al., in press).
Compensation effect within perceivers
Evidence for a compensation effect has been found not only in the way
groups are perceived on average (at the mean level) but also in the within-
perceiver correlations between the ratings of these groups on the two
fundamental dimensions of social perception. In a study already described,
Yzerbyt et al. (2005) found that the more the French rated the Belgians as
warm, the more they rated their own group as competent. Building on that
effect and on the Judd et al. (2005) experiment that showed that a
comparative context leads to compensation, Kervyn et al. (2008) designed a
study to test the idea that the perception of a group can be influenced by the
group that it is compared to and that this influence would be in a
compensatory direction.
In their first experiment (Kervyn et al., 2008, Expt. 1), half of the
participants first rated Italians (the comparison country) on warmth and
competence and then Belgians (the target country and the participants’
ingroup). The other half of the participants first rated Germans and then
Belgians. Italians are stereotypically perceived to be warm and rather
incompetent, whereas Germans are perceived to be competent and rather
Figure 6. Competence and warmth ratings of the groups in the compensation pattern and the
halo pattern condition of Kervyn et al. (2009a).
WARMTH AND COMPETENCE 175
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
cold (Cuddy et al., 2009). Results showed that when Italy was the
comparison country, Belgium was rated as more competent and colder
than when Germany was the comparison country (see Figure 7). Ratings of
the very same group were thus influenced by the comparative context and
that influence went clearly in a compensatory direction. This impact of the
comparative context was replicated in a follow-up experiment (Kervyn et al.,
2008) that used either Japan or Brazil as the comparison country and
Canada as the target country. That is, when compared to Japan, Canada
was rated as warmer and as less competent than when compared to Brazil.
In both experiments, Kervyn et al. (2008) found that there was a clear
pattern of compensation in the within-perceiver correlations. The judged
warmth of the comparison country was a positive predictor of the judged
competence of the target country. And the judged competence of the
comparison country was a positive predictor of the judged warmth of the
target country. So the warmer the comparison country was rated, the more
competent Belgium (or Canada) was rated. And the more competent the
comparison country was rated, the warmer Belgium (or Canada) was rated.
Kervyn et al. (2008) ran a regression model predicting (for instance) the
target country’s judged warmth as a function of the comparison comparison
(e.g., Germany versus Italy), the target country’s judged competence, the
comparison country’s judged warmth, and the comparison country’s judged
competence. The two significant predictors were the comparison context and
the comparison country’s competence. Similarly, when the model predicted
the target country’s competence, the significant predictors were the
Figure 7. Competence and warmth ratings of the comparison and target countries in Kervyn
et al. (2008).
176 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
comparison context and the comparison country’s warmth. In both cases, it
was the judged level of the comparison country on one dimension that
positively predicted the target country’s rating on the other dimension.
Evidence for a compensation effect in the within-perceiver correlations
was also found in experiments that used artificial groups (Judd et al., 2005;
Kervyn et al., 2009a). Judd et al. (2005) calculated two difference scores in
participant’s ratings of the two groups they judged. One of these was the
difference in the rating of the two groups on competence; the other was
the difference between the ratings of the two groups on warmth. These two
difference scores were negatively correlated: Participants who saw a
larger difference between the two groups on the manipulated dimension
also rated the two groups as more different, in the opposite direction, on the
unmanipulated dimension.
Kervyn et al. (2009a) found the same negative correlation when judging
two groups that had been presented as compensatory on the two
dimensions. For participants who were presented with a warm and
incompetent group and a competent and cold group, the more participants
differentiated between the two groups on warmth, the more they
differentiated them on competence. This correlation did not emerge in the
halo condition in which participants were presented with a warm and
competent group and a cold and incompetent group.
Interestingly, these results supporting a compensation effect at the within-
perceiver level give us further evidence that the compensation effect is not
due to a general negative correlation between warmth and competence. The
correlations reported by Yzerbyt et al. (2005) and the regression results of
Kervyn et al. (2008) show that the compensation effect leads specifically to a
positive correlation across groups and across dimensions, not to a general
negative correlation between warmth and competence. Furthermore, in the
differentiation experiment (Kervyn et al., 2009a), a negative correlation
between warmth and competence was found only when the two groups
manifested a compensation pattern, not when they manifested a halo
pattern.
Compensation and identification
The vast majority of the studies that have found support for the
compensation effect (Judd et al., 2005; Kervyn et al., 2008, 2009a, 2009b,
in press; Yzerbyt et al., 2008) are characterised by the fact that participants
were in the role of an external observer. But a number of studies and
experimental results show that compensation also emerges when social
perceivers are members of one of the two groups rated. In Yzerbyt et al.’s
(2005) first study, respondents were French and Belgians who rated both
French and Belgians. Results showed that French respondents saw less of a
WARMTH AND COMPETENCE 177
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
difference between the two groups on warmth than their Belgian counter-
parts. Similarly, compared to the French respondents, Belgian respondents
did not see as large a difference between the two groups on competence. In
other words, both groups minimised the difference on the dimension along
which their ingroup was seen more negatively than the other group. But we
note that the compensatory differences were significant nonetheless. The
other way to look at this pattern is to say that respondents overstated the
difference on the dimension that was advantageous to their group. For
Belgian respondents, this tendency to stress the difference on warmth was
even stronger for highly identified group members than for less-identified
respondents.
A similar pattern was observed in the first experiment of Kervyn et al.
(2008) where the target country was Belgium, that is, the participants’
ingroup. Here too, the authors measured participants’ identification with
Belgium. As would be expected, results showed that identification correlated
positively with the ratings of Belgium both on warmth and on competence.
However, there was no interaction of identification with the comparison
context. In other words, the ratings of Belgium made by highly identified
participants were still influenced by the comparison context in a
compensatory way.
In one of the experiments conducted by Judd et al. (2005, Exp 5),
participants were first confronted with a minimal group procedure and
induced to believe that they belonged to one of the two groups presented to
them via a list of behaviours. Specifically, whereas half of the participants
were made to believe that they belonged to the high-competence group, the
other half thought that they belonged to the low-competence group. An
equal amount of ambiguous information was provided about both groups.
Replicating the pattern obtained in the other experiments, the high-
competence group was rated as colder than the low-competence group.
However, this compensation effect was moderated by group membership.
The members of the low-competence group saw a larger difference between
the two groups on warmth than the members of the high-competence group.
So, as was the case in Yzerbyt et al.’s (2005) first study, participants
exaggerated the difference on the dimension that was favourable to them
even though a compensation effect was found for both conditions of group
membership.
Finally, in one of Kervyn et al.’s experiments (2009b, Expt. 3), two
participants were interviewed by a third one. Through a minimal group
paradigm, the two interviewees were led to believe that they were each part
of one of two groups and that they would receive questions that had been
selected to correspond to members of their groups. The questions used were
those that had been selected for each of the two groups in an earlier study
(Kervyn et al., 2009b, Expt. 1). Concretely, whereas one participant was
178 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
asked the questions that had been selected for the high-competence
(warmth) group, the other received the questions that had been selected
for the low-competence (warmth) group. Not only did the interviewer and
the other interviewee rate participants’ groups in line with the compensation
effect, this effect also emerged in the way participants rated their own group.
For instance, participants who received the high-competence questions rated
their own group as more competent but also colder than the group of the
other participant.
In sum, these various results gathered in a series of studies show that
membership in one of the two groups does not prevent the compensation
effect from emerging. Although some studies reveal the presence of a
moderating impact of membership and/or identification, group members
tend to conform to a compensatory pattern in their appraisal of the groups
involved, even while perhaps minimising differences on the dimension along
which compensation leads to less favourable evaluations of their own group.
These findings attest to the robustness of the compensation effect.
GENERAL DISCUSSION
The first portion of this review was focused on diverse lines of work that
have all supported the hypothesis that there are two fundamental
dimensions of social perception and judgement. Although the resulting
two dimensions have been defined in the literature in various ways,
fundamentally the perception of both groups and persons is organised along
two dimensions, which we have referred to as warmth and competence. The
ubiquity of these two dimensions in social perception and judgement is
striking.
Somewhat surprisingly, given the widespread agreement on these two
fundamental dimensions, there has been no systematic work that has
directly examined how the two dimensions are related to each other. There
has been some work that has indirectly examined this question. But a review
of that work yields rather inconsistent conclusions, with suggestions of both
positive and negative relations between perceived warmth and competence.
Our own prior work in the judgement of national groups (Yzerbyt et al.,
2005) was suggestive of a compensatory relationship between the two
dimensions which then motivated our more systematic approach to the
topic.
We then turned to a review of our more systematic work that has
supported, amplified, and examined the implications of this compensation
effect. This work provides unequivocal support for the existence of a
compensation effect that is defined as social perceivers’ tendency to
differentiate two social targets in a comparative context on the two
fundamental dimensions by contrasting them in a negative direction.
WARMTH AND COMPETENCE 179
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Throughout this review of our research program, we have highlighted the
fact that our results paint a picture of the compensation effect that is much
more specific than a general negative correlation between warmth and
competence and much more robust than a mere logic-of-conversation
strategy. We have identified a number of necessary conditions for the
compensation effect. And even our results on compensation at the
correlational level show a specific picture of a positive correlation across
groups and across dimensions. We also demonstrated some important
consequences of such compensatory perception of other people and groups
on the formation, maintenance, confirmation, and communication of
compensatory stereotypes.
At this stage, one crucial question remains unanswered. Why does the
compensation effect emerge at all? Why is there a tendency to differentiate
two social targets in a comparative context on the two fundamental
dimensions by contrasting them in a compensatory direction? Our research
has so far focused on identifying the compensation effect, finding its
boundaries, and highlighting its important consequences. Therefore we can
only offer conjectures about this important question. In the following we
review two lines of research that we think could lead to interesting insights
about the origin of the compensation effect. First, we propose an
interpretation of the compensation effect based on adherence to mixed
stereotypes (Reinhard, Stahlberg, & Messner, 2008, 2009). Then we reflect
on how system justification theory (Jost & Hunyadi, 2002) can offer possible
reasons for the compensation effect. We also discuss why we think that the
compensation effect should not be interpreted as a simple contrast effect
(Mussweiller, 2007; Sherif & Hovland, 1961).
Compensation effect and mixed stereotype adherence
One possible interpretation of the compensation effect is that social
perceivers have pre-conceived stereotypes that social targets are either
warm and incompetent or cold and competent. In a recent series of research,
Reinhard et al. (2008) have shown what they call a ‘‘failure as an asset’’
effect. They showed that for high-status groups members, failing in a task
that is perceived as stereotypical of the low-status group leads to a
perception of higher occupational success (Reinhard et al., 2008) and to
higher self-esteem (Reinhard et al., 2009). Additionally, those effects were
mediated by perceived stereotypicality and group identification. So, for
instance, a man who fails at a task that was introduced as a task in which
women usually outperform men is perceived as more masculine and as
having better occupational success than if he succeeds at the task or if the
task is introduced as a task in which men usually outperform women
(Reinhard et al., 2008). And when men fail at a stereotypically feminine
180 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
task, they feel higher ingroup identification which then leads to higher self-
esteem (Reinhard et al., 2009). These results were replicated outside the
gender domain with other kinds of high-status groups.
Reinhard et al.’s (2008, 2009) results only show adherence to a competent
and cold stereotype (see the link between status and competence and
between masculinity and high competence/low warmth discussed above) and
does not show a similar effect for the warm and incompetent stereotype.
Nevertheless we believe that it points towards the fact that social perceivers
expect social targets to manifest a compensated pattern. This would explain
why when we present an ambiguous situation, participants err on the side of
considering the two groups as two mixed stereotypes groups (Judd et al.,
2005; Kervyn et al. 2009b, in press) and that in the differentiation
experiment, when the pattern presented fit the mixed stereotype expectation,
participants adhere to it more than in the halo condition (Kervyn et al.
2009a). It would also explain why compensation is not observed when the
ambiguous dimension is not part of the expected mixed stereotype (Yzerbyt
et al., 2008).
Compensation effect and system justification
For Jost and colleagues, people prefer an evaluatively balanced view of
social groups in order to justify the existing social structure (Jost & Banaji,
1994; Kay, Jost & Young, 2005; for a review, see Kay et al., 2007). In one
illustrative experiment, Kay and Jost (2003) found that exposure to a ‘‘poor
but happy’’ exemplar increased system justification scores whereas exposure
to a ‘‘rich but happy’’ or to a ‘‘poor but miserable’’ exemplar decreased
system justification scores. ‘‘Poor’’ and ‘‘rich’’ are clearly related to status,
the structural variable that predicts competence (Fiske et al., 1999, 2002).
Conversely, ‘‘happy’’ and ‘‘miserable’’ are directly related to warmth. The
‘‘poor but happy’’ and the ‘‘rich but miserable’’ exemplars are thus typical of
compensated exemplars: strong on one dimension and weak on the other
dimension. In contrast, the ‘‘poor and miserable’’ and the ‘‘rich and happy’’
exemplars are either high on both dimensions or low on both.
We see the findings obtained by Kay and Jost (2003) as compatible with
the compensation effect. That is, we reinterpret Kay and Jost’s (2003) results
as showing that exposure to the compensated exemplars increased system
justification scores, whereas exposure to the halo exemplar decreased system
justification scores. When we present two groups that are differentiated on
one fundamental dimension and ambiguous on the other in (Judd et al.,
2005; Kervyn et al. 2009a, in press), it is after all an unjust ‘‘system’’ that our
participants are presented with, a system in which the high group has a
higher total amount of positive characteristics (higher on the manipulated
dimension and equal on the unmanipulated one) than the low group. Our
WARMTH AND COMPETENCE 181
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
interpretation is that participants react to—i.e., correct—this unjust system
by compensating on the unmanipulated dimension. They bias their
perception of the two groups on the dimension that was left ambiguous in
order to create a system in which both groups have strengths and
weaknesses, a situation that is closer to one in which both groups would
have an equal amount of positive characteristics. Our interpretation of such
a pattern of findings is thus that the compensation effect is a tool that the
social perceiver uses to perceive the social world as being just.
It is important to emphasise that the work on the compensation effect
and the two fundamental dimensions questions the most general, and indeed
undifferentiated, version of system justification and instead supports a
rather more restricted version. Evidence for this restricted version comes
from the third dimension experiments conducted by Yzerbyt et al. (2008). In
these studies, no compensation but rather a halo effect emerged for an
ambiguous non-fundamental dimension. This finding suggests that com-
pensation only happens on the dimensions that matter. As a matter of fact,
there is ample evidence for the fact that, in person perception, the
dimensions of competence and warmth matter the most, hence their status
as fundamental. So social perceivers consider that a social system is justified
if targets who are compared have the same total amount of positive
characteristics not on any given set of dimensions but rather on the two
dimensions that do matter.
The result of the differentiation experiment (Kervyn et al., 2009a) also fit
our restricted system justification interpretation. The compensated exem-
plars fit the expectancy of system justification in that participants are
comfortable with the differences between the two groups and ready to
perceive those differences as rather large. As a matter of fact, even if the
difference is very large it remains true that across dimensions both groups
have the same total amount of positive characteristics. In contrast, in those
cases where participants are confronted with the halo exemplars the high
group has globally more positive characteristics than the low group, a
situation that presumably leads to questions about the fairness of the system
and then perhaps to attempts to restore some fairness. Accordingly, in these
conditions participants tend to minimise the difference between the two
groups. This interpretation holds for both the mean and the correlational
results of the differentiation experiment.
Finally, the compensation effect observed in the within-perceiver
correlations (Kervyn et al., 2008) also supports our restricted view of
system justification. Indeed, we show that the perceived level of the first
group on one dimension predicts the score of the second group on the other
dimension. So the way both groups are rated may vary from one respondent
to another, but the ratings obey the rule that if the first group is higher than
the second group on one dimension then the second group is judged higher
182 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
than the first on the other dimension, presumably in order to maintain a
balanced (justified) system across the two fundamental dimensions of social
perception.
Compensation is a tool that creates and maintains compensated (mixed)
stereotypes and the motivation to create those compensated stereotypes may
be related to the motivation to perceive the system as fair and balanced. So
the compensation effect could be the process that composes the comple-
mentary part of a feedback loop. Glick and Fiske (1996) and Kay and Jost
(2003) have shown that compensated stereotypes lead to system justification.
Our interpretation is that, in turn, the motivation to perceive the system as
justified leads to the creation of compensated stereotypes. However, this link
between system justification and compensation remains to be experimentally
demonstrated.
Compensation effect and contrast
It may be tempting to reinterpret the compensation effect as a specific
instance of contrast between two targets (Mussweiller, 2007; Sherif &
Hovland, 1961). However, we believe that it is difficult to consider this as an
adequate explanation. A contrast effect is said to occur when the judgement
of a target on some dimension is influenced by the location of another target
on the same dimension (Mussweiller, 2007). In other words, if one target is
particularly competent then another contrasted target might be judged to be
particularly incompetent. This seems to us to be a rather different result
from our compensation result, which involves judgements across two
different dimensions, with the more positively evaluated target on one
dimension judged less positively on the second. It is not that one target is
contrasted from another, but that the evaluative difference is reversed across
the two fundamental dimensions. Moreover, Yzerbyt et al.’s (2008)
experiments have convincingly demonstrated that compensation does not
take place on any available second dimension but that this second dimension
has to be the second fundamental dimension.
Last but not least, the contrast interpretation is unable to account for the
evidence for a compensation effect at the correlational level found in Kervyn
et al. (2008) and in Kervyn et al. (2009a). Indeed, the contrast effects always
consists of a negative relationship between the ratings of two targets on the
same dimension (Mussweiller, 2007), whereas we find evidence of a positive
relationship between the ratings of two targets on two different dimensions.
Implications of the compensation effect for stereotype change
We wish to add one final note concerning the implications of the
compensation effect for stereotype change. Cuddy et al. (2004, 2005) have
WARMTH AND COMPETENCE 183
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
nicely illustrated the pernicious effect that the compensation effect can have
when it comes to the modification of stereotypic views. Both for working
mothers and elderly people, positive changes on one dimension (warmth and
competence, respectively) were compensated by a decrease on the other
dimension. The empirical evidence emerging from our experimental work
points to the fact that the findings reported by Cuddy et al. (2004, 2005) are
not two isolated effects but rather a widespread phenomenon that is likely to
influence any stereotype change efforts when it comes to groups associated
with mixed stereotypes. The implication would seem to be that there might be
unintended effects of stereotype change efforts devoted to a target group that is
seen relatively negatively on one of the two fundamental dimensions. Suppose,
for instance, that stereotype change efforts were devoted to convincing people
that a stereotypically low-competent group was in fact more competent. Such a
change,if successful, might have the unintended effect of convincing people that
the group is also less positive on the warmth dimension.
A more theoretical implication of the present line of research is that it
shows the importance of paying attention to the dynamic relations that exist
between warmth and competence. The research on the stereotype content
model has shown the meaningfulness of using the two dimensions of social
perception to measure stereotype content. The research on the compensa-
tion effect has shown that there are systematic biases in the way these two
dimensions relate to one another. Of course, one should not expect that the
negative relationship observed in the several studies and experiments is
always at work. As our work has shown, there are a series of necessary
conditions for this effect to emerge (Judd et al., 2005; Yzerbyt et al., 2008).
This calls for a continued theoretical and empirical effort to tackle the
crucial question of identifying the moderators of the relationship between
warmth and competence.
REFERENCES
Abele, A. (2003). The dynamics of masculine–agentic and feminine–communal traits: Findings
from a prospective study. Journal of Personality and Social Psychology,85, 768–776.
Abele, A., & Wojciszke, B. (2007). Agency and communion from the perspective of self versus
others. Journal of Personality and Social Psychology,9, 751–763.
Alexander, M., Brewer, M., & Herrmann, R. (1999). Images and affect : A functional analysis of
out-group stereotypes. Journal of Personality and Social Psychology,77, 78–93.
Anderson, N. (1965). Likableness ratings of 555 personality-trait words. Journal of Personality
and Social Psychology,9, 272–279.
Asch, S. (1946). Forming impressions of personality. Journal of Abnormal and Social
Psychology,41, 1230–1240.
Bakan, D. (1966). The duality of human existence: An essay on psychology and religion. Chicago:
Rand and McNally.
Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and
Clinical Psychology,42, 155–162.
184 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Cuddy, A., Fiske, S., & Glick, P. (2004). When professionals become mothers, warmth doesn’t
cut the ice. Journal of Social Issues,4, 701–718.
Cuddy, A., Fiske, S., Kwan, V., Glick, P., Demoulin, S., Leyens, J-Ph., et al. (2009). Stereotype
content model across cultures: Universal similarities and some differences. British Journal of
Social Psychology,48, 1–33.
Cuddy, A., Norton, M., & Fiske, S. (2005). This old stereotype: The stubbornness and
pervasiveness of the elderly stereotype. Journal of Social Issues,61, 267–285.
de Dreu, C. K. W., Beersma, B., Stroebe, K., & Euwema, M. C. (2006). Motivated information
processing, strategic choice, and the quality of negotiated agreement. Journal of Personality
and Social Psychology,90, 927–943.
Dumont, M., Yzerbyt, V. Y., Snyder, M., Mathieu, B., Comblain, C., & Scaillet, N. (2003).
Suppression and hypothesis testing: Does suppressing stereotypes during interactions help to
avoid confirmation biases? European Journal of Social Psychology,33, 659–677.
Eagly, A., & Steffen, V. (1984). Gender stereotypes stem from the distribution of women and
men into social roles. Journal of Personality and Social Psychology,46, 735–754.
Fiske, S., Cuddy, A., & Glick, P. (2007). Universal dimensions of social perception: Warmth,
then competence. Trends in Cognitive Science,11, 77–83.
Fiske, S., Cuddy, A., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content:
Competence and warmth respectively follow from the perceived status and competition.
Journal of Personality and Social Psychology,82, 878–902.
Fiske, S., Xu, J., Cuddy, A., & Glick, P. (1999). (Dis)respecting versus (dis)liking: Status and
interdependence predict ambivalent stereotypes of competence and warmth. Journal of
Social Issues,55, 473–491.
Funk, C. L. (1997). Implications of political expertise in candidate trait evaluations. Political
Research Quarterly,50, 675–697.
Geeraert, N., Yzerbyt, V. Y., Corneille, O., & Wigboldus., D. (2004). The return of
dispositionalism: On the linguistic consequences of dispositional suppression. Journal of
Experimental Social Psychology,40, 264–272.
Gilbert, D. T. (1998). Ordinary personology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.),
The handbook of social psychology (4th ed., Vol. 2, pp. 89–150). New York: McGraw-Hill.
Glick, P., & Fiske, S. (1996). The ambivalent sexism inventory: Differentiating hostile and
benevolent sexism. Journal of Personality and Social Psychology,70, 491–512.
Glick, P., & Fiske, S. (2001). Ambivalent sexism. In M. Zanna (Ed.), Advances in experimental
social psychology (Vol. 33, pp. 115–188). Thousand Oaks, CA: Academic Press.
Glick, P., Fiske, S., & Mladinic, A. (2000). Beyond prejudice as simple antipathy: Hostile and
benevolent sexism across cultures. Journal of Personality and Social Psychology,79, 763–
775.
Jost, J. T., & Banaji, M. (1994). The role of stereotyping in system-justification and the
production of false consciousness. British Journal of Social Psychology,33, 1–27.
Jost, J. T., & Hunyady, O. (2002). The psychology of system justification and the palliative
function of ideology. European Review of Social Psychology,13, 111–153.
Judd, C., James-Hawkins, L., Yzerbyt, V., & Kashima, Y. (2005). Fundamental dimensions of
social judgement: Understanding the relations between judgements of competence and
warmth. Journal of Personality and Social Psychology,89, 899–913.
Kay, A., & Jost, J. (2003). Complementary justice: Effects of ‘‘poor but happy’’ and ‘‘poor but
honest’’ stereotype exemplars on system justification and implicit activation of the justice
motive. Journal of Personality and Social Psychology,85, 823–837.
Kay, A., Jost, J., Mandisodza, A., Sherman, S., Petrocelli, J., & Johnson, A. (2007). Panglossian
ideology in the service of system justification: How complementary stereotypes help us to
rationalize inequality. In M. Zanna (Ed.), Advances in experimental social psychology (Vol.
38, pp. 305–358). Thousand Oaks, CA: Academic Press.
WARMTH AND COMPETENCE 185
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Kay, A., Jost, J., & Young, S. (2005). Victim-derogation and victim-enhancement as alternate
routes to system-justification. Psychological Science,16, 240–246.
Kelley, H. H. (1950). The warm–cold variable in first impressions of persons. Journal of
Personality,18, 431–439.
Kervyn, N., Judd, C., & Yzerbyt, V. (2009a). You want to appear competent? Be mean! You
want to appear sociable? Be lazy! Group differentiation and the compensation effect. Journal
of Experimental Social Psychology,45, 363–367.
Kervyn, N., Yzerbyt, V., Demoulin, S., & Judd, C. (2008). Competence and warmth in context:
The compensatory nature of stereotypic views of national groups. European Journal of
Social Psychology,38, 1175–1183.
Kervyn, N., Yzerbyt, V., & Judd, C. (in press). When compensation guides inferences: Indirect
and implicit measures of the compensation effect. European Journal of Social Psychology.
Kervyn, N., Yzerbyt, V., Judd, C., & Nunes, A. (2009b). A question of compensation: The
social life of the fundamental dimensions of social perception. Journal of Personality and
Social Psychology,96, 828–842.
Kinder, D. R., & Sears, D. O. (1985). Public opinion and political action. In G. Lindzey & E.
Aronson, (Eds.), Handbook of social psychology (3rd ed., vol. 2). New York: Random
House. Knutson, B. (1996). Facial expressions of emotion influence interpersonal trait
inferences. Journal of Nonverbal Behavior,20, 165–182.
Leach, C. W., Ellemers, N., & Barreto, M. (2007). Group virtue: The importance of morality
(vs. competence and sociability) in the positive evaluation of in-groups. Journal of
Personality and Social Psychology,93, 234–249.
Livingstone, R. W., & Pearce, N. A. (2009). The teddy-bear effect: Does having a baby face
benefit black chief executive officers? Psychological Science,10, 1229–1239.
Miller, A. H., & Miller, W. E. (1976). Ideology in the 1972 election: Myth or reality – A
rejoinder. American Political Science Review,70, 832–849.
Montepare, J. M., & Dobish, H. (2003). The contribution of emotion perceptions and their
overgeneralizations to trait impressions, Journal of Nonverbal Behaviour,27, 237–254.
Mussweiler, T. (2007). Assimilation and contrast as comparison effects: A selective accessibility
model. In D. A. Stapel & J. Suls (Eds.), Assimilation and contrast in social psychology (pp.
165–186). New York: Psychology Press.
Peeters, G. (1992). Evaluative meanings of adjectives in vitro and in context: Some theoretical
implications and practical consequences of positive–negative asymmetry and behavioural-
adaptive concepts of evaluation. Psychologica Belgica,32, 211–231.
Peeters, G. (2005). Communion (Solidarity) and Power conveyed by social relations: A matter of
content or structure? Paper prepared for Transfer of Knowledge Conference of the European
Social Cognition network, Vitznau, Switzerland, 1–4 September.
Phalet, K., & Poppe, E. (1997). Competence and morality dimensions in national and ethnic
stereotypes: A study in six eastern-European countries. European Journal of Social
Psychology,27, 703–723.
Poppe, E., & Linssen, H. (1999). In-group favouritism and the reflection of realistic dimensions
of difference between national states in Central and Eastern European nationality
stereotypes. British Journal of Social Psychology,38, 85–102.
Read, S. J., & Miller, L. C. (1989). Inter-personalism: Towards a goal-based theory of persons
in relationships. In L. Pervin (Ed.), Goal concepts in personality and social psychology (pp.
413–472). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.
Reeder, G., & Brewer, M. (1979). A schematic model of dispositional attribution in
interpersonal perception. Psychological Review,86, 61–79.
Reinhard, M., Stahlberg, D., & Messner, M. (2008). Failure as an asset for high-status persons –
Relative group performance and attributes occupational success. Journal of Experimental
Social Psychology 44, 501–518.
186 KERVYN, YZERBYT, JUDD
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
Reinhard, M., Stahlberg, D., & Messner, M. (2009). When failing feels good—Relative
prototypicality for a high-status group can counteract ego-threat after individual failure.
Journal of Experimental Social Psychology,45, 788–795.
Rosenberg, S., Nelson, C., & Vivekananthan, P. (1968). A multidimensional approach to the
structure of personality impressions. Journal of Personality and Social Psychology,9, 283–
294.
Semin, G., & Fiedler, K. (1988). The cognitive functions of linguistic categories in describing
persons: Social cognition and language. Journal of Personality and Social Psychology,54,
558–568.
Sherif, M., & Hovland, C. I. (1961). Social judgement: Assimilation and contrast effects in
communication and attitude change. New Haven, CT: Yale University Press.
Snyder, M. (1984). When belief creates reality. In L. Berkowitz (Ed.), Advances in experimental
social psychology (Vol. 18, pp. 247–305). Thousand Oaks, CA: Academic Press.
Snyder, M., & Swann, W. (1978). Hypothesis-testing processes in social interaction. Journal of
Personality and Social Psychology,36, 1202–1212.
Snyder, M., Tanke, E., & Berscheid, E. (1977). Social perception and interpersonal behaviour:
On the self-fulfilling nature of social stereotypes. Journal of Personality and Social
Psychology,35, 656–666.
Srull, T., & Wyer, R. (1989). Person memory and judgement. Psychological Review,96, 58–83.
Thorndike, E. (1920). Intelligence and its use. Harper’s Magazine,140, 227–235.
Triandis, H. (1995). Individualism and collectivism. Boulder, CO: Westview Press.
Wigboldus, D. H. J., Spears, R., & Semin, G. R. (2005). When do we communicate stereotypes?
Influence of the social context on the linguistic expectancy bias. Group Processes and
Intergroup Relations,8, 215–230.
Wojciszke, B. (1994). Multiple meanings of behaviour: Construing actions in terms of
competence or morality. Journal of Personality and Social Psychology,67, 222–232.
Wojciszke, B. (2005). Morality and competence in person and self perception. European Review
of Social Psychology,16, 155–188.
Wojciszke, B., Abele, A. E., & Baryla, W. (2009). Two dimensions of interpersonal attitudes:
Liking depends on communion, respect depends on agency. European Journal of Social
Psychology,39, 973–990.
Yzerbyt, V., Kervyn, N., & Judd, C. (2008). Compensation versus halo: The unique relations
between the fundamental dimensions of social judgement. Personality and Social Psychology
Bulletin,34, 1110–1123.
Yzerbyt, V., Provost, V., & Corneille, O. (2005). Not so competent but warm . . . Really?
Compensatory stereotypes in the French-speaking world. Group Processes and Intergroup
Relations,8, 291–308.
Zanna, M., & Hamilton, D. (1972). Attributes dimensions and patterns of traits inferences.
Psychonomic Science,27(6), 343–354.
WARMTH AND COMPETENCE 187
Downloaded By: [Universite Catholique de Louvain UCL] At: 15:04 16 December 2010
... It is now well established that self-, other-and groupperceptions are structured around two basic and fundamental dimensions (e.g., Abele et al., 2008;Abele & Wojciszke, 2013;Fiske et al., 2002;Kervyn, Yzerbyt & Judd, 2010). Labelled communion and agency (Abele & Wojciszke 2007) or warmth and competence (Fiske et al. 2002;Kervyn, Yzerbyt & Judd, 2010), these two basic dimensions of social judgment and their specificities have recently been integrated into a new collaborative model that defines them as the horizontal and vertical dimensions (Abele et al., 2021). ...
... It is now well established that self-, other-and groupperceptions are structured around two basic and fundamental dimensions (e.g., Abele et al., 2008;Abele & Wojciszke, 2013;Fiske et al., 2002;Kervyn, Yzerbyt & Judd, 2010). Labelled communion and agency (Abele & Wojciszke 2007) or warmth and competence (Fiske et al. 2002;Kervyn, Yzerbyt & Judd, 2010), these two basic dimensions of social judgment and their specificities have recently been integrated into a new collaborative model that defines them as the horizontal and vertical dimensions (Abele et al., 2021). The horizontal dimension describes people in a 'relational way' and captures their morality and friendliness (Abele et al., 2021). ...
... However, we did not expect this obvious and consistent opposite pattern when compared with assertiveness self-evaluations. Researchers have highlighted the process of compensation between agency and communion (Holoien & Fiske, 2013;Judd et al., 2005;Kervyn, Yzerbyt & Judd, 2010). This somewhat unexpected result invites exploration of the possible existence of compensation processes between the assertiveness and ability facets of agency as a means of supporting a meritocratic vision of the socioeconomic world (Jost & Banaji, 1994;Ledgerwood et al., 2011). ...
Article
Full-text available
Relying on the Big Two framework (Abele et al., 2016, 2021) and the distinction of agency into the facets of assertiveness and ability, three experimental studies address the hypothesis that assertiveness and ability are influenced differentially by the consequences of success or failure. In Studies 1 and 2, participants had to imagine presenting a product developed by a hospital to an audience while either knowing or not knowing that selling the product could have strong positive consequences for the hospital's budget. They further had to imagine that they had succeeded in positively presenting the product or that they had failed. Study 2 replicated the design with the participants enacting the task for real. Supporting our hypotheses, we consistently found that self-evaluation of assertiveness was higher with both success and knowledge about the economic consequences, whereas self-evaluation of ability was higher with success but without knowledge of economic consequences. These findings support the facet approach of the agency dimension and give hints on how the assertiveness versus ability facets of self-evaluation differ.
... Compared with a blank wall or virtual background, viewing the speakers' actual room led to higher impressions of trustworthiness (73%), authenticity (65%), expertise (52%) (Zandan and Lynch, 2020) leading to some experts recommending using objects to frame positive evaluations (Pelta, 2022). This compromise parallels social research that has noted a compensation effect between judgements of competence and trustworthiness (Holoien & Fiske, 2013;Kervyn et al., 2010). ...
Preprint
Full-text available
Trait inferences from first impressions are drawn rapidly and spontaneously. However, the Covid-19 pandemic forced interactions online introducing differential influential factors on first impressions. As such, there is an absence of research investigating video background on videoconferencing impression formation. This study explored the influence of video background, facial expression, and gender on first impressions of trustworthiness and competence. Video background affected trustworthy and competence perceptions with Plants, Books and Blank backgrounds scoring highly on both dimensions while the Home, Blurred Home and Novelty backgrounds consistently received the lowest ratings. Happy faces were perceived as more trustworthy and more competent while female faces were also rated as more trustworthy and more competent, regardless of the background they were using. The explanations for these findings are discussed, along with future directions for research and the implications for videoconferencing use.
... Pour appréhender l'organisation de ces croyances, les travaux princeps de Rosenberg et collaborateurs (1968) Abele & Wojciszke, 2014 ;Fiske, 2015 ;Yzerbyt, 2016 Bien que ces deux dimensions émergent immanquablement lorsqu'il s'agit de juger un membre d'un groupe, des preuves empiriques suggèrent que les jugements sur la dimension horizontale ont la primauté sur ceux relevant de la dimension verticale, à la fois car ils s'activent plus rapidement et parce qu'ils ont plus de poids dans les réactions affectives et comportementales (Cacioppo, et al., 1997 ;Peeters, 2001 ;Richetin et al., 2012 ;Roussos & Dunham, 2016 ;Wojciszke et al., 1993Wojciszke et al., , 1998Zhang & Wang, 2018, pour des données en imagerie cérébrale, voir Li et al., 2021). A noter que certains auteurs indiquent plutôt une priorité de la dimension verticale lorsqu'il s'agit de comparer les groupes entre eux (Kervyn et al., 2010 ; ou lorsque l'évaluation sociale inclut le Soi (Abele & Wojciszke, 2007Koch et al., 2016). ...
Thesis
Full-text available
Ce travail de thèse est pensé comme une contribution originale à la recherche sur le handicap, en adoptant une approche neuro-socio-cognitive de la perception sociale. L’objectif général est de poursuivre l’analyse des facteurs contribuant à comprendre les barrières à l’inclusion sociale auxquelles font face les personnes en situation de handicap. Dans ces recherches, nous discutons l’opérationnalisation du handicap comme une catégorie univoque. Nous abordons également la question centrale de la mesure en psychologie sociale et comment celle-ci peut approcher au mieux la réalité sociale. A partir de ces réflexions, ce travail de thèse suit une ambition double. D’une part, élargir l’objet d’étude du handicap en prenant en compte l’(in)visibilité de la déficience. D’autre part, diversifier les outils d’investigation concernant notre objet d’étude. Pour répondre à ceci, la thèse s’organise autour de deux volets principaux. Le premier volet, centré sur la perception de Soi des individus en situation de handicap, propose quatre études dont deux font l’objet de publications. Globalement, les résultats mettent en évidence le rôle déterminant de facteurs subjectifs, tels que les jugements de soi en terme de compétence. Le deuxième volet de cette thèse se focalise sur l’étude de la perception sociale à l’égard des personnes en situation de handicap. Il s’organise autour de sept études dont quatre s’inscrivent dans deux publications. Les résultats mettent en évidence l’impact du handicap sur la perception d’autrui, modulé par la visibilité de celui-ci. Pour mettre en évidence l’ensemble de ces résultats, les études se sont appuyées sur des outils relevant de la psychologie cognitive, de la cognition sociale et des neurosciences sociales. Cette approche neuro-socio-cognitive permet d’aborder les questions soulevées par le handicap de façon transversale tel que le préconise actuellement la recherche dans ce domaine.
... Agency refers to the notion of competency or self-efficacy, whereas communion refers to the concept of interpersonal connectedness or warmth toward others (Grant and Gino, 2010). As an agentic-oriented emotion (Agrawal et al., 2013), pride tends to foster one's focus on internal feelings and signals a valued self-achievement (Agrawal et al., 2013) based on the assessment of one's ability (Kervyn et al., 2010). In contrast, empathy is communal in nature. ...
Purpose Focusing on the corporate social responsibility (CSR) communication context, the present research aims to understand when and why featuring pride versus empathy in a hospitality brand’s social media post can effectively boost consumers’ loyalty intention. Design/methodology/approach Two experimental studies examined the congruence effects between emotional appeal and sense of power, where power was made situationally salient within the social media post (Study 1) or measured as a personality trait (Study 2). Findings Emotional appeals featuring pride (vs empathy) will lead to higher loyalty intention for individuals with a situational or chronic sense of high (vs low) power. A further examination into the psychological mechanism reveals that such congruence effects are serially mediated through consumers’ perceived brand authenticity and brand trustworthiness. Practical implications Understanding how the sense of power may influence consumer response to social media posts using different emotional appeals can provide useful guidance for marketers about how to creatively segment customers and curate appropriate targeting messages for effective CSR communication and relationship building on social media. Originality/value Extending the message framing research on schema congruity, this research is the first to reveal the congruence effects of emotional appeal and sense of power in CSR communications and uncover the serial mediating roles of perceived brand authenticity and brand trustworthiness in relationship marketing on social media.
... The morality and competence dimensions showed a positive linear relationship, indicating that more morally positive behavior statements were rated as more competent. This replicates prior research [29,42,43], and suggests a halo effect [10,44] whereby favorable judgments on the morality dimension positively influence judgements on the competence dimension (or vice versa). The informativeness dimension showed a quadratic relationship to the morality and competence dimensions: behavior statements rated as more extreme in morality or competence (i.e., extreme positive or extreme negative) were associated with an increase in informativeness. ...
Article
Full-text available
To investigate impression formation, researchers tend to rely on statements that describe a person’s behavior (e.g., “Alex ridicules people behind their backs”). These statements are presented to participants who then rate their impressions of the person. However, a corpus of behavior statements is costly to generate, and pre-existing corpora may be outdated and might not measure the dimension(s) of interest. The present study makes available a normed corpus of 160 contemporary behavior statements that were rated on 4 dimensions relevant to impression formation: morality, competence, informativeness, and believability. In addition, we show that the different dimensions are non-independent, exhibiting a range of linear and non-linear relationships, which may present a problem for past research. However, researchers interested in impression formation can control for these relationships (e.g., statistically) using the present corpus of behavior statements.
... Agency refers to the notion of competency or self-efficacy, whereas communion refers to the concept of interpersonal connectedness or warmth toward others (Grant and Gino, 2010). As an agentic-oriented emotion (Agrawal et al., 2013), pride tends to foster one's focus on internal feelings and signals a valued self-achievement (Agrawal et al., 2013) based on the assessment of one's ability (Kervyn et al., 2010). In contrast, empathy is communal in nature. ...
Purpose – Focusing on the CSR communication context, the present research aims to understand when and why featuring pride versus empathy in a hospitality brand’s social media post can effectively boost consumers’ loyalty intention. Design/methodology/approach – Two experimental studies examined the congruence effects between emotional appeal and sense of power, where power was made situationally salient within the social media post (Study 1) or measured as a personality trait (Study 2). Findings – Emotional appeals featuring pride (vs. empathy) will lead to higher loyalty intention for individuals with a situational or chronic sense of high (vs. low) power. A further examination into the psychological mechanism reveals that such congruence effects are serially mediated through consumers’ perceived brand authenticity and brand trustworthiness. Originality/value – Extending the message framing research on schema congruity, this research is the first to reveal the congruence effects of emotional appeal and sense of power in CSR communications and uncovered the serial mediating roles of perceived brand authenticity and brand trustworthiness in relationship marketing on social media. Practical implications – Understanding how the sense of power may influence consumer response to social media posts using different emotional appeals can provide useful guidance for marketers about how to creatively segment customers and curate appropriate targeting messages for effective CSR communication and relationship building on social media.
Article
When choosing between political candidates of different educational levels, do voters show ingroup bias or base their vote choice on candidates’ perceived competence? We aim to investigate how (fictional) political candidates of different educational levels are evaluated and voted for, how this is affected by voters’ educational level, and the role of perceived (Study 1) and manipulated competence (Study 2). Higher educated participants preferred higher to less educated candidates over and above their level of competence, particularly when they identified strongly with their educational level. This reflects ingroup bias among the higher educated. Less educated participants preferred higher educated candidates in Study 1, but did not prefer higher educated candidates when competence was manipulated independently from education in Study 2. The less educated, unlike the higher educated, therefore, seem to show deference to the assumed competence of the higher educated, because it disappears when more reliable competence information is available.
Thesis
Full-text available
Bu tezde, kalıpyargı içeriği modeli (KİM) ve gruplar arası duygu ve kalıpyargı temelli davranış haritasının (yanlılık haritası) öngörülerini Türkiye'deki Suriyeliler bağlamında test etmek, kalıpyargı-davranış ilişkisinde duyguların aracı rolüne ilişkin tutarsız bulguların yanlılık haritası yazınının veri toplama, analiz stratejisi ve yetersiz istatiksel güçle ilgili yöntemsel sınırlılıklarıyla olan ilişkisini incelemek ve bu modellerde işaret edilen kalıpyargı-davranış ilişkisindeki meşrulaştırıcı mekanizmayı ahlaki dışlama açısından test etmek amaçlanmaktadır. İlk çalışmada, katılımcılar (N = 217) 20 Suriyeli alt-grubu KİM'in yapısal, kalıpyargı içeriği boyutları ve duygu değişkenleri açısından değerlendirmiştir. Beklendiği gibi, grupların çoğunluğu boyutların birinde yüksek diğerinde düşük algılanarak acımanın veya hasedin; geriye kalanı ise her iki boyutta yüksek veya düşük algılanarak hayranlığın veya hor görmenin belirgin hedefi olmuştur. Rekabet, kapsamı sembolik ve gerçekçi tehditlerle genişletildiğinde cana yakınlığı daha iyi yordamıştır. İkinci çalışmada, katılımcılar (N = 528) Suriyeli alt-grupları kalıpyargı içerikleri, duygu, davranışsal eğilim ve ahlaki dışlama değişkenleri açısından değerlendirmiştir. Yanlılık haritasıyla uyumlu olarak, acıma yüksek cana yakınlık-düşük yetkinlik kümesinde cana yakınlık-aktif yarar ilişkisine, hor görme ise düşük cana yakınlık-düşük yetkinlik kümesinde cana yakınlık-aktif zarar ilişkisine anlamlı düzeyde aracılık etmiştir. Ancak, hayranlık ve haset diğer iki kümede cana yakınlık-davranış ilişkilerine anlamlı düzeyde aracılık edememiştir. Yetkinlik-davranış ilişkilerine de hiçbir kümede hiçbir duygu anlamlı düzeyde aracılık edememiştir. Ahlaki dışlamanın aracı etkisi ise yalnızca çelişik/karışık kalıpyargı kümelerinde anlamlılık düzeyine ulaşabilmiştir. Birlikte değerlendirildiğinde, bulguların KİM'in kültürler arası genellenebilirliğini ve Türkiye'deki Suriyeliler bağlamına uygulanabilirliğini desteklediğini, duyguların aracı rolüne ilişkin tutarsız bulguların bu araştırmada etkisi araştırılan yöntemsel sınırlılıklarla ilgili olmayabileceğini ve davranışları meşrulaştırma ihtiyacının çelişik/karışık kalıpyargılanan kümelerde ortaya çıkma eğiliminde olduğunu işaret ettiği söylenebilir.
Article
Full-text available
To investigate the role of competence and morality in stereotypes, a cross-national research was set up in six eastern-European countries. Study 1 measured the perceived desirability of stereotype attributes in ingroup versus outgroup members. Across countries, biopolar Competence and Morality components emerged. It was found that the perceived desirability of ingroup attributes was primarily competence-based, while desirability perceptions of outgroup attributes were mostly morality-based. In Study 2, participants in the six countries rated the occurrence of competence- and morality-related stereotype attributes among 10 national and ethnic target groups. Study 2 also assessed general evaluative attitudes and perceptions of power and conflict in inter-nation relations. Competence and morality dimensions fully explained the evaluative structure of national and ethnic stereotypes, generating a four-fold typology of sinful–loser, sinful–winner, virtuous–loser and virtuous–winner stereotypes. This typology was strongly related to perceptions of power and conflict between national groups. © 1997 John Wiley & Sons, Ltd.
Article
Full-text available
Consistent with a theory stressing the behavioral-adaptive function of cognitive-affective biases, evaluative meanings of adjectives in isolation were found to reflect the meanings endorsed by the adjectives in positive rather than in negative contexts. Dramatic shifts of meaning were observed in negative contexts. However, those negative context effects seemed restricted to one specific evaluative dimension (good/bad for the self or 'self-profitability') that was to be distinguished from another, more genuinely evaluative, dimension (good/bad for others or 'other-profitability'). The psychological relevance of both dimensions is established and practical implications regarding the construction and interpretation of evaluative rating scales are presented
Article
Stereotype research emphasizes systematic processes over seemingly arbitrary contents, but content also may prove systematic. On the basis of stereotypes' intergroup functions, the stereotype content model hypothesizes that (a) 2 primary dimensions are competence and warmth, (b) frequent mixed clusters combine high warmth with low competence (paternalistic) or high competence with low warmth (envious), and (c) distinct emotions (pity, envy, admiration, contempt) differentiate the 4 competence-warmth combinations. Stereotypically, (d) status predicts high competence, and competition predicts low warmth. Nine varied samples rated gender, ethnicity, race, class, age, and disability out-groups. Contrary to antipathy models, 2 dimensions mattered, and many stereotypes were mixed, either pitying (low competence, high warmth subordinates) or envying (high competence, low warmth competitors). Stereotypically, status predicted competence, and competition predicted low warmth.
Article
To investigate the role of competence and morality in stereotypes, a cross-national research was set up in six eastern-European countries. Study 1 measured the perceived desirability of stereotype attributes in ingroup versus outgroup members. Across countries, biopolar Competence and Morality components emerged. It was found that the perceived desirability of ingroup attributes was primarily competence-based, while desirability perceptions of outgroup attributes were mostly morality-based. In Study 2. participants in the six countries rated the occurrence of competence-and morality-related stereotype attributes among 10 national and ethnic target groups. Study 2 also assessed general evaluative attitudes and perceptions of power and conflict in inter-nation relations. Competence and morality dimensions fully explained the evaluative structure of national and ethnic stereotypes, generating a four-fold typology of sinful-loser, sinful-winner, virtuous-loser and virtuous-winner stereotypes. This typology was strongly related to perceptions of power and conflict between national groups.
Article
Rosenberg, Nelson, & Vivekananthan (1968) have suggested that Asch’s (1946) “warm-cold” effect can be meaningfully interpreted within the framework of two dimensions (social and intellectual desirability) which underlie personality impressions. Their analysis suggested that the manipulation of unique information on one dimension would affect trait inferences on that dimension alone. This hypothesis was tested by replicating and extending Asch’s experiment in terms of these two dimensions. The results strongly supported the hypothesis.