Children’s identification with the group and their
acquisition of self-relevant knowledge: the case of the nation
Poster presented at the 9th European Conference on Developmental
Psychology, Spetses, Greece, August 1999.
Mark Bennett Evanthia Lyons
University of Dundee University of Surrey
Fabio Sani Martyn Barrett
University of Dundee University of Surrey
Mark Bennett and Fabio Sani, Department of Psychology, University of Dundee,
Scotland; Martyn Barrett and Evanthia Lyons, Department of Psychology, University
of Surrey, England.
Sincere thanks are conveyed to all the teachers and children who participated in
our research. We are extremely grateful to Mariangela Bati, Sian French, Nick
Messing, Katie Neale, Sarah Scott, Dawn Taylor and Sophie Whitehouse for their
assistance in data collection and extend our warm thanks to them.
The research reported in this paper was supported by a grant received from the
Commission of the European Communities DGXII Human Capital and Mobility
(Networks) Programme (Grant No. CHRX-CT94-0687). We are greatly indebted to
our colleagues who contributed to the design of this research: Luciano Arcuri,
Almudena Gimenez de la Pena, Annamaria Silvana de Rosa and Ignasi Vila.
Correspondence should be directed to Mark Bennett, Department of Psychology,
University of Dundee, Dundee DD1 4HN, Scotland (electronic mail,
M.Bennett@Dundee.ac.uk)
1
Children’s identification with the group and their
acquisition of self-relevant knowledge: the case of the nation
Abstract
Much previous research has shown that children's gender self-categorization as male
or female plays an important role in guiding the acquisition of self-relevant
information. The present study seeks to establish whether this reflects a general
phenomenon in identity development and looks at an abstract and relatively less
salient aspect of identity than gender, viz. nationality. Subjective identification with
the national group and knowledge of one's own and other national groups were
examined in 459 children between 6 and 15 years. Data are presented which indicate
that even with a level of identity which is highly abstract and not chronically salient,
identification with the group is strongly associated with increased acquisition of self-
relevant information.
2
Introduction
Within developmental psychology, research concerned with the self has been largely
descriptive, focusing for example on the acquisition and development of children's
self-conceptions. Although much is known about the developmental course of the
self-concept (Damon & Hart, Livesley & Bromley, 1973; McGuire McGuire, 1987;
Peevers, 1987), relatively less is understood about the cognitive consequences of
variations in children's self-concepts. In short, little work has examined process-
related issues. Thus far, such research, guided by social identity theory (Tajfel &
Turner, 1986), has shown how children's self-conceptions influence their in- and
outgroup evaluations (Bennett, Lyons, Sani & Barrett, in press; Bigler, Jones &
Lobliner, 1997; Powlishta, 1994; Yee & Brown, 1992). However, the role of self-
processes in children's acquisition of social information has been little-explored by
developmental psychologists. It is this latter issue which the present study addresses.
Within social psychology, cognitive approaches to the self have given prominence
to self-schemas, that is, "cognitive generalizations about the self derived from past
experience that organize and guide the processing of self-related information"
(Markus, 1977, p. 63). There now exists a substantial body of research which
demonstrates that self-schemas are implicated in various forms of information-
processing (e.g., Markus, Crane, Bernstein, Siladi,1982). Markus (1977) has shown
that individuals who are schematic on particular personality dimensions, such as
independence, are able to make more rapid self-related judgments on those dimension
than on other dimensions. Self-schemas also guide judgment about others, so that for
example the "independence schematic" would be more likely to invoke this dimension
in the perception of others than would someone aschematic for that dimension (Park
& Hahn, 1988; Shapiro, 1988). For the present purposes, it is particularly important to
note that self-schemas play a role in the encoding and storage of information. Thus,
the independence schematic is much more likely than the aschematic to have a set of
memories of his or her behaviors which exemplify that dimension (Markus, 1977).
Similarly, "weight schematics" will typically acquire greater knowledge about dieting,
the caloric values of different foods, etc., than will aschematics (Fong & Markus,
1982).
Within developmental psychology, the major theory which has drawn attention to
the consequences of self-conceptions for information-processing is gender schematic
processing theory (Martin & Halverson, 1981). This theory proposes that children's
sex-typing begins with the mere categorization of the self as male or female.
Following this, and as a direct consequence of categorization of the self, children are
increasingly oriented towards socially available information with which to elaborate
their gender schema. Following an initial period of establishing rudimentary
characteristics of both male and female behavior, Martin and Halverson propose that
children focus increasingly upon information relevant to their own sex, thereby
becoming more expert with respect to what is considered appropriate to that category.
This theory has been well supported empirically (e.g. see Durkin (1995) for a
summary) and can account for a broad range of phenomena, such as self-evaluation in
the light of gender stereotypes about behavior (Cramer & Skidd, 1992), toy and
activity preferences based on gender-typed labeling ("this is for boys/girls") (Masters
et al, 1979), distortions in the assimilation of counter-stereotyped information (e.g.
taking a female doctor to be a nurse; Carter & Levy, 1988; Liben & Signorella, 1993),
and the finding that children who learn gender labels early in life know more about
3
what is appropriate for their sex than do those who learn labels later (Fagot, Leinbach
& O'Boyle, 1992).
Martin and Halverson explicitly present the theory as a domain-specific account,
that is, they do not make general claims about the role of identity schemas in
children's information processing. Arguably, however, the theory's basic assumptions
are applicable to other domains; children's self-categorization with respect to a
particular group may encourage the acquisition of knowledge relevant to that group.
The extent to which the theory's assumptions may be generally applicable may
nevertheless be circumscribed by the particular characteristics of different categories,
especially their abstractness. We propose that categorization at the level of gender
may be an optimal level in terms of demonstrating the role of schemas in the
acquisition of new information; other types of identity schema may not facilitate this
process to anything like the same extent. This possibility is suggested on the basis of
three principal considerations, which we now outline.
Martin and Halverson note that schemas differ in their salience and suggest that
gender schemas are likely to be particularly salient: "Gender is an obvious and stable
human characteristic. Other types of self-defining groupings such as national origin
are not so apparent" (p.1127). We contend that the central point here is that, as a
highly abstract and inclusive category, nationality has significantly lower utility than
does gender. Whereas gender is an effective basis for distinguishing between persons
within one's typical social contexts, the same is not true of nationality since the
overwhelming majority of one's peers, relatives, teachers, et. al. are likely to be of the
same nationality as oneself. As a basis for distinguishing others in one's everyday
contexts, then, nationality has very low utility indeed. Moreover, understanding of
one's national identity is relatively late-appearing, undergoing substantial development
between six and ten years (Lambert & Klineberg, 1967). Thus, nationality schemas are
unlikely to be as chronically salient as gender schemas and may not therefore play as
important a role in guiding cognition (1). In view of this difference it seems plausible
to argue, from a purely cognitive perspective, that the role of self-schemas in
information acquisition may be a function of categories' utility; categories of low
utility may play a less significant role in this respect than those of high utility.
Apart from cognitive considerations, it is important too to consider social factors,
since the cognitive processes associated with sex-typing are likely to be supported and
augmented by social processes (Maccoby, 1988). Thus, early self-labeling as a girl or
a boy, and subsequent sex-typed behavior, will be reinforced by agents of
socialization; similarly, sex-inappropriate behaviors will frequently be punished
(Fagot & Hagan, 1992). Although this general point is almost certainly true for all
socially significant categories, we suggest that it is particularly pertinent for gender
categories, since, as Banaji & Prentice (1994) have argued, gender is "the most
fundamental of human categories" (p. 315). To illustrate the point, consider the case
of a boy who fails to make sex-typed discriminations in play, for example between
dolls and airplanes. We suggest that such a child is much more likely to be subject to
censure than is the child who fails to make nationality-typed discriminations, perhaps
between pretending to be both Prime Minister Blair and President Clinton. That is,
norms surrounding gender are particularly potent, prescribing very strongly what is
appropriate and inappropriate behavior for category members (Bem, 1981). Such
prescriptions at the level of nationality are much less clearcut. Thus, we suggest that
in terms of adult demands upon children, those pertaining to gender are likely to be
more significant and frequent than those related to nationality (at least in the absence
4
of serious international conflicts). This bolsters our previous point and supports our
contention that gender schemas are likely to be more available than many others, and,
as such, that Martin and Halverson may have identified an optimal context for the
demonstration of the role that self-processes might have in the acquisition of
knowledge.
Finally, "unlike many other types of self-defining categories, gender categories
are... dichotomous. Information about either group can be used in defining the self"
(Martin & Halverson, 1981, p.1127). Thus, unlike a system of categorization based on
nationality, which involves a multiplicity of groups (and groups within groups, e.g.
England, Scotland and Wales as parts of Britain), that based on gender is both simple
and, potentially at least, directly informative with respect to the self.
All these considerations suggest that, in terms of guiding information-processing,
gender identification may differ from more abstract and inclusive forms of
categorization, such as at the level of nationality. We suggest that the generality of
this phenomenon in children's development should therefore be subject to empirical
study. Conceivably, this phenomenon may be limited to highly salient categories
which discriminate between others within one's immediate social sphere. Thus, the
study that follows examines the generality of the application of cognitive approaches
to the development of the self, looking at the extent to which children's knowledge of
the national group is a function of their identification of themselves as group-
members. In particular, it aims to determine whether children identifying with their
national group are more knowledgeable about their group than those not identifying
with it; and whether their knowledge is particular to the ingroup rather than outgroups,
as would be expected under Martin and Halverson's account. The study thus addresses
the issue of whether children's subjective identification as members of their national
group has a bearing on the acquisition of various types information relevant to the
group.
Method
Participants: The sample comprised 459 predominantly white British children drawn
from schools in the London and Dundee areas. Participants were divided into four
age-groups: 6-, 9-, 12- and 15-year-olds. There were 108 6-year-olds (
M = 78
months,
SD= 2.8; 51 boys & 57 girls), 110 9-year-olds (M = 113, SD= 2.9; 59 boys
and 51 girls), 122 12-year-olds (
M = 149 SD= 3.4; 62 boys and 60 girls) and 119 15-
year-olds (
M = 184, SD = 3.3; 61 boys and 58 girls). Children's names were drawn
randomly from class registers. The only criteria for inclusion in the sample were that
the children were British (either by birth, parentage or passport held)
Procedure: Children were seen individually as part of a broader study concerned with
children's beliefs and feelings about their own and other national groups. For the
purpose of the present study, two types of measure were employed, those concerned
with subjective identification with the group, and those examining knowledge of
national groups.
Subjective identification
Children were provided with two measures of subjective identification with the
national group, one open-ended, the other forced-choice. The former always preceded
the latter.
Open-ended task. Children were presented with 15 cards, each with a potential
self- descriptor on it: British, French, Italian, German, Spanish, Scottish, English,
5
Dundonian, Londoner, girl, boy, 6-years-old, 9-years-old, 12-years-old and 15-years-
old. The cards were spread, in a randomized order, on a table in front of the child. The
child's task was to identify only those cards which might be used to describe the self.
The interviewer instructed him or her to Have a look at these cards. All these words
can be used to describe people. Which ones do you think could be used to describe
you, which ones do you think are you? You can choose as many as you like. (And in
the case of 6- and 9-year-olds) Shall I help you to read them?
When the child had made his or her selection, the remaining cards were removed.
He or she was then asked to place all the selected cards on the table. The interviewer
then asked the child, If you had to choose just one of these cards because it was the
most important to you, which would you choose? The selected card was then removed
from the table and the procedure repeated until only one card remained.
Scoring for this task was in two stages, as follows. First of all, the first-ranked card
was assigned a score of one, the second 2, and so on. The value assigned to each of
the cards not chosen from the total set of 15 cards was the average of the remaining
ranks up to rank 15. For example, in the case of a child who had selected only 4 cards,
ranks 5 to 15 would remain. The average of these ranks is established as follows:
(5+6+7+8+9+10+11+12+13+14+15) / 11 = 10. The rank assigned to the British card
was then used as a basis for creating a factor reflecting the important of British
identity: Ranks 1 and 2 were taken as indicating that British identity was very
important; ranks 3-6 fairly important; less than 6 (i.e. in practice, unranked)
unimportant. The conversion of these ranks to a factor with three levels was to enable
comparison with data resulting from the forced-choice task.
Forced-choice task. In order to assess the degree of children's identification with
the category British, they were presented with the following four cards, placed in front
of them, from left to right: very British; little bit British; not at all British; don't know.
The interviewer asked, Which one do you think best describes you? Are you (pointing
to the cards) very British, a little bit British, not at all British, or don't you know?
Answers were recorded in terms of the response options indicated above. To enable
comparison with the previous variable, we combined those children who denied being
group members with those who expressed ignorance of group membership. This
decision was guided by social identity theory (Tajfel & Turner, 1986), the main
assumption of which is that to define the self in terms of the group, that is, at a social
categorical rather than personal, idiosyncratic level, is the psychological process
responsible for group behavior. Thus, the two response types (denial and ignorance of
group membership) are theoretically equivalent in that they both indicate a lack of
subjective identification with the group and therefore both imply the absence of
group-relevant cognitions and behavior.
To establish the extent of convergence between the two measures, a test of
association was conducted on responses to the open-ended and forced choice tasks.
Rankings of the British card were found to be highly associated with responses to the
question about the extent of Britishness: Cramers V = .227,
p< .0001. Thus, children
who ranked the British card highly were likely to respond that they felt very British;
those who did not rank the card, or who gave it a low ranking, were likely to respond
that they did not feel very British.
Knowledge of national groups
Children's knowledge of the following five European countries was assessed:
Britain, France, Germany, Italy and Spain. For each country, three distinct types of
6
knowledge were examined: geographical knowledge, knowledge of significant
symbols, and knowledge of famous people.
Geographical knowledge
Children were presented with an A3-sized outline map of Europe indicating national
boundaries. For each of the five countries (which were considered in a random order),
they were asked (in the following order):
a) to point to the named country. Responses were recorded as either correct or
incorrect. ('Don't know' responses were recorded as incorrect.)
b) to name the capital of the country. Again, responses were recorded as either correct
or incorrect.
c) to mark with a cross where they thought the capital of the country was. Only those
responses within a 30-mile radius of the capital were accepted as correct.
For each country, responses from each of these questions were summed to give a
measure of children's geographical knowledge.
Knowledge of symbols
Children were presented with a set of cards, each of which depicted what a group of
15 adults had judged to be the most prototypical exemplars for each country under the
following headings: buildings (e.g. for Britain, Tower Bridge; for France, the Eiffel
Tower), food (e.g. for Britain, an English breakfast of eggs, bacon, sausages, etc; for
Spain, paella), scene (e.g. for Britain, a thatched cottage in a rural village; for Italy, a
hilltown in the countryside of Tuscany) and event (e.g. for Britain, the Changing of
the Guards; for Germany, the Oktoberfest). To this set was added the national flag for
each of the countries.
Children were presented with the total card set (n = 25) in a random order and were
asked to place each picture into one of five boxes representing each of the countries
(i.e. the boxes were labeled Britain, France, Germany, Italy and Spain). There was
also a further box marked "Don't know".
For each country children were assigned a score between 0 and 5, depending on the
number of correct responses given.
Knowledge of famous people
For each country, children were asked to provide the names of as many famous
people as they could. Although the names given were typically familiar to
interviewers (usually being heads of state, royals, pop stars, or soccer players),
problematic cases where the interviewer had no knowledge of named person (less than
5% of the total) were referred to at least two colleagues in an attempt to determine
whether it was appropriate to record them as correct responses.
For each country, frequencies of between 0 and 5 correct responses were recorded.
Few children gave more than 4 or 5 names per country; where the number exceeded 5,
this was often the result of the names of entire soccer teams being given. To avoid
artificially inflating means with these outliers, it was thus decided to limit the highest
score to 5.
Results
Three principal sets of analyses will be addressed. First of all, attention will be given
to the relationship between age and level of identification with the national group.
Next, age and level of knowledge about each of the countries will be considered.
Since these initial analyses are not of central importance in the present context, they
will be covered relatively briefly as important precursors to our examination of the
relationship between level of identification and knowledge. In particular, these initial
7
analyses demonstrate the highly significant effect of age on both identification and
knowledge, and guide the form of statistical analysis employed to investigate the
hypothesized relationship between identification and knowledge.
Age and identification
Both measures of identification were found to be highly age-related. For the open-
ended measure, a Kruskal Wallis test was conducted on the rankings of the British
card. Older children ranked this card more highly than did younger children: X2 (3) =
26.12, p<0.001 (6-year-olds mean rank = 8.87 (i.e., unranked); 9-year-olds, 7.22 (i.e.,
unranked); 12-year-olds, 6.2; 15-year-olds, 6.2). For the fixed-choice question, age
was significantly associated with the level of identification: X2 (6) = 158.6,
p<0.0001. As can be seen from Table 1, whereas many of the youngest children failed
to identify with the national group, this was rare amongst the older children, who
typically identified with the group. It should be noted however that 12-year-olds
identified more strongly than did 15-year-olds: X(2) 2 = 6.3, p<0.05.
Table 1 about here
Age and knowledge
A multivariate analysis of variance for scores from the three knowledge measures,
over each of the countries, showed significant effects of age for every variable
(p<0.001). A summary of means, over age-groups, along with univariate F values for
every variable, is given in Table 2. Post hoc testing (Tukey's critical distance - CD)
showed that there were significant differences between all age groups (p<.05) for all
variables except geographical knowledge of Italy (where there were no differences
between 12- & 15-year-olds); knowledge of famous people with respect to France,
Italy and Germany, where there were no differences between 6- and 9-year-olds'
knowledge of famous people; and for knowledge of famous Spaniards, where there
were differences only between the two older and younger groups.
Table 2 about here
Identification and knowledge
Since age is so strongly related to both identification and knowledge, it is not
included as a factor in the following analyses. Rather, the relationship between
identification and knowledge is examined by means of analysis of covariance, with
age entered as a covariate. To permit comparability between the findings from the
open-ended and fixed-choice measures, the rankings from the open-ended task were
converted to a factor, where rankings of 1 or 2 were taken as indicating 'very British',
rankings 3-6 as 'a little bit British' and less than 6 as 'not at all British' (since no
children used more than 6 cards in the self-description task this would indicate that
the British card had not been selected).
Univariate F tests were conducted on each of the variables under the three
knowledge types. Knowledge of the ingroup was found to be strongly related to
identification with the ingroup. Thus, significant effects of British identification on
knowledge about Britain were found for both the open-ended and fixed-choice
measures of identification with respect to geographical knowledge, knowledge of
symbols, and knowledge of famous people (see Tables 3, 4 & 5 for means and F
values). Level of identification was not found to have consistent effects on knowledge
of countries other than Britain. Exceptions, however, were: knowledge of French and
Italian geography, which were found to be related to identification under both
measures; knowledge of Italian symbols under both measures; knowledge of French
symbols under the fixed choice measure; knowledge of famous Italians under the
8
open-ended measure, and knowledge of famous Germans when using the fixed choice
measure.
Tables 3, 4 & 5 about here
Post hoc testing showed that in all cases but one, where there was a significant effect
of level of identification, differences were between children not identifying with the
category "British" and those who identified with it to a greater or lesser degree.
(p<.05). The exception was with respect to knowledge of famous Germans, where,
when considering the fixed-choice measure of identification, it was found that those
children who described themselves as "a little bit British" typically knew more names
than did those who described themselves as "very British"(p<.05).
Discussion
The results of this study indicate that knowledge of the ingroup is related to
identification with the ingroup, even where that group is a highly abstract one, as in
the case of the nation. Thus, significant effects of national identification (as measured
by two independent techniques) on knowledge about one's nation were found for three
different types of knowledge: geographical knowledge, knowledge of symbols, and
knowledge of famous people. In each case, children identifying with the national
group knew more than those not identifying (though degree of identification appeared
not to be important). Moreover, as would be expected under a schematic account of
identity-relevant information-processing (e.g. Martin & Halverson, 1981),
identification was not found to have consistent effects on knowledge of countries
other than one's own.
In view of the fact that age was confounded with both knowledge and identification
with the national group, it was not possible to enter it as a variable in the principal
analyses of this study. Despite this, we suggest that developmentally the import of the
present findings is that subjective identification can be accorded an important role in
guiding the course of nascent identities: Identifying oneself as a group member would
appear to play a part in assembling the body of knowledge that adult group members
may be expected to have. This contention is made in the light of the fact that our
findings are consistent with a schematic account of identity development even under
the most stringent of circumstances (i.e. involving a highly abstract type of identity),
and using both production and recognition measures. Given thess findings, it is
suggested that future research on identity development should assign the process of
subjective identification a major role, especially in view of the fact that, elsewhere,
too, studies have shown how this process plays a fundamental role in children's
intergroup cognitions (e.g., Powlishta's (1995) work on perceptions of outgroup
homogeneity).
Although identification with the group was consistently related to knowledge of
one's one group, it is noteworthy that it appears also to be related to geographical
knowledge of other countries. Thus, under both measures of identification it was
found that children who identified themselves as British were more knowledgeable
about French and Italian geography than those not identifying. Why should national
identification be related to geographical knowledge but not other types of knowledge?
One possibility is that geographical knowledge is intrinsically relational. Knowledge
about symbols and famous people, however, clearly is not. One cannot identify one's
own country on a map without being able to locate it in relation to other countries;
however it is possible to know of the symbols and famous people of one's country
whilst knowing nothing of others. Although at a general level this account appears
9
plausible, it leaves unaddressed the issue of why identification was found to be related
to geographical knowledge of France and Italy in particular. On this matter we are less
confident, but speculate that France is likely to be significant because of its proximity
as a neighbor to Britain; and Italy, because of distinctive shape, may provide a notable
cue to organizing knowledge of European geography.
Although a number of findings emerged which suggested that ingroup
identification may be related to knowledge pertaining to groups other than one's own,
for only one was this effect consistent over both measures of identification:
knowledge of Italian symbols. We are unable to provide a plausible account for this.
An important objection that might be leveled at our conclusion about the
relationship between ingroup identification and knowledge is that it assumes a
particular causal relation: that identifying with the group leads to acquisition of
information relevant to one's group. However, an alternative interpretation is that
knowledge of one's group may drive identity development: As one acquires more
knowledge of the group, so identification with the group is heightened. Evidence
contradicting this latter account is that 12-year-olds identify significantly more
strongly with the national group than do 15-year-olds. Over all three knowledge
measures, however, they fare significantly less well than 15-year-olds. Thus, it
appears unlikely that identification with the group is determined by knowledge of it.
However, there must remain some uncertainty over this issue given that, in order to
test the theory formally, it would be necessary to perform experimental manipulations
of gender and other forms of identity, which is clearly an impossibility.
Previous research concerned with children's identity development has taken a
discrete view of identities. For example, one either recognizes that one is a boy or
girl, or one does not. In this study, identification was conceived as graded, so that one
might identify with one's national group very strongly, somewhat, or not at all. A
secondary aim of the study was thus to attempt to determine whether the extent of
identification may be related to the extent of knowledge. Our data indicate that the
crucial distinction is between identifying and not identifying with the group: It was
not the case that differences in knowledge emerged between children strongly and
moderately identifying with the group, but between these children and those who
failed to identify with the group. It would therefore appear to be defensible for future
research to adopt a straightforward approach which treats children's identities
discretely.
Footnote 1. Consistent with social identity theory, we take the view that the
salience of an identity is not unchanging but contextually variable. There are clearly
circumstances under which national identity can be chronically salient, as in cases of
prolonged international conflict, such as wars. However, during the lives of the
children studied here, Britain has not been involved in a major and sustained
international conflict. Given this, it seems defensible to argue that national identity
will not have been chronically salient.
10
References
Bem, S.L. (1981) Gender schema theory: A cognitive account of sex typing.
Psychological Review, 88, 354-64.
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (In press) Children's subjective
identification with the group and ingroup favoritism. Developmental Psychology
Bigler, R.S., Jones, L.C., & Lobliner, D.B. (1997) Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530-543.
Bigler, R.S., Jones, L.C., & Lobliner, D.B. (1997) Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530-543.
Carter, D.B. & Levy, G.D. (1988) Cognitive aspects of early sex role development:
the influence of gender schemas on prescholers' memories and preferences for sex-
typed toys and activities. Child Development, 59, 782- 92.
Cramer, P., & Skidd, J.E. (1992) Correlates of self-worth in pre-schoolers: the role
of gender stereotyped styles of behavior. Sex Roles, 26, 369-390.
Damon, W. & Hart D (1988) Self-understanding in Childhood & Adolescence.
Cambridge: Cambridge University Press.
Durkin K. (1995) Developmental Social Psychology. Oxford: Blackwell.
Fagot, B.L., Leinbach, M.D. & O'Boyle C. (19992) Gender labeling, gender
stereotyping and parenting behaviors. Developmental Psychology, 28, 225-230.
Fagot, B.L., & Hagan, R. (1991) Observations of parent reactions to sex-
stereotyped behaviors: Age and sex effects. Child Development, 62, 1617-28.
Fong G.T., & Markus, H. (1982 ) Self-schemas and judgments about others. Social
Cognition, 1, 191-205.
Fiske, S.T., & Taylor, S.E. (1991) Social Cognition. New York: McGraw Hill.
Lambert, W., & Klineberg, O. (1967) Children's views of foreign people. New
York: Appleton-Century-Crofts Meredith.
Liben L.S. & Signorella, M.L. (1993) Gender schematic processing in children: the
role of initial interpretations of stimuli. Developmental Psychology, 29, 141-9
Livesley, W.J. & Bromley, D.B. (1973). Person Perception in Childhood and
Adolescence. London: Wiley
Maccoby, E.E. (1988) Gender as a social category. Developmental Psychology, 24,
755-765.
Markus, H. (1977) Self-schemata and processing information about the self.
Journal of Personality & Social Psychology, 35, 63-78.
Markus, H. ,Crane, M., Bernstein, S., & Siladi, M. (1982) Self-schemas and
gender. Journal of Personality and Social Psychology, 32, 38-50.
Martin, C.L. & Halverson, C.F. (1981) A schematice processing model of sex
typing and stereoptyping in children. Child Development, 52, 1119-34.
Masters, J.C., Ford, M.E., Arend, R., Grotevant, H.D., & Clarke, L.V. (1979)
Modeling and labeling as integrated determinants of children's sex-typed imitative
behavior. Child Development, 50, 364-71.
McGuire, W.J., & McGuire, C.V. (1987) Developmental trends and gender
differences in the subjective experience of the self. In T. Honess & K. Yardley (Eds.)
Self and identity: Perspectives across the lifespan. New York: Routledge.
Park, B. & Hahn, S. 91988) Sex role identity and the perception of others. Social
Cognition, 6, 61-87.
Peevers, B.H. (1987) The self as observer of self: A developmental analysis of the
subjective self In T.Honess & K. Yardley (eds) Self & Identity: Perspectives across
the Lifespan. London: Routledge.
11
Powlishta, K.K., (1995) Intergroup processes in childhood: Social categorization
and sex role development. Developmental Psychology, 31, 781-88.
Shapiro, J.P. (1988) Relationships between dimensions of depressive experience
and evaluative beliefs about people in general. Personality & social Psychology
Bulletin, 14, 388-400.
Tajfel, H., & Turner, J.C. (1986) The social identity theory of intergroup behavior.
In S.Worchel & W.G. Austin (Eds.) The Psychology of Intergroup Relations (Vol. 2).
New York: Nelson Hall.
Yee, M.D, & Brown, R (1994) The development of gender differentiation in young
children. British Journal of Social Psychology, 33, 183-96.
12
Table 1
Children's degree of subjective identification as British over age-groups
Self-reported importance of 'Britishness'
Age-group
Not at all/DK
A little bit
Very
6-year-olds
81
7
20
9-year-olds
34
23
52
12-year-olds
7
52
63
15-year-olds
18
51
50
13
Table 2
Extent of children's knowledge over age-groups, countries and types of knowledge
Type of
knowledge
& country
6-yr-olds
9-yr-olds
12-yr-olds
15-yr-olds
F value (df
= 3, 458),
p<.0001 in
all cases
Geograph-
ical
knowledge
Britain
.21
1.32
1.90
2.32
119.8
France
.15
.67
1.48
1.82
128.6
Germany
.00
.22
.85
1.21
82.82
Italy
.02
.52
1.42
1.56
108.56
Spain
.06
.53
1.65
2.00
105.26
Symbol
knowledge
Britain
1.43
3.01
3.82
4.32
139.5
France
.82
1.51
2.51
2.82
96.56
Germany
.70
1.15
1.83
2.50
68.11
Italy
.75
1.48
2.63
2.82
137.00
Spain
.91
1.31
2.12
2.45
58.91
Knowledge
of famous
people
Britain
.28
.92
2.29
3.1
77.07
France
.03
.07
.47
1.01
47.67
Germany
.00
.09
.72
1.11
60.58
Italy
.00
.16
.57
.92
28.38
Spain
.00
.05
.17
.26
10.05
14
Table 3
Children's geographical knowledge over level of national identification,
countries and measures
Country
Measure
Level
of
Identifica
tion
F value (Df
= 2, 453
probability
(fixed-
choice
vs. open-
ended)
1
(not at all
British)
2
(a little bit
British)
3
(very
British)
Britain
fc
.57
1.95
1.82
19.8
.001
oe
1.10
1.95
2.07
10.92
.001
France
fc
.40
1.49
1.25
9.17
.001
oe
.75
1.47
1.49
10.41
.001
Germany
fc
.25
.92
.62
2.5
ns
oe
.43
.81
.82
.56
ns
Italy
fc
.31
1.41
1.01
8.5
.001
oe
.67
1.30
1.10
3.7
.05
Spain
fc
.41
1.66
1.22
4.04
.01
oe
.82
1.49
1.47
1.35
ns
15
Table 4
Children's symbol knowledge over level of national identification,
countries and measures
Country
Measure
Level
of Identific
ation
F value (Df
= 2, 453)
probability
(fixed-
choice vs.
open-
ended)
1
(not at all
British)
2
(a little bit
British)
3
(very
British)
Britain
fc
1.94
3.69
3.80
31.7
.001
oe
2.71
3.82
4.00
10.91
.001
France
fc
1.17
2.36
2.24
5.84
.01
oe
1.64
2.37
2.40
2.77
ns
Germany
fc
1.00
1.96
1.74
.69
ns
oe
1.36
1.84
1.96
.42
ns
Italy
fc
1.11
2.50
2.21
9.31
.001
oe
1.61
2.51
2.29
6.81
.001
Spain
fc
1.21
2.00
1.90
.68
ns
oe
1.49
2.01
2.10
1.47
ns
16
Table 5
Children's knowledge of famous people over level of national identification
countries and measures
Country
Measure
Level
of Identific ation
F value (Df
= 2, 453)
probability
(fixed-
choicevs
open-
ended)
1
(not at all
British)
2
(a little bit
British)
3
(very
British)
Britain
fc
.62
2.36
2.03
3.92
.05
oe
1.23
2.35
2.31
3.47
.05
France
fc
.15
.62
.45
.20
ns
oe
.29
.64
.45
2.42
ns
Germany
fc
.17
.83
.51
3.06
.05
oe.
.34
.73
.68
1.16
ns
Italy
fc
.17
.65
.46
.53
ns
oe
.28
.70
.44
3.29
.05
Spain
fc
.06
.19
.12
.66
ns
oe
.09
.17
.16
.06
ns