The moral dimension of children’s and
adolescents’ conceptualisation of
tolerance to human diversity
R. T. Witenberg*
Australian Catholic University, Victoria, Australia
This study examined the kinds of justifications children and adolescents used to support tolerant
and intolerant judgements about human diversity. For the tolerant responses, three main belief
categories emerged, based on the beliefs that others should be treated fairly (fairness),
empathetically (empathy) and that reason/logic ought to govern judgements (reasonableness).
Fairness emerged as the most used belief to support tolerant judgements and the most commonly
used combination of beliefs was found to be fairness/empathy, linking tolerance to moral
reasoning, rules and values. Specifically noticeable was that 6–7-year-olds appealed to fairness
more often in comparison to the 11–12 and 15–16-year-olds. Older students used a larger
repertoire of beliefs to support tolerance, indicating developing cognitive maturity. There was also
a tendency for females to appeal to fairness/empathy more often than males. The major constraint
to positive tolerance was not prejudice toward the target groups but the adolescents’ beliefs in
freedom of speech as a democratic right, pointing to a conflict in values between tolerance and
other human rights.
Considering current events in the world and the global increase of racial1and
cultural diversity, better understanding about inter-group relations has never been
more important. The overall goal of the present research was to examine judgements
and beliefs about tolerance and intolerance to human diversity in relation to patterns
of thinking about real-life critical events. The influence of age and gender were also
Much of the research about the development of children’s understanding of
tolerance of others who are different from them has been examined through research
about prejudice and not through the moral domain. Blum (1999) argues that the
relationship between race and morality in the US has been inadequately considered
*School of Psychology, Australian Catholic University, Victoria, Australia 3065. Email: rivka.
Journal of Moral Education
Vol. 36, No. 4, December 2007, pp. 433–451
ISSN 0305-7240 (print)/ISSN 1465-3877 (online)/07/040433-19
# 2007 Journal of Moral Education Ltd
by moral educators. In contrast, there is an important and large body of research that
has examined prejudice, using social cognitive and social identity theories (Doyle &
Aboud, 1995; Aboud & Levy, 2000) amongst others and, from a developmental
perspective, has primarily explored the development of prejudice in young children
(Nesdale, 2001). For example, research showed that some white American and
Canadian children between the ages of four and seven tended to show same-group
favouritism and that they also held moderate negative attitudes toward people from
other racial or ethnic groups. Prejudice was thought to decline after seven years of
age (Aboud, 1988).
In Australia, studies have failed to uncover ethnic bias in young children as in the
case of American and Canadian studies (Robinson et al., 2001). However,
Australian research about children aged between 7 and 12 is less conclusive, with
some indications that negative attitudes remain until the age of 12 in some young
people and with other research showing that negative attitudes decline from seven
years onward (Black-Gutman & Hickson, 1996). The current research aimed to
increase knowledge about these issues and, in particular, to better understand how
both positive and negative beliefs toward others who are different in racial
characteristics, ethnicity and nationality are conceptualised and developed beyond
12 years of age.
Definitions of tolerance are difficult and may have contributed to limiting the
study of tolerance in favour of studying prejudice. However, unlike prejudice,
tolerance can be grounded in theories of morality, which provide a positive approach
to examining inter-group relations (Vogt, 1997; Blum, 1999; Witenberg, 2002a). In
the psychological research literature, prejudice is defined as ‘an unfavorable
judgment toward a particular group’, while ‘discrimination involves behaving
differently, usually unfairly, toward the members of a group’ (Robinson et al., 2001,
p.5). In comparison, tolerance is a much more ambiguous concept, open to several
interpretations ranging from full or indiscriminate acceptance to forbearance or
‘putting up with’. Although seemingly an ideal form of tolerance, indiscriminate
acceptance in its most extreme form could lead to acceptance of questionable
practices and human rights violations; for instance, if freedom of speech is extended
to all forms of intolerant views including neo-Nazi propaganda (Oberdiek, 2001).
However, based on its Latin origin, tolerance is most commonly viewed negatively as
endurance or ‘putting up with’ something we dislike or even abhor. Tolerance as
forbearance leads to individuals who act without discrimination out of restraint but
in fact remain intolerant in thought and belief. (For a more detailed account of
definitional issues around tolerance see Vogt, 1997; Blum, 1999; Robinson et al.,
2001 and philosophical writings, including Oberdiek, 2001; Walzer, 1997).
An alternative way to think of tolerance is to place it within the moral domain.
Philosophers argue that today most of us should and do regard tolerance as a positive
civic and moral duty discharged between individuals who are of equal value. It is a
moral obligation bound by mutual respect and consideration between people
(Rawls, 1972; Walzer, 1997; Dusche, 2002) and, perhaps most importantly, it
entails respect for the autonomy of the individual. It is an essential element in social
R. T. Witenberg
cohesion and an antidote to intolerance and prejudice (Vogt, 1997; Walzer, 1997;
Blum, 1999; Dusche, 2002). The idea that tolerance is a moral duty had been
acknowledged by earlier civil libertarians, such as John Locke, Baruch Spinoza, John
Stuart Mill and others, who argued that tolerance presupposes the value of the
individual, his or her autonomy and freedom of choice. While there is contemporary
debate in the philosophical literature about the asymmetry of social and political
power that affects tolerance (Blum, 1999), many recent philosophers have linked
tolerance with respect, equality and liberty. This allows for the coexistence of
conflicting claims of beliefs, values and ideas as long as they fit within a scheme of
moral values (Dusche, 2002). When tolerance is placed within the moral domain
pertaining to equality, justice and respect and avoiding harm to others, it should be
viewed as positive in nature.
Thus for the purpose of this research, tolerance was defined as the conscious
affirmation of favourable judgments and beliefs involving principles of justice,
equality, care and consideration for the plight of others or, more concisely, according
respect and equality to others who are different through racial characteristics,
ethnicity and nationality. This definition involves an active, conscious and reflective
agent and for these reasons was adopted for this research.
While it is important to know whether young people are tolerant or not, how they
reason about taking a tolerant (or intolerant) stance is equally important.
Judgements about important sociomoral or moral issues are not made in a vacuum
but as a function of our belief system (Goldman, 1986). Personal beliefs are the basis
from which decisions and judgements are made (Kuhn, 1991) because they are used
as a personal yardstick of propositional truth or possible truth and represent an
individual’s thought (Macnamara, 1986). Beliefs are acquired and can be changed
(Colby & Damon, 1995) but they do usually imply a personal stance, rather than a
chance conviction (Goldman, 1986). Believing, therefore, is an intellectual
judgement, which involves a specific conviction. By implication, holding favourable
and/or unfavourable attitudes and beliefs toward others who are different from us in
racial characteristics, ethnicity and nationality, influences our decisions or
judgements. Ponterotto and Pedersen (1993) have argued that prejudicial attitudes
are related to ‘an overgeneralized or erroneous belief’ (p.10) that may not be
expressed openly. The corollary is also a realistic proposition in that we hold positive
beliefs about others and about the reasons for being tolerant that we rarely express
and, from an empirical perspective, we know less about.
In developmental psychology, particularly in the pro-social and moral domains,
there is a tradition to explore the way children, adolescents and young adults reason
about and justify their views in order to understand the underlying rationale behind
their judgements. Typically, such a methodology maps children’s, adolescents’ and
young adults’ reasoning and judgements, using stories or dilemmas that present a
character with conflicting events requiring resolution (for example, Piaget, 1932/
1965; Kohlberg, 1981; 1984; Dunn et al., 2000). This is a particularly important
aspect of children’s tolerant judgements and whether they can reason about and/or
justify their stance. While a child may score highly on a particular tolerance scale, his
Tolerance to human diversity
or her reasoning for such a stance is not explicit and may have little to do with
tolerance. The same argument could be advanced for adolescents and even young
adults. Thus, examining the reasoning process could give a better insight into how
tolerance to human diversity emerges and is supported. However, there appears to
be limited research that has examined the way tolerant judgements are reasoned
about or the relationship between tolerant judgements and beliefs.
Studies in areas akin to tolerance to human diversity have at times also taken this
approach. For example, in a series of studies using a cognitive developmental
approach, Enright and Lapsley (1981) asked children, adolescents and young adults
to judge the worth of others who held beliefs discrepant from their own on a range
of social, moral and political dilemmas. They concluded that belief-discrepancy
tolerance involved a four-level, age-related progression from less to more tolerance,
consistent with Piagetian development.
Just over a decade later, Sigelman and Toebben (1992) examined the
development of both political and belief-discrepancy tolerance of Grade 2, 5 and
8 students and showed that situational information contributed to variations in
responses. For example, responses varied considerably for the younger children
dependent on the popularity of the idea held by those with discrepant beliefs.
Overall, they found differences in tolerance levels increasing with age but, when
context was considered, no global construct for tolerance emerged. A later study by
Wainryb et al. (1998) examining belief-discrepancy tolerance only, also found age-
related differences that were mediated by situational information; this suggests
domain-specific approaches may explain the data better than global stage theories.
Thus, one way of measuring tolerance to human diversity is through the use of
dilemmas. In 2002 an Australian study (Witenberg, 2002b) examined tolerance to
diversity using dilemma-like stories, with a similar methodology to the current
research. Respondents to the study, aged from 11–22 years of age, expressed a very
high level of tolerance towards others different from them in racial characteristics,
ethnicity or nationality. Similar findings were evident in an Israeli study using the
same methodology but culture-specific stories (Witenberg & Cinamon, 2006).
Further cluster analysis of the Witenberg data, with the aim of identifying distinct
groups of respondents, found that up to one-half of the sample, including up to one-
third of young adults (18–22 years), consistently made tolerant judgements
regardless of the situational information embedded in the different dilemma-like
stories (Witenberg, 2002a).
While tolerance and intolerance do co-exist (Doyle & Aboud, 1995; Wainryb
et al., 1998), these findings support the contention that some individuals are
consistently tolerant, irrespective as to who and under what circumstances they
are asked to extend their tolerance, at least on an explicit level. Yet we know less
about what underpins positive judgements and beliefs that guide the reasoning
process about tolerance than we know about intolerance or prejudice. A stronger
focus on tolerance in its own right would increase our understanding of tolerance
and its moral dimensions, as well as the relationship between tolerance and
R. T. Witenberg
How to measure beliefs about tolerance must be considered, bearing in mind the
current debate about implicit measures about prejudice. Fazio and Olson (2003)
have highlighted the importance of recognising that both implicit and explicit or
‘direct’ measures (as they prefer to call more deliberative measures) ‘can be
predictive of judgments and behavior’ (p.305) and that it is the measures, and not
necessarily the processes underlying them, that are implicit or explicit in nature.
Wittenbrink et al., (2001) have suggested that explicit measures are conceptual tasks,
which tap into cognitive beliefs, in contrast to automatic activation of stereotyping
and prejudice, as the research into that topic suggests. Knowing more about the
kinds of explicit cognitive beliefs individuals use to underpin tolerant judgements,
their relationship to moral beliefs and their age progression (together with those used
to support intolerant judgements) could have implications for curriculum design and
social policy alongside efforts to reduce prejudice.
In adopting a cognitive developmental approach, this study was concerned with
the way individuals reason about critical issues that can be resolved with either
tolerance or intolerance. Thus the main aim of this research was to explore the kind
of beliefs young people between the ages of 6 and 16 used to guide their reasoning
and whether these beliefs remained the same across age and in different situational
In the light of the lack of a substantial body of research about the conceptualisa-
tion of tolerance to human diversity, this study should be seen as exploratory.
However, it was expected that children as young as six years would favour a tolerant
stance. A pilot study showed that young children have a tendency to make tolerant
judgements and support their judgements with appeal to fairness. It is also
anticipated that younger and older adolescents would use different kinds of beliefs in
reasoning about tolerance and intolerance, but that how they reason about a tolerant
stance would remain, to a degree, unclear.
There were 112 participants in this study, 45 males and 67 females. Of the
participants, 40 were aged between 6 and 7 years (M56.5, SD50.35), 32 aged
between 12 and 13 years (M512. 5, SD50.34) and 40 aged between 15 and 16
years (M515. 6, SD50.49). They were recruited as intact Grade 1, 6 and 10 class
groups from two private schools and two state-run schools (two elementary and two
secondary) in the northern and south-eastern regions of Melbourne. While the
sample ranged from lower to upper middle class, none of the teaching institutions
were situated in areas of extreme socioeconomic deprivation. Based on the
demographic information gathered, they represented a cross-section of the diversity
of Australian society. Of the participants, 64. 5% had parents born in Australia,
11. 8% in Asia and South-East Asia, 20.1% in Europe or the UK and 2.5% in the
Middle East or the Indian subcontinent. Less than 1% of the parents were born in
Africa or South America. All of the participants were fluent in English.
Tolerance to human diversity
Measures and procedure
The stimuli to assess tolerance consisted of three short dilemma-like stories. To
assess the influence of situational information (context effect), each story dealt with
a different event or situation depicting a form of intolerance/tolerance relevant to the
Australian context. The events were based on real-life incidents gathered from
reports in newspapers, individual experiences and from official sources, such as the
Australian Human Rights and Equal Opportunity Commission’s (2004) study into
racial prejudice in the Australian environment.
One story concerned a storekeeper who wants to serve a person with an Asian
background last because the storekeeper believes that Asians do not belong in
Australia. Another story involved an employer who will not hire a young English
person because the employer believes that English people are lazy. Yet another story
was about a teacher who believes in giving lower marks to Aboriginal students
because they are black. A pilot study found that the use of the terms ‘Asian’,
‘English’ and ‘Aboriginal’ posed no conceptual or ethical problems to the
participants. To avoid social desirability the stories were presented with competing
considerations. That is, a negative belief that involved some form of racial
intolerance was followed by a positive belief presenting the opposite view (see
below for a more detailed account of the stories). Each story could be resolved using
either tolerant or intolerant views. For example:
I know a storekeeper who believes that you should serve people from Asian backgrounds
last because they do not belong in Australia. Another storekeeper believes it is wrong to
serve people from Asian backgrounds last because they do not belong in Australia.
The order of the presentation of the stories was systematically counterbalanced for
each subgroup of age by gender.
As the main purpose of the study was to determine how tolerance was
conceptualised through the examination of justifications used, a series of questions
was designed to probe participants’ judgements, explanations and justifications. The
questions are presented below. In each instance, when responding to the first
question, participants were asked to make a judgement about the event presented in
the story. The subsequent three questions aimed to elicit participants’ reasoning
underlying their judgements/decisions. Questions 2 and 3 asked for explanations for
the specific decision/judgement and for participants to express their own ideas about
the event. The final question aimed to obtain justifications for the specific stance
that had been taken. The format of the questions was based on previous empirical
and theoretical analyses found useful in eliciting responses reflecting the reasoning
process (Kuhn, 1991) and Decision Theory (Abelson & Levi, 1983) as well as
examination of underlying beliefs (Witenberg, 2000). Hence the format of the
questions remained the same across all three stories. Participants were asked to
respond to the following four questions in turn after reading each story, with each
presentation of questions modified to the event under examination:
1. Do you think it is all right or not all right to give students from Aboriginal
backgrounds lower marks because they are black? (OR to serve people from
R. T. Witenberg
Asian backgrounds last because they do not belong in Australia? OR to not
employ English people because they are lazy?)
Can you explain why?
What do you actually think about such a belief?
If I told you I would have done the very opposite, how would you convince me
that you made the right choice and my choice is wrong?
The two older age groups made written responses. The youngest age group was
interviewed, as described below. At the beginning of the session, which was
conducted in the participants’ respective classrooms, they were assured that
individual questionnaires could not be identified and they were encouraged to
respond with honesty. At the end of the session, participants were debriefed and
informed that the researchers did not endorse the negative views presented in the
The youngest participants were presented with the same story format and
content but the incidents were adjusted to be conceptually accessible to children
aged 6–7 years, with the help of the participating school principals and Grade 1
co-ordinators. The essential difference in the stories was to place the protagonist
in a situation with which the children were familiar, otherwise the stories remained
exactly the same. For example, as 6–7-year-olds do not receive grades or marks,
lower marks were replaced by fewer stickers, which suited the experience of
children of that age group. The age-relevant modified stories were found to be
useful in drawing out information about children’s tolerance judgements and their
reasoning about their decisions, without eliciting any adverse effects. Skilled
interviewers used to working with children questioned the youngest participants
Due to the young age of the participants, the investigator first asked a series of
preliminary questions to set the stories in a context that 6–7 year old children could
understand, such as ‘Do teachers at school give you stickers for doing good work?’
The investigator then read out each story. After the story was read, and any
misunderstanding and ambiguities were explained or clarified, each participant was
asked to verbally respond to the following set of three questions without any
prompting from the interviewer.
1. Do you think that it is all right that Aboriginal children should get fewer stickers
because they are black? Or do you think that it is not all right?
2. Can you tell me why?
3. What if I said I believe that Aboriginal children should get fewer stickers than
everyone else because they are black? What would you say to me to get me to
think that I am wrong and you are right?
As with the older participants the questions varied depending on the context and
whether the child made a tolerant or intolerant response. The children’s responses
were transcribed and the transcripts were used for content analysis alongside the
completed booklets of the older participants.
Tolerance to human diversity
The transcripts were read several times. The first reading was to gain an overall view
of the transcript. The second reading aimed to analyse the response patterns either
supporting or rejecting tolerance (affirm or disaffirm tolerance) when responding to
Question 1 for each story. Responses were classified into two categories: ‘tolerant’
responses (affirming tolerance) and ‘intolerant’ responses (disaffirming tolerance).
For example, affirmation of tolerance included responses such as ‘It’s not OK to
believe such a thing’, ‘It’s wrong to stop them’ or It’s not all right at all’;
disaffirmation included such comments as ‘It’s OK, I have no problem with it’ or
‘It’s fine with me. He can believe what he wants.’
The third reading aimed to determine the kind of explanations and justifications
participants used to support their judgement for each of the stories. Therefore the
analysis focused primarily on the responses to Questions 2, 3 and 4. Reasoning
responses were analysed using the coding schemes presented in Table1. The coding
schemes were developed based on pilot studies and previous research (Witenberg,
2002a). The categories were designed to reflect underlying beliefs that guided
participants’ reasoning. Participants’ responses were assigned to one or more of the
categories that were not mutually exclusive of each other.
The transcripts were further analysed for content through a process of colour
coding to establish the most frequently used category or categories for each story.
Colour coding is a type of content analysis that is a manual but systematic research
tool to determine themes, ideas underlying beliefs and concepts emerging from the
text. It has been used to interpret moral narratives, conflicts and choices (Brown
et al., 1989). For the ‘tolerant’ responses, affirming tolerance, the following
categories emerged: judgement made on the basis of fairness; judgement made on
the basis of fairness and empathy where appeal to both fairness and empathy was
used with similar frequency; judgement made on the basis of unreflective thinking
(reasonableness); and idiosyncratic responses. The dual response of fairness and
empathy was classified as one category labelled fairness/empathy.
A similar process was undertaken for the ‘intolerant’ responses, disaffirming
tolerance. The following belief categories emerged in this part of the analysis:
judgement made on the basis of freedom of opinion and/or freedom of speech
(freedoms); judgement made on the basis of prejudicial beliefs; and idiosyncratic
The final reading aimed to assess variations in the response pattern across the three
stories. Based on colour coding, the analysis revealed that the contextual information or
the content of the stories did not influence overall belief categories used to support
either a tolerant or an intolerant stance across the three stories. That is, participants
expressed a predominant belief or a combination of beliefs across the three stories,
irrespective of the context. For example, one participant used the following
justifications for the three stories ‘It is not fair that they [students from Aboriginal
backgrounds] should get lower marks; it should be decided on the way they do their
work’; ‘Asian people may look different but we are all the same. We should treat them
fairly’; and ‘It’s not fair to stop someone from getting a job just because they are
R. T. Witenberg
English’. As the examples illustrate, a respondent may justify his or her judgements
with story-specific references, but the underlying belief that guided the reasoning
remained the same across the three stories and in this case involved a belief about the
principle of fairness. In light of this finding, participants were assigned to either one of
the belief categories that predominated their reasoning across the three stories.
However, in some instances both fairness and empathy were used in equal measures
and, thus, a new combined category was formed for the purpose of the analysis. Thus
the categories were fairness, fairness/empathy and reasonableness.
An independent judge coded a randomly selected one-quarter of the responses,
using the belief categories described in Table1. The judge was not instructed and no
discussion took place about the individual categories. Inter-rater reliability was
found to range between 0.75 and 0.82 using Kappa for the four major belief
Table 1. Belief categories and examples of justifications
Belief category Description and examples of justifications
Fairness Expressed through appeal to equality, similarities, differences and rights.
(‘It’s not fair’; ‘We are all equal’; ‘We are all different’ and ‘We are all
alike’; ‘I would explain that we are all equal and it shouldn’t matter where
you come from or what you are’; ‘We should all be treated fairly and
equally’; ‘It’s wrong because black people are just the same as us no
different’; ‘People are different from each other. It doesn’t matter if you are
black, white or Asian we should all be treated fairly’.)
Empathy Expressed through appeal to personal feelings, perspective taking or hurt.
(‘Your choice is wrong because if all the Aboriginal people were living on
one street and they didn’t let you live there, how would you feel?’; ‘I would
ask you what you would do if you stood in the Asian’s shoes. How would
you feel?’; ‘They shouldn’t be allowed to say these sorts of things because it
could make Asian young people depressed and hurt’; ‘It can be hurtful and
give them [people from an Aboriginal background] a complex’.)
Reasonableness Expressed through suggestions of unreasonable or ‘unreflective’ ideas and
through challenging generalisations and assumptions (‘I personally think that
it’s stupid that a person can think like that’; ‘Because racism is bad and it’s
stupid and illogical’; ‘This person is misinformed. I think that this is a really
irrational attitude’; ‘This person is making a rash generalisation. I would point
out that some Asian people have lived here longer than some white people’.)
Freedoms Expressed through appeal to freedom of speech and freedom of opinions.
(‘Everybody has the right to have their own opinion and talk about them’;
‘If this person wants to think such things he can do so because this is a free
society and we have no right to persuade him otherwise—he has the right to
think what he wants’; ‘I believe she/he has the right to state her/his opinion
and can share her/his opinion with others. We live in a free society and we
have the right to say what we believe’.)
Prejudice Expressed through appeal to traditional or personal biases.
Others Miscellaneous and unelaborated responses that cannot be coded.
Tolerance to human diversity
categories of fairness, fairness/empathy, unreflective thinking and freedoms. Hence,
a good to high rate of agreement was reached between the two raters in each case
(Landis & Koch, 1977).
The major aims of this study were to assess the relationship within and between
belief categories and the influence of age and gender. Due to the nature of the data,
nonparametric and categorical data analyses were used. Table2 presents the
response pattern for the belief categories of fairness, fairness/empathy, reason-
ableness, freedoms and prejudice for the whole group and within age groups.
The association between the belief categories
To examine if there were any significant differences in the distribution of the belief
categories, a one-sample chi-square goodness-of-fit test was conducted. The data
showed that the observed and expected distributions differed significantly from one
another, x2(4, n5112)5127. 91; p,0.001. The residuals (see Table3) indicate
Table 2. Percentages of whole group and within-age group responses to belief categories across
Group 6–7 11–13
Table 3. Observed and expected count and residuals of belief categories; one-sample chi-square
x2(4, n5112)5127.91; p,0.001.
R. T. Witenberg
significant differences in the use of belief categories with fairness used more often
than expected by chance, followed by the use of fairness/empathy with reason-
ableness used less often and freedoms and prejudice used substantially less often
The association between belief categories, age and gender
To assess possible multi-way associations between age and gender on belief
categories, a Log Linear Analysis was conducted on belief categories that comprised
above 10% of the total response pattern for each belief category (see Table2). On
this basis, only fairness, fairness/empathy and reasonableness were included in all
subsequent analyses. The analysis revealed that the most parsimonious model was
one where age and gender were independent of each other. Thus two separate Chi-
square analyses were conducted to assess two-way associations of belief categories
with age and gender respectively.
The association between belief categories and age
Figure1 presents the within-age group response patterns for each belief category
Figure1 shows that all three belief categories also show clear variations in the
response pattern for age. Specifically noticeable is that within age groups, 84.2% of
all the 6–7-year-olds appealed to fairness, with none of them appealing to
reasonableness. In contrast, only 29% of the 11–13-year-olds and 13. 9% of the
oldest participants appealed to this mode of thinking. To examine the association
between belief categories and age, a Chi-square test of association was conducted
which revealed that the observed distribution differed significantly from an equal
likelihood distribution, x2(4, N5105)515. 76; p,0.01, as shown in Table4. Thus
Figure 1. Percentages of belief categories within age groups
Tolerance to human diversity
the data indicates that the belief categories were not independent of age. To assess
where the relevant differences lie, the adjusted residuals (see Table4) were
examined. Adjusted residuals of 2 or exceeding 2 are a confirmation of a statistically
significant deviation from what would be expected by chance alone. Similarly to z-
scores, adjusted residuals of 2 or larger are significantly different from zero at 95%
The adjusted residuals indicate an association between the age groups and beliefs.
The data shows that 6–7-year-olds appealed to fairness and that 11–13-year-olds
appealed to reasonableness more often than was expected by chance.
The association between belief categories and gender
The relationship between beliefs and gender was also considered. Figure2 presents
the within-gender response pattern for each belief category.
Table 4. Observed and expected count of belief categories for each age group*
Total 682314 105
x2(4, N5109)517.64; p,0.001.
*Adjusted residuals in parentheses and italic.
Figure 2. Percentages of belief categories within gender
R. T. Witenberg
Variations in response pattern are also evident with gender. Appeal to fairness/
empathy seemed to be favoured by females. In contrast, more males appealed to
fairness only and reasonableness. A second Chi-square test of association was
conducted which revealed that the observed distribution differed significantly from
an equal likelihood distribution, x2(2, N5105)514. 10; p,0.001, as shown in
Table5, indicating that the beliefs used were not independent of gender.
In particular, the adjusted residuals (see Table5) indicate an association between
females and fairness/empathy, showing that females appealed to fairness/empathy
more often than was expected by chance. A trend is also indicated that males
appealed to beliefs about reasonableness more often than was expected by chance
This research explored how tolerance of human diversity was conceptualised by
examining the underlying beliefs that guided the reasoning process to support
tolerant and intolerant judgements about racial characteristics, ethnicity and
nationality of children and adolescents between the ages of 6 and 16 years. While
high levels of tolerance were expressed, responses were influenced by both within-
and between-subject variations. For the tolerant responses, three main belief
categories emerged based on the beliefs that others should be treated fairly (fairness),
empathetically (empathy) and that reason/logic ought to govern judgements
(reasonableness). Variations in response patterns between these belief categories
emerged. Use of underlying beliefs that others should be treated fairly and
empathetically indicates overlapping relationships with moral reasoning. There were
also between-group age-related variations in response patterns, pointing to patterns
of development. Gender differences also emerged. Two main belief categories
emerged from the intolerant responses based on beliefs about freedoms (speech and
opinion) and prejudicial beliefs.
By their inherent nature, conflicts can only be resolved through a decision or a
judgement (Hart & Killen, 1995) where the individual takes a stance in favour of one
Table 5. Observed and expected count of belief categories by gender*
Total 68 2314 105
x2(4, N5109)516. 34; p,0.000.
*Adjusted residuals in parentheses and italic.
Tolerance to human diversity
or the other of the two propositions. In this study the majority of the participants
endorsed tolerance and used a set of beliefs across the three stories to support their
stance, irrespective of the contextual information. While previous studies showed
that contextual differences influence judgements (Wainryb et al., 1998; Witenberg,
2002b), the findings of the current study did not support context effect when
underlying beliefs were examined. At least on a general level, the similarity in the use
of the three key positive beliefs across the dilemma-like stories could suggest that
participants were tapping the same fundamental social and socio-moral beliefs
relevant to tolerance of human diversity. A similar argument can be made for beliefs
that were used to reject a tolerant stance. This seems particularly true for freedom of
opinions and speech.
The most commonly used justifications were based on appeal to the principle of
fairness, which is closely associated with equality and justice, suggesting a moral
dimension to tolerance. This supports the argument that tolerance is viewed as a
moral obligation between people and is more than endurance (Dusche, 2002).
Justice, equality and fairness are also critical for moral development and reasoning
(Rawls, 1972; Kohlberg, 1984). Barrow (2001) argues that ‘…the principle of
fairness or impartiality is absolutely fundamental to any coherent moral philosophy’
(p.237). Moral beliefs have also been shown to influence political tolerance (Avery,
1988). Unpacking this relationship may give further insight into the development of
tolerance generally and its relationship to morality.
In previous work, Witenberg (2002a) found that fairness, empathy and reason-
ableness predicted 14% of the variance on judgements about tolerance, with fairness
accounting for 11% of the variance. In the current study, appealing to the principle
of fairness was found to be used across the three age groups, but it also clear that 6–
7-year-olds used appeal to fairness more often than the other two groups, indicating
a possible developmental trend specific to this area.
While differences in the level of elaborations emerged in how appeal to fairness
was expressed among the three age groups, the underlying message that we should
treat others who are different from us fairly was unmistakable. Appeal to the
principle of fairness was also found in the reasoning of young adults (aged between
18 and 22 years) about the same or similar stories (Witenberg, 2002a; Thomas &
Based on theories of social cognition, Fiske (1991) argues that fairness becomes
salient at about the age of four and is generalised to social issues due to the
maturation process leading to the externalisation of innate cognitive models.
Cognitive models, whether innate or acquired, imply a theory about a fair world.
Kuhn (2002) argues that, rather than using theories as metacognitive tools, ‘young
children think with their theories’. It is not within the scope of this work to be able to
determine the processes that lead adolescents to adopt alternative beliefs to support
tolerance. However, speculatively, it could be suggested that age brings both
increased knowledge and the emergence of metacognitive structures to aid the
reasoning process. Younger reasoners encode a situation more narrowly and tend to
focus on the most salient issue, often disregarding the complexity of the problem, in
R. T. Witenberg
contrast to older adolescents and young adults whose processing capacities allow
them to consider several aspects of a problem simultaneously (Bjorklund, 2000).
Precisely how and why this happens is a challenge for future research about
Appealing to empathy and fairness with equal measure was another way to
support a tolerant stance. In addition to fairness, the underlying message about
empathy was related to considering how it would feel to be treated badly. Empathy is
a motivator of pro-social and altruistic behaviour (Hoffman, 2000) and is a basis by
which prejudice can be reduced (Batson et al., 2002). Hoffman further argues that
empathy and perspective-taking are both implicated in moral development and
particularly relevant to the moral care orientation proposed by Gilligan (1983). In
fact, in the current research the use of both fairness and empathy in equal measures
to support tolerance was most prevalent in the responses of the female participants.
One of the enduring debates in the last two decades has been whether morality is
gender specific (Gilligan, 1983; Wark & Krebs, 2000; Skoe et al., 2002). The current
data showed that, whilst both females and males used beliefs about fairness and
empathy equally, there was a significant trend for females to support tolerance in this
manner. However, it is too early to predict gender differences in tolerance of
diversity, especially since there is a substantial body of research that disputes the
claims of gender differences in moral reasoning (for example, Walker, 1984) which
cannot be ignored. Future research will help to clarify this relationship.
The least prevalent belief was that reason/logic ought to govern judgements. When
reasonableness was used, participants appealed to rationality and logic and indicated
the stupidity of holding prejudicial beliefs. They also questioned assumptions and
generalisations or asked for evidence. It is harder to link reasonableness to clear
moral beliefs, such as treating others fairly and empathetically but, implicitly, it
suggests a sense of indignation or questioning of stereotypical beliefs and
discriminatory acts. The youngest age group did not use reasonableness, which is
not surprising. Questioning assumptions and generalisations and seeking evidence is
only available developmentally to cognitively more mature individuals (Kuhn,
1991). However, there was a tendency for males to appeal to reasonableness more
often than females. On the basis of the current study it is difficult to determine why
appealing to reasonableness is favoured by males from 11 years of age onward. It is
possible that patterns of socialisation in which girls are more encouraged to express
emotions contributed to the adolescent girls’ empathic responses, whilst the
‘masculinity of reason’ has been acknowledged both in the feminist and
psychological research (e.g. Jaffe & Hyde, 2000; Applebaum, 2001). This needs
further investigation if we are to understand the development of tolerance.
The major constraint to tolerance was not prejudice toward the target groups but
older participants’ beliefs in freedoms (of speech and opinion) as democratic rights.
This was unanticipated and differed from the intent of endorsing traditional
prejudicial beliefs. In previous work, Witenberg (2004) found that the subordination
of tolerance to freedom of speech was least evident in 11–12-year-olds and most
pronounced in the 18–22-year-olds. Interestingly, Helwig (1998) also found that
Tolerance to human diversity