What is the role of teachers in helping children and teen’s development? What is the influence of peer groups on children and teens? What is the influ

Belonging to and exclusion from the

peer group in schools: influences on

adolescents’ moral choices

Luba Falk Feigenberg*a, Melissa Steel Kinga, Dennis J. Barra,b

and Robert L. Selmana

a Harvard University, USA; bFacing History and Ourselves, Boston, USA

This paper reports on a mixed methods study of adolescents’ responses to case material about

social exclusion. First, a qualitative coding method is presented that describes the way adolescents

choose and justify strategies to negotiate such situations. The responses were then analysed

quantitatively using chi square tests and multinomial logistic regression. Findings indicate that

adolescents’ interpretation of their social context was a significant factor in their choice of strategy.

Those adolescents who invoked normative rules and conventions as the most salient justifications

were more likely to recommend bystanding rather than joining in the exclusion. However,

adolescents who viewed the protagonist’s own choice as an opportunity for making long-lasting

positive changes in the social environment were more likely to recommend helping the victim.

Gender and school context also were associated with adolescents’ choice of strategy. Implications

for research in moral development as well as practical implications for school-based programming

are discussed.

Introduction: the problem of peer group social exclusion and bullying in the

schools

In her ninth grade class on ethics and history, Eve Shalen wrote an essay about her

involvement with an incident of social exclusion and ostracism that occurred in early

adolescence. About her experience with social relations in middle school, she says:

My eighth grade consisted of 28 students most of whom knew each other from the age

of five or six. Although we grew up together, we still had class outcasts. From second

grade on, a small élite group spent a large portion of their time harassing two or three of

the others. I was one of those two or three, though I don’t know why…The harassment

was subtle. It came in the form of muffled giggles when I talked and rolled eyes when I

turned around. If I was out in the playground and approached a group of people, they

*Corresponding author. Harvard Graduate School of Education, Larsen Hall 610, Appian Way,

Cambridge, MA 02138, USA. Email: [email protected]

Journal of Moral Education

Vol. 37, No. 2, June 2008, pp. 165–184

ISSN 0305-7240 (print)/ISSN 1465-3877 (online)/08/020165-20

# 2008 Journal of Moral Education Ltd

DOI: 10.1080/03057240802009306

often fell silent. Sometimes someone would not see me coming and I would catch the

tail end of a joke at my expense.

There was another girl in our class who was perhaps even more rejected than I. One

day during lunch…one of the popular girls in the class came up to me to show me

something she said I wouldn’t want to miss. We walked to a corner of the playground

where a group of three or four sat. One of them read aloud from a small book, which I

was told was the girl’s diary. I sat down and, laughing till my sides hurt, heard my voice

finally blend with the others. Looking back, I wonder how I could have participated in

mocking this girl when I knew perfectly well what it felt like to be mocked myself. I

would like to say that if I were in that situation today I would react differently, but I

can’t honestly be sure. (Facing History and Ourselves, 1994, pp. 29–30).

Experiences of social exclusion such as this are quite common to early adolescents

who struggle to navigate social relationships every day. Much of the discourse on

social exclusion focuses on why adolescents who may ‘know better’ still join in when

they see their peers mocking or harassing another student (Bosworth et al., 1999;

Nansel et al., 2001). However, there is little research that examines what might help

adolescents ‘think beyond’ not joining in and, instead, influence them to get

involved to help the victim. In other words, what factors might influence adolescents

to choose to stand up for themselves or others?

In this paper, we explore the way early adolescents think about the choice to

bystand, join the perpetrators or defend the victim in a situation of social exclusion,

such as the one Eve Shalen describes in her school. First, we review the research

literature on how various factors—social cognitive, cultural and contextual—influence

adolescents’ choices about difficult social actions and decisions. We then describe the

construction and validation of a framework to classify both the range of strategies

adolescents recommend in response to the In Group Assessment, a qualitative

measure based on the Eve Shalen case material, as well as the justifications they give

for their choice (Barr, 2005). Next, we present findings about the influence of school

contextual factors on the variation in social choices. We conclude with a discussion of

the implications of our findings for research and practice in moral development and

educational programming to reduce social exclusion.

Social exclusion: definitions of the problem

Often defined as a physical, verbal or psychological action intended to cause fear,

distress or harm to the victim, social exclusion during childhood and adolescence

occurs within the context of an asymmetric power relationship, where a more

powerful child, or group of children, oppresses the less powerful one(s) (Olweus,

1993). Social exclusion in adolescence often includes ostracism, teasing, harassment

and bullying (Swain, 1998). Research in the US suggests between 5 and 27% of

adolescents admit to having excluded a peer (Nansel et al., 2001; Dake et al., 2003).

It is now recognised that social exclusion is seldom the action of only one

individual. The group dynamics required for and created by such behaviour

contributes to the overall culture and climate of schools and social groups. Not

surprisingly, schools with higher rates of exclusion are perceived as less safe (Astor

et al., 2002; Dupper & Meyer-Adams, 2002). Faced with such statistics, schools

166 L. F. Feigenberg et al.

struggle to find effective responses to social exclusion and to prevent its negative

consequences in their hallways, cafeterias, gyms and classes.

Further, rigid rules and inflexible consequences in schools typically address the

behaviour of those identified as the ‘perpetrators’ but not that of the ‘bystanders’—

those students who either passively watch or actively incite the exclusion (Staub, 2002).

In fact, even though most adolescents believe social exclusion is wrong, they often do

not try to intervene and usually stand by passively (Tisak et al., 1997; O’Connell et al.,

1999). In order to address this problem, it is important to understand what may cause

adolescents to make choices about their own behaviour in these contexts.

Theoretical orientations and empirical evidence: a brief review

Psychological theories that focus on child and adolescent social cognitive

development primarily describe the quality of individuals’ thought processes that

relate to or influence their social behaviour. Social information processing models,

for example, portray the cognitive steps necessary for individuals to make decisions

about social action (Dodge, 1986; Dodge & Price, 1994; Crick & Dodge, 1996).

Adolescents engage in a series of thought processes and ultimately choose an action

from a range of perceived possibilities. Antisocial behaviour, or social exclusion, may

result from misperceptions of the actions and intentions of others, a deficiency that

may occur at any point in the social information process (Fontaine et al., 2002).

According to these models, adolescents’ choices are due to internal cognitive

structures and abilities where the selection of exclusionary behaviour may be an

indicator of inaccurate or distorted social perceptions (Camodeca et al., 2003).

Other cognitive developmental approaches examine adolescents’ thought processes

as related to social interactions or the understanding, negotiation and meaning of

social relationships over time (Selman, 2003). This approach focuses specifically on

the conditions under which children develop and use the ability to coordinate different

social perspectives (Collins, 2002), rather than conceptualising social competence as a

sequence of social cognitive information processing steps (Selman, 1980; Keller &

Edelstein, 1991). For example, adolescents who have more difficulty coordinating

their own and others’ points of view are at greater risk for peer conflict and may be

more likely to participate in social exclusion (Selman et al., 1992, 1997). Like social

informational processing models, however, many earlier social cognitive-develop-

mental theories (Kohlberg, 1971) located the impetus for adolescents’ choice of social

action primarily in the minds of individuals or at the individual level of analysis. By

focusing mainly on individuals’ social cognitions, these models often do not consider

how other factors may influence adolescents’ social choices, especially under

challenging, complex or ambiguous conditions (Steinberg, 2003).

Differences between adolescents’ social viewpoints and actions can also be attributed

to population level or cultural factors. Research on individuals’ membership of broad

social systems suggests that group affiliation, such as gender, race or socioeconomic

status, may shape the way they think about their choices about behaviour. For example,

gender has been shown to be an important influence on social exclusion. When faced

Moral choices 167

with such situations, early adolescent boys tend to prefer to join in with the

perpetrators, while girls tend to side with the victim (Nansel et al., 2001; Seals &

Young, 2003). In contrast, research that focuses specifically on social aggression—

defined as the manipulation of a relationship in order to damage ‘another’s self-esteem,

social status or both’ (Underwood, 2003, p. 23)—suggests girls are much more likely

than boys to be the perpetrators of socially aggressive acts (Crick et al., 1996; Galen &

Underwood, 1997). Further, girls are expected, by both boys and girls, to be the

perpetrators of socially aggressive acts (Crick & Grotpeter, 1995). Lastly, there appear

to be gender differences in preferred coping strategies and desire for adult intervention

in situations of social exclusion (Naylor et al., 2001; Gamliel et al., 2003).

While gender plays an important role in adolescents’ behaviour in social exclusion,

such behaviour does not appear to vary by race or ethnicity, at least in the US (Leff et

al., 1999; Seals & Young, 2003). Yet, when cultural characteristics are considered as

part of the larger context within which exclusion occurs, adolescents view race, as well

as gender, as important factors in thinking about when such behaviour may be

appropriate. For example, adolescents consider others’ racial background as a valid

reason for excluding individuals from particular social relationships or situations, such

as school-based clubs or peer groups (Killen & Stangor, 2001; Killen et al., 2002).

The roles individuals assume within the peer group context also affect the way

they think about social choices. Individuals tend to take on particular roles, across

both situations and time, that contribute to the structural components of intergroup

dynamics. In particular, adolescents tend to view their moral choices differently

depending on whether they are witness to an event or a perpetrator (Salmivalli et al.,

1996; Sutton & Smith, 1999). In other words, adolescents’ role in the social context

influences their choices for behaviour. In this sense, context extends beyond the

impact of external social structures and systems on social and relational behaviour.

Rather, context can be thought of as the system of ‘social activities and cultural

meanings’ in which an individual participates (Serpell, 2002). The way students

understand their choices, then, is grounded in their context and social experiences

(Burton et al., 1996).

The school context, for one, plays an especially important role in the way students

perceive risk behaviour (Kuperminc et al., 1997). For example, the level of safety

adolescents perceive in their schools and communities is related to their thinking

about the meaning of threatening behaviour (Espelage et al., 2000). When students

believe their social environment is unsafe they are more likely to interpret others’

behaviour as hostile or aggressive. The climate fostered in the school plays an

important role in students’ choices around peer group actions (Salmivalli & Voeten,

2004), such as whether they join the perpetrator(s) or defend the victim.

Research questions

This study is guided by the following research questions:

1. What strategies do students recommend for negotiating a situation of social

exclusion and what justifications do they offer for their chosen strategy?

168 L. F. Feigenberg et al.

2. Do students’ recommended strategies and justifications for negotiating a

situation of social exclusion differ by their role in the social context, i.e.

whether they are asked to assume the role of witness or perpetrator?

3. Do students’ recommended strategies for negotiating a situation of social

exclusion vary as a function of their perceptions of the social context, i.e.

school or classroom climate?

To address these questions, we designed a mixed-methods study. First, we explored

students’ open-ended responses to case material about a situation of social exclusion

for the strategies they recommend and the justifications they offer for their choices.

Next, we translated the qualitative codes into quantitative categories in order to

examine the influences on students’ choice of strategy. We interpret the implications

of our findings for research in social development as well as school-based social

development programming.

Methods

The research context and procedure

The research described in this paper is part of an ongoing collaborative project with

Facing History and Ourselves (hereafter Facing History),1 an international

organisation that provides resources for teachers and a program for students that

focuses on human behaviour and ethics as a bridge between history and the self

(Tollefson et al., 2004).

Schools with teachers who were known to use the Facing History program and

who were willing to be included in a quasi-experimental study were recruited to

participate in this study. A second school was then recruited within the same town

that matched for similar characteristics. All students in the selected classrooms were

eligible to participate in this research. Only students who returned written parental

consent forms were included in the study; participation rates were above 75% across

the classrooms. Measures were administered during class time by the teachers in the

study and were then turned over to members of the research team. No identifying

information was collected and students’ confidentiality was guaranteed. The

students completed the measures before they had any contact with Facing History

materials and so the data are not construed as evaluation outcome data.

Setting and participants

The participants in this study were students in five public middle schools in

Massachusetts. The schools were all located in districts with a predominantly white,

middle-class student population. As Table 1 shows, there is some variation in

socioeconomic and educational characteristics across the schools, as suggested by

the percentage of students who qualify for free or reduced price lunch (information

about the students in this study and the schools they attended was provided by the

Facing History organisation).

Moral choices 169

The sample for this study includes 168 students in five Grade 8 classrooms, one

from each school in the study. One hundred and four girls and 64 boys participated.

Students ranged in age from 12–13 years old. No other individual-level demographic

information is available.

Measures

After reading the case material about Eve Shalen, participants completed the In

Group Assessment. This is a survey measure, with two primary questions based on

the case study of Eve Shalen and seven follow-up open-ended items. The measure

was designed to challenge students to assume different vantage points, such as

bystander or victim, about a situation of social exclusion, as well as to capture their

thinking about the issues of interpersonal relationships and social exclusion more

generally (Barr, 2005). The content and language of the assessment questions rely

on the discourse used in Facing History, while the structure of the questions is

driven by a psychological theory about how children develop the capacity to

coordinate various social perspectives, which is central to their understanding and

negotiation of social relationships as well as their developing social awareness

(Selman et al., 1997; Selman, 2003).

For the purposes of this study, we analysed students’ responses to two questions

on the In Group Assessment. The first question asked students to respond as a

witness or bystander to the social exclusion: 1(a) List at least two different ways that

Eve could have acted when she witnessed her classmates picking on other students.

(b) Which would be the best way? (c) Why would that be the best way?

In the second question, students were asked to suggest choices when invited to

join the perpetrators: 2(a) List at least two things Eve could have done when she was

invited to join in the teasing of the other girl. (b) Which would be the best way? (c)

Why would that be the best way?

In each of the two questions, students were asked to write about the choices they

perceived for negotiating the situation and their justifications for each choice. This

allowed us to compare strategies and justifications across each of the two vantage

points. Data for this study include all students who complied with the instructions

for both questions.

Table 1. Sample demographics (n5168)

School n Race (% white of the total

population)1 % Eligible for free/reduced

price lunch1

A 34 81.0 9.6

B 63 92.0 4.8

C 19 88.0 7.1

D 14 80.0 30.5

E 38 80.0 27.7

Notes: 1Massachusetts Department of Education (2004)

170 L. F. Feigenberg et al.

Analysis plan

Qualitative analysis: coding development

To address the first research question, we initially examined the responses to code

for the strategies students suggested for negotiating the choice to witness or join in

the social exclusion. Using a semi-grounded approach (Glaser & Strauss, 1967), we

allowed the data, rather than theory, to guide these analyses. First, we open-coded

the responses by grouping similar strategies together. We then clustered the

suggested strategies by salient themes, which resulted in three strategy categories.

The codebook for these strategy categories, including exemplars of each, is shown in

Table 2.

To examine the second portion of the responses, where students justify why their

chosen strategy would be the best one for the situation, we focused on the factors in

the social environment students perceive to be most salient when choosing a

strategy. We used an inductive approach (Boyatzis, 1998) to allow both the data and

theory to guide the coding. This analysis involved a greater amount of interpretation

on our part than the analysis for strategy codes, as we attempted to capture key

elements about how the strategy was explained and the reason given for its selection.

Table 3 provides the codebook for the four justification categories as well as

examples of each.

A three-person team coded the data. Each phase of coding began with a discussion

of the codebooks for strategy and justification (as shown in Tables 2 and 3), the

establishment of anchor responses and independent scoring until an acceptable level

of inter-rater agreement was reached. Both the strategy and justification categories

had over 92% direct agreement and Kappas greater than .9 (Bakeman & Quera

Table 2. Codebook for strategy categories

Code description Anchor response

Upstand N Requires intervention in the existing

situation of ostracism

‘She could of told them to stop, and

that they were being mean.’

N Articulates an action that assists the

victim, such as standing up to the group

or comforting the victim

N Aligns against the mocking of the girl

‘She should make friends with that

girl.’

‘She should of just told a teacher what

was going on.’

Perpetrate N Aligns with the group mocking the girl ‘She could play along and make fun

of the kids too.’

‘She should go with the girls.’

‘Ignore them and just go on with her

business.’

‘Make an excuse and walk away.’

N Action implies that the invitation to join

the mocking has been accepted

N Contributes to the existing situation of

ostracism

Bystand N Aligns with neither the victim nor the

group doing the mocking

N Avoids involvement with the existing

situation of ostracism

N Uninvolvement, active as in walking away,

or passive, by minding one’s own business

Moral choices 171

[1995] suggest .6 to .8 to be adequate and above .8 to be excellent). Approximately

20% of the sample was used during training and reliability procedures. Once the

training was completed, the remaining surveys were divided among the raters and

scored independently. All of the surveys were used in subsequent analyses.

Quantitative analysis

In order to examine the relationships between strategy and justification categories,

we created categorical variables and assigned each code a numerical value (0–2 for

Table 3. Codebook for justification categories

Code description Anchor response

Conventional N References social norms, conventions or

rules (formal or informal) as the main

guiding principle

‘It is the right thing to do.’

‘It would be easier.’

‘Because I think it’s the best.’

N Highlights the efficiency or expediency

of the recommended strategy

‘It would keep everyone out

of trouble.’

N Does not explicate reasoning beyond

simple explanations of cost-benefit

analyses that imply one action is simply

‘better’ than another

Safety N Indicates protection as a priority ‘So they don’t start picking

on her too.’

‘That way nothing bad

happens to the victim.’

N Perceives an immediate threat to one’s

emotional or physical well-being

N Indicates that the main goal is to stop

the current situation of ostracism ‘To make sure she doesn’t get

hurt.’N Does not reference long term

consequences or implications of

recommended strategy

Relational N Highlights the formation or maintenance

of interpersonal relationship(s)

‘And I could be considered

the ‘‘Popular’’ girl.’

N Articulates desire for belonging or

connectedness with another person or

with a group of people

‘Because she’d feel like she fit

in.’

‘Because she knows what it

feels like.’N Identifies a connection between people’s

experiences or emotions

Prosocial

Transformational

N Explains connections between the

recommended action and possible future

consequences or implications

‘They might realize they’re

doing the wrong thing and

not do it again.’

N Speculates about the possible

development of or changes in other

people’s thinking or beliefs

‘I believe that if enough

people are willing to do

something about a

problem, the problem

would not exist anymore.’

‘She would make a good

influence on other people.’

N Articulates opportunities for group

dynamics to shift as a result of the

recommended action

N Implies that the recommended action

coul