Categorization in everyday life: The effects of positive and
negative categorizations on emotions and self-viewsy
NAOMI ELLEMERS* AND
MANUELA BARRETO
Leiden University, The Netherlands
Abstract
This study investigates how everyday categorization experiences affect people's emotional responses
and self-views. A representative Dutch population sample (N  463) was asked to recount a situation in
which they were categorized by others. This resulted in a range of categories that were spontaneously
evoked by research participants. Participants were asked to think of a situation either where the
categorization resulted in negative or in positive expectations about the self. Positive categorization
elicited more positive emotions and agreement than negative categorization. However, when positive
expectations about the self were formed, people found it less easy to detect that these were based on
external categorizations, and were less likely to protest. Mediational analyses showed that because
detection was impaired, exposure to positive categorization resulted in lower self-confidence than
exposure to negative categorization. Copyright # 2006 John Wiley & Sons, Ltd.
Theory and research on the effects of social categorization (Tajfel, 1978) has mainly examined the
problems people encounter when they are categorized as members of socially devalued groups, such as
ethnic minorities (Ethier & Deaux, 1994), homosexuals (Simon et al., 1998), overweight women
(Crocker, Cornwall, & Major, 1993), or disabled people (Frable, Platt, & Hoey, 1998). The present
research takes a broader view, as it systematically compares how people's emotions and self-views are
affected by positive versus negative expectations that can be derived from the social categories people
belong to. Unlike previous researchers we will not limit ourselves to a single category membership, but
examine the broad range of categorizations that people can experience in everyday life.
European Journal of Social Psychology
Eur. J. Soc. Psychol. 36, 931­942 (2006)
Published online 23 August 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ejsp.333
*Correspondence to: Naomi Ellemers, Social and Organizational Psychology, Leiden University. P.O. Box 9555, 2300 RB
Leiden, The Netherlands. E-mail: Ellemers@fsw.leidenuniv.nl
y
The present research was carried out in the context of a public education project on categorization (SIRE-champagne `Denk
ruimer dan in hokjes').
Contract/grant sponsor: Dutch Science Foundation (NWO, Vernieuwingsimpuls).
Copyright # 2006 John Wiley & Sons, Ltd.
Received 22 April 2005
Accepted 22 December 2005
CATEGORIZATION IN EVERYDAY LIFE
Because social categorizations do not only apply to members of stigmatized groups, but the
experience of being categorized is common to all of us, the first goal of the present study is not to
focus on a single (devalued) group membership, but to investigate a variety of different social
categorizations that people experience in everyday life. Because social categorization is used to
simplify information about different people we encounter (Oakes, Haslam, & Turner, 1994), by
definition the resulting category-based expectations are not always descriptive of the particular
individual under consideration. This can be the case because even though the individual belongs to
the group, this does not necessarily imply that all group characteristics apply to each group member.
For instance, while men generally tend to be more competitive than women, the select group of
women who have been successful in a typically masculine work setting describe themselves as more
competitive than their male colleagues (Ellemers, Van den Heuvel, De Gilder, Maass, & Bonvini,
2004). Additionally, people may prefer to be categorized in a different group, because the way they
are categorized by others does not reflect self-chosen group memberships (Barreto & Ellemers, 2002,
2003). For instance, immigrants may think of themselves as members of the host community where
they live, while others continue to perceive them in terms of their country of origin (Barreto, Spears,
Ellemers, & Shahinper, 2003). Finally, people tend to resist being treated in terms of a category
membership they do not find relevant to the situation at hand, even when they would not contest the
fact that they belong to this particular social category in other situations (Ellemers, Spears, & Doosje,
2002). This is the case for instance when a female professional resents being treated in terms of her
gender in a work situation, while she is perfectly happy to be regarded as such when going out with
friends (Ellemers, De Gilder, & Haslam, 2004).
Thus, independently of the identity threat that results from the inclusion of an individual in a socially
disadvantaged or otherwise unattractive group (value threat), theorists have proposed that people can feel
threatened when they are viewed by others as interchangeable category members, in situations where they
think they should be treated as unique individuals, or as members of another category (categorization
threat; Barreto & Ellemers, 2003; Branscombe, Ellemers, Spears, & Doosje, 1999; Ellemers, Spears, &
Doosje, 2002). Because research so far has not systematically addressed the consequences of social
categorization in this more general sense, we aim to assess how people react when others consider them in
terms of a broader variety of categorizations, including seemingly `innocent' group memberships that can
elicit positive as well as negative expectations about individual group members.
In doing this, we will build on the work of Deaux, Reid, Mizrahi, and Ethier, 1995, who developed
a broad taxonomy of different kinds of social categorizations. On the basis of a previous investigation
where respondents had generated different identities (Deaux, 1991), Deaux et al. (1995) asked a
student sample to rate a total of 64 different social identities in terms of their similarity, and in terms
of characteristic trait properties. On the basis of these ratings, Deaux et al. (1995). distinguished
between five different clusters of social categorizations: (1) relationships (e.g., husband, daughter),
(2) vocation or avocation (e.g., salesperson, intellectual), (3) stigma (e.g., deaf, alcoholic), (4)
ethnicity/religion (e.g., Jewish, Hispanic), and (5) political affiliation (e.g., feminist, republican).
Deaux et al. (1995) conclude that the clarity with which these clusters emerge, attest to the
heterogeneous nature of social categorizations as well as the social identities that are derived from
them, and emphasize that this should be taken into account in future research. In the present study,
where we explicitly address the question whether similar psychological processes may operate across
a broad range of social categorizations, we use this taxonomy to organize the variety of social
categorizations that is spontaneously mentioned by our research participants into a limited number of
classes.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
932 Naomi Ellemers and Manuela Barreto
POSITIVE VERSUS NEGATIVE GROUP-BASED EXPECTATIONS
Research shows that group members have mental images of the stereotypes and expectations that others
associate with their group (e.g., Sigelman & Tuch, 1997). These meta-stereotypes may be positive,
neutral, or negative in valence, and just like ingroup and outgroup stereotypes they are typically
sensitive to contextual variations (Vorauer, Main, & O'Connel, 1998). Departing from this knowledge,
the second goal of our investigation is to systematically compare the effects of positive versus negative
category-based (meta-) expectations for the self. As we indicated above, in most research to date, social
categorization was expected to have harmful effects for individual group members because the groups
that were examined were characterized by socially devalued features (as in research on social
stigmatization, Crocker, Major, & Steele, 1998), or because of immoral behavior of the group in the
past (as in research on collective guilt; e.g., Doosje, Branscombe, Spears, & Manstead, 1998). In the
present investigation we aim to examine the broader effects of social categorization, by systematically
comparing the effects of positive and negative category-based expectations.
Recent research on different forms of prejudice against women has revealed that views emphasizing
positive characteristics of women (e.g., benevolent sexism) are regarded as more acceptable and less
offensive than negative expressions of sexism (e.g., hostile sexism, see Barreto & Ellemers, 2005b; Glick &
Fiske, 1996). Nevertheless, such category-based expectations can still be harmful for the well being of the
individuals who are implicated in these views. Furthermore, recent research on implicit sexism revealed that
even when certain acts or statements of others are so subtly sexist that they are not recognized as
discriminatory by the women who are exposed to them, they do undermine subjective well-being, and
impair individual performance (Barreto & Ellemers, 2004, 2005; Barreto, Ellemers, & Palacios, 2004).
The present investigation builds on this previous research, by systematically comparing how
categorizations that evoke negative expectations about the individual (negative categorizations) and
categorizations that result in positive expectations about the individual (positive categorizations) affect
the individuals who are exposed to such categorization. Our main prediction is that, while both forms
can represent categorization threat, and as a result have the potential to undermine the individual self,
each form of categorization can have its own distinct effects. Thus, we predict that the specific effects of
categorization are not necessarily determined by the nature of the group one is categorized in, or even
by whether or not targets find this categorization contextually appropriate, but depend on whether this
is seen to lead to positive or negative expectations about the self.
Our reasoning is based on the assumption articulated in self-categorization theory (Turner, 1987),
that the comparative value of any social group is not an inherent feature of that particular group, but can
vary over time, across comparative domains, or across social contexts. Thus, similarly valenced
expectations can derive from different group memberships. Likewise, the same categorization can be
seen as a source of either positive or of negative group-based expectations, depending on the
comparison group and the comparative dimension that seems most relevant in that situation (Haslam &
Turner, 1992). For instance, psychology students can derive a positive identity from their group
membership when they compare to arts students in terms of intelligence, or when they compare to
science students in terms of creativity. However, this same categorization as a psychology student can
yield a negative identity when comparing to arts students in terms of creativity, or to science students in
terms of intelligence (Spears, Doosje, & Ellemers, 1997). Although these differences can be the mere
result of prejudiced beliefs, this is not always the case, as they often correspond to socially valid
differences between groups. We predict that each social category membership in principle can yield
positive as well as negative expectations about individual category members, and that the way people
respond to a particular categorization depends on the positive or negative valence of these group-based
expectations in a particular context, instead of being determined by the nature of the categorization.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
Categorization in everyday life 933
EFFECTS ON EMOTIONS AND SELF-VIEWS
Our third goal is to examine more explicitly than before how external categorizations impact upon
people's emotional responses and self-views. In parallel to our findings with respect to different forms
of sexism (Barreto & Ellemers, 2004, 2005a, 2005b), we hypothesize that a distinction should be made
between emotional responses to the categorization on the one hand, and the consequences of
categorization for people's self-views on the other. More specifically, we argue that­while positive and
negative categorizations both can have negative effects for the self­the nature and focus of these effects
differs depending on the extent to which the categorization is recognized as a (contextually)
inappropriate means of deriving expectations about the self.
When there is a discrepancy between the way people are treated by others and the way they prefer to
see themselves, they need to realize that this discrepancy derives from the behaviors of these others
(i.e., the act of categorization), to be able to maintain their internal self-views. In other words, to
maintain confidence in one's own sense of self, it is essential that people detect that categorization is
taking place (see also Barreto & Ellemers, 2004). This reasoning is based on previous work showing
that people tend to show confirmation of views about the self that are held by others when it is unclear
how these differ from self-views, whereas conditions that clarify discrepancies between self-views and
views held by others are more likely to result in self-verification responses (Kray, Thomson, &
Galinsky, 2001; Swann & Ely, 1984).We argue that the valence of the categorization (i.e., whether it
results in positive or negative expectations about the self) is an important factor that either facilitates or
impedes the recognition that categorization is taking place.
When categorization results in negative group-based expectations about the self, self-enhancement
motives are challenged (Sedikides, 1993). Thus, the desire to maintain a positive view of the self facilitates
the recognition that categorization is taking place, as should be evident from responses indicating
disagreement and detection. In addition, we expect that negative categorizations tend to elicit negative
emotional responses and lead the individual in question to engage in active coping strategies that focus on
the other, not the self (e.g, hostility and protest). However, when the external categorization elicits positively
evaluated group-based expectations about the self, the same self-enhancement tendencies will cause people
to find these expectations more acceptable as a description of the self (Kunda, 1990). Accordingly, they are
less likely to recognize that they are being categorized. Furthermore, because the positive nature of the
expectations induces agreement and elicit positive emotions, other-directed coping responses associated
with hostility and protest are less likely to be triggered (Schmitt, Ellemers, & Branscombe, 2003). Thus,
when focusing on emotional responses, positive categorization can be seen to have favorable consequences
as it tends to elicit positive emotions and induce a sense of agreement, whereas negative categorization
results in hostility and protest. However, the adverse effects of positive categorization emerge when
addressing people's self-views. That is, due to the greater difficulty to detect that categorization is taking
place we propose that when people who are subjected to positive category-based expectations, this affects
the confidence they have in their self-views, instead of calling into question the appropriateness of the
categorization to derive expectations about the self.
In sum, we predict that emotional responses and judgmental responses occur relatively
independently of each other (Barreto & Ellemers, 2005a). That is, whereas negative group-based
expectations on the one hand lead people to experience negative emotions, on the other hand the
negative valence of these expectations makes it easier to detect the fact that they are (inappropriately)
categorized, and helps them to maintain confidence in their own internal self-views. On the one hand
exposure to positive group-based expectations elicits positive emotional responses, on the other hand
people are less likely to detect that these favorable expectations about the self are category-based, and
as a result they will have less confidence in their internal self-views.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
934 Naomi Ellemers and Manuela Barreto
METHOD
Sample
The sample for this study was obtained through an internet community that is maintained by a national
advertising agency with the aim of obtaining representative opinions of the Dutch population on
various consumer products. From the complete database, consisting of 40,000 individuals, we drew a
stratified sample (N  463) for this study. This sample is representative for the Dutch population in
terms of gender (49% male, 51% female), level of education completed (14% primary education, 60%
secondary education, 20% higher professional education, 7% university education), and age category
(23% younger than 23, 24% between 23 and 33 years, 26% between 33 and 42 years, and 27% over 42
years).
Procedure
Members of the internet community from which we drew our sample, enter the website of the internet
community once a week, to complete that week's questionnaire on line. When participants in the
present study went on line, through their internet connection they received information explaining that
the purpose of the present investigation was to assess their responses to social categorization. Social
categorization was defined as a situation where one is labeled by others in a certain way, causing these
others to hold specific behavioral expectations about the self. It is however important to note that the
Dutch expression used to refer to external categorization (`in hokjes plaatsen') implicitly conveys
disagreement with that categorization. Participants were then invited to recount a situation in which
they were subjected to categorization by others (`in een hokje geplaatst'). Half the participants were
asked to think of a situation they had experienced where (unwanted) categorization resulted in positive
expectations about the self (positive categorization), and half the participants were instructed to evoke
an experience where categorization resulted in negative expectations about the self. This constituted
our experimental manipulation of positive versus negative categorization. Participants were then asked
to type-up a brief description of the situation they had in mind. Subsequently, the description they had
provided of their own positive or negative categorization experience was projected at the top of their
computer screen, and they were asked to keep this situation in mind when answering the questions that
followed. All questions were phrased in such a way that they asked participants to indicate how they felt
and how they responded in the situation that they had described.
Measures
All measures were taken by asking participants to indicate their agreement with specific statements on
rating scales, ranging from 1 (not at all) to 5 (very much).
Agreement with Categorization
We constructed six items to assess the extent to which research participants considered the
categorization appropriate (e.g., I belong in this category, alpha  0.89).
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
Categorization in everyday life 935
Detection
Three items intended to assess the extent to which participants were able to detect the fact that they had
been categorized (e.g., I immediately realized that I had been categorized; others had to point out to me
that I had been categorized (reverse-coded), alpha  0.62).
Emotions
With 16 emotion terms (adapted from PANAS, Watson, Clark, & Tellegen, 1988) respondents were
asked to indicate how they felt in the situation they had reported. Principal components analysis
revealed that these represented three orthogonal factors, accounting for 67% of the variance in the
individual items. Accordingly, three emotion scales were constructed: one containing eight positive
emotions (happy, proud, confident, strong, satisfied, secure, cheerful, elated; alpha  0.92), four
negative emotions representing feelings of hostility (angry, indignant, frustrated, irritated;
alpha  0.89), and four negative emotions associated with anxiety (ashamed, insecure, weak, tense;
alpha  0.82; see also Russell, 1980).
Coping Responses
Nine items intended to tap respondents' coping responses. A principal components analysis indicated
that these represented two orthogonal factors, which accounted for 58% of the variance in the
individual items. Five items captured the use of active coping strategies (e.g., protesting against the
categorization, alpha  0.84), whereas the remaining four items indicated the inclination to engage in
more passive coping (e.g., avoiding the situation, alpha  0.70).
Self-Views
Eight items intended to assess respondents' self-views. A principal components analysis indicated that
these represented two underlying factors, accounting for 49% of the variance in the individual items.
Accordingly, we constructed a four-item scale indicating participants' self-confidence (e.g., I can
achieve just as much as anyone else, alpha  0.62), and a four-item scale assessing participants' selfdoubt
(e.g., I feel I cannot live up to other people's expectations, alpha  0.62).
RESULTS
Types of Categorization
Based on the classification provided by Deaux et al. (1995), we coded the categorization experiences
that participants generated into five classes (see Table 1). This revealed that the majority of the
examples (about 65%) that were provided referred to relatively innocuous categorizations, based on
participants' vocations or avocations (e.g., truck driver, teacher, student), or on their personal
relationships (e.g., father, daughter), whereas a minority of the experiences (about 25%) were
associated with categories that tend to be seen as problematic, as stigmatized categories (e.g.,
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
936 Naomi Ellemers and Manuela Barreto
homosexual, elderly, handicapped), or ethnicity. Although the latter classes of categorizations (stigma
and ethnicity) were mentioned relatively more often in the negative categorization condition than in the
positive categorization condition (x2
(5)  40.07, p < 0.0001), still more than half of the participants
who recounted a negative experience referred to a categorization based on their (a)vocation or personal
relationships. That is, whereas previous research on stigmatization has emphasized the ways people can
suffer from their membership in groups they did not choose, because they were born in them (such as
ethnic groups), or were included due to factors beyond their control (as in the case of disabled people),
the everyday categorization experiences recounted by our participants indicate that even self-selected
category memberships, such as professional groups can be experienced as a source of negative
expectations about the self, and hence can represent identity threat.
Further inspection of the examples that were generated by participants revealed that the same
category membership was mentioned by some to illustrate the occurrence of negative group-based
expectations, while others referred to this same group membership to recount an experience where
categorization had induced positive expectations about the self. For instance, some indicated that
because they were civil servants, people thought they would be reliable (positive expectations), while
others indicated that their status as a civil servant made other people think that they would be lazy
(negative expectations). Likewise, the same category membership (overweight) was used to recount
positive group-based expectations (because I am fat, people think I am fun-loving) by some, whereas it
served as an example of negative categorization for others (I am fat, so people think I lack self-
discipline).
Thus, the coding of the categorization examples spontaneously provided by our research
participants first underline that categorization is an everyday experience. That is, even those who do not
belong to social groups that are considered as disadvantaged can and do experience that others
(unjustly) derive expectations about them, based on their social category membership. Second, these
results confirm the notion that in principle any social category can yield positive as well as negative
expectations, depending on the comparative dimension that is relevant and the comparative context that
is salient (e.g., Oakes, Haslam, & Turner, 1994).
Agreement with Categorization
To examine our predictions about the effects of positive versus negative categorization, we conducted
one-way ANOVA's on our dependent measures. For the indicator of agreement with the categorization,
this resulted in a significant effect of categorization condition (F(1,436)  123.78, p < 0.0001,
h2
 0.221). The relevant means (see Table 2) show that participants generally agreed more with the
positive categorization than with the negative categorization, as predicted. It is however important to
Table 1. Category frequencies, according to the classification developed by Deaux et al. (1995)
Positive Negative Total
Relationships 10 (7.1%) 24 (14.7%) 34 (11.2%)
(A)Vocation 97 (68.8%) 66 (40.5%) 163 (53.6%)
Stigma 18 (12.8%) 45 (27.6%) 63 (20.7%)
Ethnicity 0 (0%) 14 (8.6%) 14 (4.6%)
Political 1 (0.7%) 2 (1.2%) 3 (1.0%)
Other 15 (10.6%) 12 (7.4%) 27 (8.9%)
Total 141 163 304
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
Categorization in everyday life 937
note that the relatively low mean scores indicate that people generally considered the categorization
inappropriate rather than appropriate, irrespective of valence. This is consistent with what we
requested participants to recount, and indicates that the reactions described here must be seen as
reactions to category-based perceptions of self that deviate from situationally preferred self-views and
hence represent an identity threat, regardless of whether they elicit positive or negative expectations
from others.
Detection
A one-way ANOVA revealed a significant effect of categorization (F(1,435)  10.45, p < 0.001,
h2
 0.023), showing that detection was impaired in the positive categorization condition compared to
the negative categorization condition, as we predicted (see Table 2).
Emotions
The three composite emotion scores were subjected to a one-way (condition: positive/negative)
MANOVA, revealing significant effects of condition for all three emotions. These confirm that
participants who had been asked to recount an instance of positive categorization reported more
positive emotions than participants in the negative categorization condition (F(1,441)  119.12,
p < 0.0001, h2
 0.213). Additionally, less hostility-related emotions were reported in the positive
categorization condition than in the negative categorization condition (F(1,441)  112.07, p < 0.0001,
h2
 0.203). A similar effect was obtained for the anxiety-related emotions (F(1,441)  13.26,
p < 0.0001, h2
 0.029), although the latter effect was substantially smaller in size than the effects on
positive emotions or on hostility-related emotions.
Coping Responses
A MANOVA on the two composite scores indicated that, as predicted, respondents in the negative
categorization condition were more inclined to use active coping strategies than those in the positive
categorization condition (F(1,4408)  34.06, p < 0.0001, h2
 0.072), while participants in both
categorization conditions were equally likely to display passive coping (M  2.55, F < 1, ns).
Table 2. Mean differences for positive versus negative categorization
Positive categorization Negative categorization
Agreement 2.82 (0.82) 1.98 (0.76)
Detection 3.41 (0.85) 3.67 (0.83)
Emotions:
Positive 3.22 (0.75) 2.42 (0.78)
Hostility cluster 2.41 (0.96) 3.39 (0.97)
Anxiety cluster 2.38 (0.81) 2.69 (0.98)
Active coping 2.59 (0.85) 3.09 (0.94)
Self-confidence 3.65 (0.63) 3.78 (0.63)
Standard deviations are given in brackets.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
938 Naomi Ellemers and Manuela Barreto
Self-views
A one-way MANOVA on the two composite scores for self-confidence and self-doubt revealed that
participants in the negative categorization condition reported more self-confidence than those in the
positive categorization condition (F(1,441)  4.88, p < 0.028, h2
 0.011), as we predicted.
Categorization condition did not significantly affect reported levels of self-doubt (overall
M  2.98, F(1,441)  1.74, ns).
Mediation
Finally, through structural equation modeling (using EQS 6.0) we examined whether the observation
that positive categorization resulted in less self-confidence than negative categorization, could be
ascribed to the fact that positive categorization was less likely to be detected as such. For this analysis,
the categorization manipulation was dummy-coded, with positive categorization coded as 1, and
negative categorization as 2. First, we tested a model where detection mediated the relation between
categorization and self-confidence but no direct relation was allowed between categorization and selfconfidence.
This fully mediated model (x2
(19)  25.25, p  0.15) indicated that categorization was a
significant predictor of detection (b  0.19, p < 0.05), and that detection was a significant predictor of
self-confidence (b  0.19, p < 0.05). Furthermore, it showed a good fit to the data (Non-Normative Fit
Index NNFI  0.97; Comparative Fit Index CFI  0.98; Root Mean-Square Error of Approximation
RMSEA  0.027; Hu & Bentler, 1999). In the alternative model, where in addition to the mediated path
a direct relationship between categorization and self-confidence was allowed (x2
(18)  22.32,
p  0.22), this direct relation was not significant (b  0.11, ns), and indeed this more elaborate model
did not represent a significant improvement over the fully mediated model (Dx2
(1)  2.93, ns). From
these analyses, we conclude that the effect of positive versus negative categorization on participants'
self-confidence is fully mediated by the detection of the categorization. That is, the reason that people
respond with lower self-confidence to positive categorization than to negative categorization, is that
they find it more difficult to realize that they are being categorized in the case of positive categorization.
DISCUSSION
The present research has increased our understanding of the effects of categorization in everyday life.
First, because we asked research participants to generate their own examples of instances of
categorization, we were able to establish that people's everyday experiences are permeated with the
effects of categorization, even if we would not suspect this to be the case, because they do not belong to
the socially disadvantaged or stigmatized groups that are the focus of most research in this area. Indeed,
although some of our research participants clearly belong to stigmatized social groups (e.g., due to their
sexual preferences or physical disabilities), what strikes us is the relatively unproblematic and innocent
nature of most of the categorizations that were mentioned, which also included self-selected category
memberships that are generally seen as harmless, such as professional groups (Deaux et al., 1995;
Frable, 1993).
This complements what we know from previous research suggesting that categorizations based on
gender and ethnicity dominate the way we perceive others (Fiske, 1998). The argument here is that
these category memberships are readily visible, and can even be inferred from people's names or voices
when there is no face-to-face interaction. Furthermore, these categories are associated with a clear and
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
Categorization in everyday life 939
widely shared system of expectations about the traits and behaviors that are characteristic for the
members of these groups (Williams & Best, 1990). This is used to explain why gender and ethnic
categories tend to determine people's responses even in situations where they are clearly irrelevant
(Becker & Eagly, 2004). However, the present data show that a variety of social categories is used to
derive expectations about individuals, indicating that social categorization is a much more general
phenomenon.
Second, it is important to note that the identity of the group is not the decisive factor that determines
how people experience the fact that they are categorized. Instead, the predicted results of positive
versus negative categorization emerged across different category classes, and the same social
categories were mentioned as examples of negative as well as positive categorization. Indeed, the
consistency of the effects we observed is all the more meaningful, given the broad variety of
categorizations that was generated by the participants in this study. Furthermore, because research
participants indicated how they felt as a result of categorization experiences that had actually occurred
we feel confident that we have captured experiences that people considered subjectively meaningful,
which is less obvious when participants are required to focus on particular category memberships or on
traits or behaviors that researchers consider relevant.
A third important conclusion is that­as predicted­positive as well as negative categorization can
have negative effects for the self, even if these effects emerge in different ways. This is likely to be the
case when the categorizations are unwanted in a given context, irrespective of their valence. Indeed,
overall our participants indicated a lack of agreement with the fact that they were categorized,
indicating that the categorization in itself can constitute a source of identity threat, which should be
separated from the threat that stems from the association of the self with a lower status or morally
suspect social group (Branscombe et al., 1999; Ellemers et al., 2002). This complements recent
research demonstrating that positive versus negative expressions of sexism each in their own ways have
negative effects on the individuals who are exposed to such views (Barreto & Ellemers, 2003, 2005a,
2005b). The present investigation shows that similar processes can also emerge with respect to other
types of categorizations, and are not restricted to perceived gender differences.
The final conclusion from this study is that those categorizations that evoke the most negative
emotional responses (and generally tend to be seen as most problematic), at the same time are most easy
to cope with, as they elicit disagreement and invite protest (Deitch et al., 2003). By contrast, seemingly
favorable categorizations that are less easily detected are more likely to affect people's self-views as
they result in lower individual self-confidence. These findings clearly illustrate the relevance of
studying the range of responses of interest, as we cannot infer how people's self-views are affected by a
particular categorization from the emotional responses they show. Indeed, although some of the effects
reported here are relatively small, it is important to note that­compared to the way valence of
categorization affected people's emotional responses­we observed a significant reversal of the effect
when addressing their self-confidence. This is consistent with our research on expressions of sexism
(Barreto & Ellemers, 2004, 2005a) which also indicates that when group-based expectations are
presented in a `positive' way, positive emotional responses can co-occur together with negative effects
on self-confidence and task performance. Likewise, in a study where Asian American women were
subtly reminded that Asian Americans tend to have superior mathematical skills, these women showed
impaired performance on a maths test, which the researchers ascribed to the pressure of positive groupbased
expectations (Cheryan & Bodenhausen, 2000; Kray et al., 2001).
While research on stereotype threat has by now established that group-based expectations can
undermine task performance even for members of groups with no history of stigmatization or social
disadvantage (Maass & Cadinu, 2003), this evidence has been restricted to cases where the group
stereotype was negative with respect to the task domain (e.g., when men were compared to women in
terms of affective information processing, Leyens, De´sert, Croizet, & Darcis, 2000, or when whites
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
940 Naomi Ellemers and Manuela Barreto
were compared to blacks in terms of their athletic performance, Stone, Lynch, Sjomeling, & Darley,
1999). Thus, we think that the results of the present investigation indicating that positive stereotypes
can also be harmful for individual group members are novel and thought-provoking, and deserve
further attention in future research examining the conditions under which and the processes through
which this occurs.
ACKNOWLEDGEMENTS
This study was supported by funding from the Dutch Science Foundation (NWO, Vernieuwingsimpuls)
awarded to Manuela Barreto. We would like to thank Dana Muijlwijk, Ilse de Winne, and Daan Stam,
for their assistance in collecting, coding, and analyzing the data.
REFERENCES
Barreto, M., & Ellemers, N. (2002). The impact of respect vs. neglect of self-identities on identification and group
loyalty. Personality and Social Psychology Bulletin, 28, 493­503.
Barreto, M., & Ellemers, N. (2003). The effects of being categorised: The interplay between internal and external
social identities. European Review of Social Psychology, 14, 139­170.
Barreto, M., & Ellemers, N. (2004). The secret success of subtle sexism: How it contributes to the maintenance of
gender inequalities. Manuscript under editorial consideration.
Barreto, M., & Ellemers, N. (2005a). The perils of political correctness: Responses of men and women to oldfashioned
and modern sexist views. Social Psychology Quarterly, 68, 75­88.
Barreto, M., & Ellemers, N. (2005b). The burden of `benevolent' sexism: How benevolent sexism legitimizes
gender inequalities. European Journal of Social Psychology, 35, 633­642.
Barreto, M., Ellemers, N., & Palacios, M. (2004). The backlash of token mobility: The impact of past group
experiences on individual ambition and effort. Personality and Social Psychology Bulletin, 30, 1433­1445.
Barreto, M., Spears, R., Ellemers, N., & Shahinper, K. (2003). Who wants to know? The effect of audience on
identity expression among minority group members. British Journal of Social Psychology, 42, 299­318.
Becker, S. W., & Eagly, A. E. (2004). The heroism of women and men. American Psychologist, 59, 163­178.
Branscombe, N. R., Ellemers, N., Spears, R., & Doosje, B. (1999). The context and content of social identity threat.
In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity: Context, commitment, content (pp. 35­58).
Oxford: Basil Blackwell.
Cheryan, S., & Bodenhausen, G. V. (2000). When positive stereotypes threaten intellectual performance: The
psychological hazards of model minority status. Psychological Science, 11, 399­402.
Crocker, J., Cornwall, B., & Major, B. (1993). The stigma of overweight: Affective consequences of attributional
ambiguity. Journal of Personality and Social Psychology, 64, 60­70.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.),
Handbook of social psychology (4th ed., Vol. 2, pp. 163­177). Boston: McGraw-Hill.
Deaux, K. (1991). Social identities: Thoughts on structure and change. In R. C. Curtis (Ed.). The relational self:
Theoretical convergences in psychoanalysis and social psychology (pp. 77­93). New York: Guilford Press.
Deaux, K., Reid, A., Mizrahi, K., & Ethier, K. A. (1995). Parameters of social identity. Journal of Personality and
Social Psychology, 68, 280­291.
Deitch, E. A., Barsky, A., Butz, R. M., Chan, S., Brief., A. P., & Bradley, J. C. (2003). Subtle yet significant: The
existence and impact of everyday racial discrimination in the workplace. Human Relations, 56, 1299­1324.
Doosje, B., Branscombe, N. R., Spears, R., & Manstead, A. S. R. (1998). Guilty by association: When one's group
has a negative history. Journal of Personality and Social Psychology, 75, 872­886.
Ellemers, N., De Gilder, D., & Haslam, S. A. (2004). Motivating individuals and groups at work: A social identity
perspective on leadership and group performance. Academy of Management Review, 29, 459­478.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
Categorization in everyday life 941
Ellemers, N., Spears, R., & Doosje, B. (1997). Sticking together or falling apart: Group identification as a
psychological determinant of group commitment versus individual mobility. Journal of Personality and Social
Psychology, 72, 123­140.
Ellemers, N., Spears, R., & Doosje, B. (2002). Self and social identity. Annual Review of Psychology, 53, 161­186.
Ellemers, N., Van den Heuvel, H., De Gilder, D., Maass, A., & Bonvini, A. (2004). The underrepresentation of
women in science: Differential commitment or the Queen-bee syndrome? British Journal of Social Psychology,
43, 315­338.
Ethier, K. A., & Deaux, K. (1994). Negotiating social identity when contexts change: Maintaining identification
and responding to threat. Journal of Personality and Social Psychology, 67, 243­251.
Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.),
Handbook of Social Psychology (4th ed., pp. 357­411). New York: McGraw Hill.
Frable, D. E. S. (1993). Dimensions of marginality: Distinctions among those who are different. Personality and
Social Psychology Bulletin, 19, 370­380.
Frable, D. E. S., Platt, L., & Hoey, S. (1998). Concealable stigmas and positive self-perceptions: Feeling better
around similar others. Journal of Personality and Social Psychology, 74, 909­922.
Glick, P., & Fiske, S. (1996). The ambivalent sexism inventory: Differentiating hostile and benevolent sexism.
Journal of Personality and Social Psychology, 70, 491­512.
Haslam, S. A., & Turner, J. C. (1992). Context-dependent variation in social stereotyping 2: The relationship
between frame of reference, self-categorization and accentuation. European Journal of Social Psychology, 22,
251­277.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria
versus new alternatives. Structural Equation Modeling, 6, 1­55.
Kray, L. J., Thomson, L., & Galinsky, A. (2001). Battle of the sexes: Gender stereotype confirmation and reactance
in negotiations. Journal of Personality and Social Psychology, 80, 942­958.
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480­498.
Leyens, J. P., De´sert, M., Croizet, J. C., & Darcis, C. (2000). Stereotype threat: Are lower status and history of
stigmatization preconditions of stereotype threat? Personality and Social Psychology Bulletin, 26, 1189­1199.
Maass, A., & Cadinu, M. (2003). Stereotype threat: When minority members underperform. European Review of
Social Psychology, 14, 243­275.
Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality. Oxford: Blackwell.
Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39, 1161­1178.
Schmitt, M. T., Ellemers, N., & Branscombe, N. (2003). Perceiving and responding to gender discrimination at
work. In A. Haslam, D. Van Knippenberg, M. Platow, & N. Ellemers (Eds.), Social identity at work: Developing
theory for organizational practice. New York: Psychology Press.
Sedikides, C. (1993). Assessment, enhancement, and verification determinants of the self-evaluation process.
Journal of Personality and Social Psychology, 65, 317­338.
Sigelman, L., & Tuch, S. A. (1997). Metastereotypes: Black's perceptions of Whites' stereotypes of Blacks. Public
Opinion Quarterly, 61, 87­101.
Simon, B., Loewy, M., Stu¨rmer, S., Weber, U., Freytag, P., Habig, C., Kaupmeier, C., & Spahlinger, D. (1998).
Collective identification and social movement participation. Journal of Personality and Social Psychology, 74,
646­658.
Spears, R., Doosje, B., & Ellemers, N. (1997). Self-stereotyping in the face of threats to group status and
distinctiveness: The role of group identification. Personality and Social Psychology Bulletin, 23, 538­553.
Stone, J., Lynch, C. I., Sjomeling, M., & Darley, J. M. (1999). Stereotype threat effects on Black and White athletic
performance. Journal of Personality and Social Psychology, 77, 1213­1227.
Swann, W. B., & Ely, R. J. (1984). A battle of wills- self-verification versus behavioral confirmation. Journal of
Personality and Social Psychology, 46, 1287­1302.
Tajfel, H. (1978). Differentiation between social groups: Studies in the social psychology of intergroup relations.
New York: Academic Press.
Turner, J. C. (1987). A self-categorization theory. In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, & M. S.
Wetherell (Eds.), Rediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.
Vorauer, J. D., Main, K. J., & O'Connel, G. B. (1998). How do individuals expect to be viewed by members of
lower status groups? Content and implications of meta-stereotypes. Journal of Personality and Social
Psychology, 75, 917­937.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and
negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063­1070.
Williams, J. E., & Best, D. E. (1990). Measuring sex stereotypes: A multi-nation study. Newbury Park: Sage.
Copyright # 2006 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 36, 931­942 (2006)
942 Naomi Ellemers and Manuela Barreto