ArticlePDF Available

Partner Exploitation and Violence in Teenage Intimate Relationships

Authors:
Partner exploitation and violence
in teenage intimate relationships
Christine Barter, Melanie McCarry, David Berridge and Kathy Evans
October 2009
www.nspcc.org.uk/inform
Barter et al Partner exploitation and violence in teenage intimate relationships Acknowledgements
Acknowledgements
First and most importantly, our deepest thanks go to all the young people who
participated in the research, especially those who talked with us so openly about their
experiences. Due to confidentiality their names must remain anonymous. We are
indebted to the eight schools that agreed to participate in the research, for their
enthusiasm and for always making us feel welcome despite the inevitable disruption
such a partnership entails.
This research was funded by the Big Lottery Fund and we are very grateful for all the
assistance and support they provided over the three years of the project. Our great
thanks go to the members of our young people’s advisory group, who worked with us
on all aspects of the research from the beginning of the project until completion; their
contribution cannot be overstated. We also express our appreciation for the support
and advice provided by members of the adult advisory group and especially Donna
Chung and Nicky Stanley for reviewing the final report on our behalf. We would also
like to thank Annabel Lander for her administrative support and help over the
duration of the research. We are grateful to Demi Patsios, Marsha Wood and Dinithi
Wijedasa for assisting with the statistical analysis. Any omissions or limitations in the
research and this report of course remain our responsibility.
Our gratitude also extends to colleagues from the Violence Against Women’s
Research Group, University of Bristol, who generously volunteered to assist in
undertaking the survey in a number of schools. Finally, we appreciate the
encouragement and support of colleagues at the NSPCC, especially Lorraine Radford
and Phillip Noyes, and the University of Bristol.
©NSPCC 2009 2
Barter et al Partner exploitation and violence in teenage intimate relationships Contents
Contents
Acknowledgements ......................................................................................................2
Contents........................................................................................................................3
List of figures................................................................................................................5
List of tables..................................................................................................................6
Introduction..................................................................................................................7
US research on “dating” violence..........................................................................7
UK context.............................................................................................................9
Professional response to teenage partner violence...............................................10
Research aims and methodology .........................................................................11
Section 1: Survey sample...........................................................................................26
Gender and age of participants ............................................................................26
Nationality of participants....................................................................................26
Ethnicity of participants.......................................................................................27
Religion of participants........................................................................................28
Family composition .............................................................................................29
Disability..............................................................................................................30
Section 2: Young people’s experiences of family and peer violence......................31
Family violence....................................................................................................31
Peer violence........................................................................................................34
Family and peer violence.....................................................................................36
Summary points...................................................................................................36
Section 3: Young people’s relationships ..................................................................38
Gender and age ....................................................................................................38
Ethnicity and religion...........................................................................................38
Family violence....................................................................................................39
Disability..............................................................................................................39
Survey results for respondents who had been in a relationship...........................40
Summary points...................................................................................................43
Section 4: Recipients of teenage partner violence...................................................44
Physical partner violence.....................................................................................44
Emotional partner violence..................................................................................55
Sexual partner violence........................................................................................65
Section 5: Instigation of teenage partner violence..................................................74
Physical partner violence.....................................................................................74
Emotional partner violence..................................................................................78
Sexual violence....................................................................................................81
Section 6: Further multivariate statistical analysis................................................85
Victimisation........................................................................................................86
Instigation ............................................................................................................89
Section 7: Interview findings.....................................................................................92
Interview sample..................................................................................................92
Part 1: Interview findings – girls .........................................................................94
Part 2: Interview findings – boys.......................................................................156
Section 8: Discussion and implications ..................................................................178
Recognition of teenage partner violence as a significant child welfare issue...178
Impact of teenage partner violence – the gender divide....................................179
Different forms of partner violence ...................................................................182
©NSPCC 2009 3
Barter et al Partner exploitation and violence in teenage intimate relationships Contents
Family and peer violence...................................................................................188
Older boyfriends ................................................................................................188
Disadvantage......................................................................................................189
Ethnicity and religion.........................................................................................190
Same-sex partners..............................................................................................192
Enhancing help-seeking.....................................................................................193
Conclusion .........................................................................................................196
Bibliography.............................................................................................................198
Appendix I: Logistic regression – methodological note........................................206
Appendix II: Summary of logistic regression models...........................................208
©NSPCC 2009 4
Barter et al Partner exploitation and violence in teenage intimate relationships List of figures
List of figures
Figure 1 Age of young people.....................................................................................26
Figure 2 Nationality of young people .........................................................................27
Figure 3 Ethnicity of young people.............................................................................28
Figure 4 Religion of young people..............................................................................29
Figure 5 Family violence by gender............................................................................32
Figure 6 Impact of physical partner violence by gender (per cent).............................46
Figure 7 Who young people told about physical partner violence (per cent) .............49
Figure 8 Who young people told about emotional violence (per cent).......................61
Figure 9 Impact of sexual violence (per cent).............................................................68
©NSPCC 2009 5
Barter et al Partner exploitation and violence in teenage intimate relationships List of tables
List of tables
Table 1 Schools sample..............................................................................................14
Table 2 Family compositions and gender..................................................................30
Table 3 Have any of your partners ever used physical force such as pushing,
slapping, hitting or holding you down? ........................................................44
Table 4 Have any of your partners ever used any more severe physical force such
as punching, strangling, beating you up or hitting you with an object? .......45
Table 5 Young people’s experiences of emotional violence.....................................57
Table 6 Have any of your partners ever pressured you into kissing, touching or
something else?.............................................................................................66
Table 7 Have any of your partners physically forced you into kissing, touching
or something else?.........................................................................................66
Table 8 Have any of your partners ever pressured you into having sexual
intercourse?...................................................................................................67
Table 9 Have any of your partners physically forced you into having sexual
intercourse? ...................................................................................................67
Table 10 Have you ever used physical force against your partner? ............................74
Table 11 Have you ever used more severe physical force against your partner?........75
Table 12 Have you ever pressured your partner into kissing, touching or something
else?...............................................................................................................81
Table 13 Have you ever physically forced your partner into kissing, touching or
something else?.............................................................................................82
Table 14 Have you ever pressured your partner into having sexual intercourse? .......82
Table 15 Have you ever physically forced your partner into having sexual
intercourse?...................................................................................................82
Table 16 Interview sample...........................................................................................93
©NSPCC 2009 6
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
Introduction
Although a substantial body of UK evidence exists on adult experiences of domestic
violence, including the impact on children and professional practice (Mullender et al
2002; Hester et al 2007; Radford and Hester 2007), we know very little about
teenagers’ own experiences of partner violence in their intimate relationships. In
contrast, within the USA a significant amount of research has emerged over the past
two decades on young people’s experiences of partner or “dating” violence and abuse.
US research on “dating” violence
Most of the US and indeed wider international literature has adopted the term
“dating” to describe this area of research. However, this terminology does not transfer
well to the UK context, as young people do not use or indeed recognise this term. In
addition, dating seems to imply a degree of formality that does not necessarily reflect
the diverse range of young people’s intimate encounters and relationships. Similarly,
Brown et al (2007), argue that the term is outdated and too restrictive. The term
“partner exploitation and violence” seems more appropriate (we acknowledge this
also has some limitations); to aid brevity “partner violence” will be used in this report.
Terminology is not simply an interesting academic exercise as, unless appropriate
definitions are used by professionals, young people will not perceive any intervention
as being relevant to themselves or their specific circumstances. In addition, the terms
“young people”, “adolescents” and “teenagers” are used interchangeably to refer to
children under the age of 18.
Primarily, US research has focused on physical or sexual violence, although
psychological and coercive control mechanisms have now begun to be explored. US
research confirms the high prevalence of partner violence in young people’s
relationships and its negative impact (Jackson et al 2000; Hickman et al 2004). One of
the current authors has published a detailed critique of the “dating” violence literature
(Barter 2009), so we will not repeat this here in any depth. Previous research findings
on teenage partner violence are included alongside our findings throughout the report.
©NSPCC 2009 7
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
Some findings, primarily from US survey-based studies, suggest that boys and girls
use similar levels of physical and emotional violence towards their partners (Foshee
1996; Arriaga and Foshee 2004; Roberts et al 2005), resulting in propositions that
teenage partner violence demonstrates a greater degree of gender symmetry compared
to adult domestic violence where women are predominantly the victim. However,
research also shows that girls are more likely to be the recipient of serious physical
and sexual violence than are boys (Lane and Gwartney-Gibbs 1985; Roscoe and
Callahan 1985; Muehlenhard and Linton 1987; Gamache 1991; Foshee 1996;
Silverman et al 2001; Ackard et al 2003; Arriaga and Foshee 2004). In addition, while
both boys and girls use verbal violence and control mechanisms, the impact of these
on girls appears to be much greater than for boys (Lavoie et al 2000; Sears et al 2006).
What we know about teenage partner violence testifies to its serious consequences for the
wellbeing of victims and their future life prospects (Tangney et al 1992; Harway and Liss
1999; Smith et al 2003). US research has also identified that teenage partner violence is
associated with a range of adverse outcomes for young people, including mental health,
depression and suicide (Silverman et al 2001; Collin-Vézina et al 2006). Studies indicate
that adolescent partner violence is strongly associated with experiencing domestic
violence in adulthood (O’Leary et al 1989; Cleveland et al 2003).
Furthermore, US research has identified a range of risk factors that may increase a
teenager’s susceptibility to partner violence. These risk factors include previous
experiences of parental domestic violence, physical and sexual abuse, and violent peer
groups (Roscoe and Callahan 1985; O’Keefe et al 1986; Smith and Williams 1992;
O’Keefe and Treister 1998; Wolfe et al 2001a; Simonelli et al 2002; Whitfield et al
2003). While less researched, it appears that parental neglect, especially lack of
supervision and involvement/interest in their teenage children’s lives, also impacts
negatively on young people’s vulnerability to partner violence (Lavoie et al 2002;
Pflieger and Vazsonyi 2005), although what constitutes “neglect” for adolescents has
still to be fully explored within the research literature (see Rees and Stein 1999).
The majority of US studies have used large-scale surveys to determine the incidence
of or prevalence rates for teenage partner violence, often across different sample
communities. However, this approach has meant, with a few noticeable exceptions
©NSPCC 2009 8
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
(see Banister et al 2003; Chung 2005; Jackson 1999; Sears et al 2006; Silverman et al
2006; Próspero 2006), that young people’s own experiences, views and agency have
been neglected (James et al 1998; Mullender et al 2002). Furthermore, some
commentators argue that past dependence on quantitative methodologies, primarily
based on measures created for adults, means that research has been unable to capture
the range and complexity of this social problem (Jackson 1999; West and Rose 2000;
Sears et al 2006). For example, Connolly et al (2000) and Chung (2005) both argue
that research needs to reflect the fluidity of teenage relationships, producing
typologies to describe different teenage intimate encounters. Researchers, therefore
face considerable challenges in operationalising teen relationships in a manner that
captures their variation in form and meaning. Thus, according to Hickman et al
(2004), in comparison with adult studies, research involving adolescents may require
a more resource-intensive methodology to capture this meaning.
UK context
Teenagers’ experience of partner violence has not received the same degree of
attention in the UK, either within research or practice. The small number of UK
studies undertaken confirms its impact and seriousness for young people’s welfare
(Hird 2000). Given the critical nature of adolescence as a developmental period, it is
surprising that so little attention has been given to this social problem (O’Keefe et al
1986; Williams and Martinez 1999. This is despite the fact that young people
themselves have repeatedly identified peer relationships, especially those involving
abuse and violence, to be among the main areas of anxiety and unhappiness in their
lives (Utting 1997; Barter et al 2004).
Young people have also clearly articulated that they do not feel their views and
wishes are taken seriously or acted on by professionals. Consequently, professional
practice may not be responding to or reflecting young people’s own concerns, fears
and wishes regarding the impact of peer violence. A number of obstacles have been
presented to explain this omission, including viewing behaviour as experimental, fear
of stigmatisation, a mistaken view that peer abuse is less harmful than abuse by
adults, the unawareness of adolescent abuse generally and low reporting levels (Hird,
2000; Barter 2009).
©NSPCC 2009 9
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
Professional response to teenage partner violence
Only a minority of US studies (Brown et al 2007), and it appears none within the UK,
have addressed multi-agency professional practice within this area of child welfare.
Intervention work on teenage partner violence has been developed in some schools –
see Bell and Stanley (2006) for an evaluation of school initiatives in this area. This
general disregard of the area may be due to the emphasis within child welfare research
and practice on adult–child interactions, be they neglectful/abusive or in a
professional context. However, recent governmental guidance in Working Together to
Safeguard Children (HM Government 2006) has, for the first time, officially
recognised the need for professionals to safeguard children from harm arising from
abuse or violence in their own relationships. This is, at least in part, a response to the
Bichard Inquiry (Bichard 2004) into the killings in Soham of two 10-year-old girls by
their school caretaker. The inquiry raised, among other things, concerns about the way
that professionals dealt with previously known and alleged cases of violence and
under-age sex involving the murderer.
The guidance states that, in cases involving sexually active children under the age of
13, there is a presumption that the case will be reported to social work services (HM
Government 2006, 5.25). For young people under 16, consideration must be given as
to whether there should be a discussion with other agencies regarding the need for a
referral, including where both parties are below this age and in a consensual
relationship (5.26). To assist with this assessment a risk-checklist is provided, which
includes the child’s living circumstances and background, age imbalance, overt
aggression or power imbalances, evidence of coercion, attempts at secrecy by the
partner, and denial, minimisation or acceptance of the violence by the victim (5.27).
However, we have no information regarding practitioners’ awareness of this guidance,
how it is being interpreted and operationalised, either nationally or locally, and what
multi-agency decision-making models and protocols are in place.
The need for professionals both to recognise and to respond to partner violence within
young people’s relationships is indisputable (Barter 2006a). This research seeks to
provide evidence on which to base more effective intervention and prevention
©NSPCC 2009 10
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
programmes aimed at safeguarding all young people in their teenage intimate
relationships.
Research aims and methodology
This remaining part of this introduction outlines the aims of the study, the
methodology used, ethical issues and the sampling framework.
Young people’s advisory group
First, it is important that the work of the young people’s advisory group (YPAG) is
acknowledged. At the proposal stage of the project it was clear that for the research to
be successful we needed to work closely with young people. With assistance from an
NSPCC school team, the researchers approached a local school, which agreed that we
could ask if any pupils would like to volunteer to be a member of a research advisory
group for a two-year period. Young people who were interested in participating were
required to return a consent form from a parent. Eventually we recruited 12 young
people, all aged 14, to be on the group. The YPAG was convened within three months
of the start of the project and followed the research through until completion.
Although the aims of the study were determined before the YPAG was convened, the
group was influential in refining these, ensuring that suitable terminology was being
used and young people’s views were reflected in the research aims. The YPAG helped
to design and develop all of the data collection tools, provided consultation on
fieldwork issues, analysis and commented on findings and dissemination. For
example, the research team had developed an outline survey, which had been
positively reviewed by the adult advisory group. However, the YPAG made
substantial comments on the type and range of questions, the format and style of the
questionnaire itself and how the instructions were presented. Overall, acting on their
comments, 20 questions were removed, others were rephrased and the survey format
and instructions were simplified. We piloted the survey with 60 pupils, aged 15 and
16, from the same school. All provided very positive comments, which endorsed the
changes initiated by the YPAG.
©NSPCC 2009 11
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
Some members of the YPAG also made a short DVD in which they introduced the
research, explained why the research was important, what participation entailed,
including consent and confidentiality, and how the findings would be used. The
researchers played the DVD to pupils in participating schools to introduce the
research and to show how young people were involved in the project. At the
completion of the project the YPAG came to the University of Bristol for the day to
provide feedback on findings, discuss practice implications and to say what they
thought should be done next. Certificates of attendance were presented at the final
meeting. Although, as with any advisory group, not everyone agreed on certain points,
the group worked to ensure that all comments were given equal weight within the
research process. We are very grateful to the YPAG for all their hard work and
enthusiasm over the two-year project. We are also very grateful to their school,
especially to one particular teacher, who provided valuable support to us and gave us
a great deal of their time to ensure the group ran smoothly.
Alongside the YPAG, we also had a professional advisory group, which met twice at
the beginning of the project and provided detailed suggestions and comments
regarding the research tools. Throughout the research individual members of the
group were asked to comment on particular research issues relevant to their specific
expertise. The group also provided very important links in relation to gaining access
to schools. The researchers undertook presentations at the request of some members
of the group. These presentations provided valuable feedback on how different
professionals viewed the emerging findings and what were seen as the main issues for
their own practice in this area. Members of the group also provided detailed
comments on the final draft report. The researchers are very grateful for their valuable
contribution to the project.
Research aims
The overall aim of the research was to examine in depth young people’s experiences
of partner violence and how best to respond to this problem. The study utilised a wide
definition of violence (Stanko 2000), which incorporated emotional, verbal, physical
and sexual forms of violence. These were looked at both in isolation and as they co-
exist within young people’s relationships.
©NSPCC 2009 12
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
The more specific objectives included exploring:
the nature, frequency and dynamics of different forms of violence in teenage
relationships and the contexts in which they occur and persist
young people’s own understandings, attitudes to and evaluations of such violence
if any groups are particularly vulnerable to becoming victims and/or assailants
young people’s perceptions of the impact of violence on their wellbeing, the
coping strategies they used, their effectiveness, as well as factors that enabled
young people to leave violent relationships
the “reasons” why assailants used violence and their perceptions of its impact on
their victims and themselves
young people’s views on prevention and how support services could best respond
to preventing violence
wider cultural or social processes that underpin experiences of exploitation and
violence.
Multi-method approach
To fulfil the above objectives the research used a multi-method framework combining
both quantitative and qualitative approaches (Plano Clark and Creswell 2008). Two
data collection techniques were used:
self-completion survey
face-to-face, semi-structured, in-depth interviews with vignettes.
Sample
Schools sample
The schools sample was chosen to reflect previous research findings that suggest
deprivation and locality are possible risk factors associated with some forms of
teenage partner violence victimisation (Hird 2000). Eight secondary schools
participated in the research. Four schools were in England, two in Scotland and two in
Wales (see table below). We wanted some geographical spread so as to represent a
range of circumstances and to build on previous research relationships. Four schools
were in suburban areas surrounding large cities, three schools were situated in
predominantly rural areas and one was in an inner city area. The criterion of free
©NSPCC 2009 13
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
school meals was used to gauge the level of social derivation in each of the school’s
catchment areas. Applying this criterion, two schools were located in areas of very
high deprivation, two in areas of high deprivation and four in areas of average to low
deprivation. All eight schools were in the public sector. Only one school was faith-
based (Catholic). Three of the schools had sixth forms attached.
Table 1 Schools sample
School Deprivation level Location Sixth form
England 1 (E1) High Suburban Yes
England 2 (E2) Very high Suburban No
England 3 (E3) Average Inner city Yes
England 4 (E4) Low Rural Yes
Scotland 1 (S1) Very high Suburban No
Scotland 2 (S2) High Suburban No
Wales 1 (W1) Average Rural No
Wales 2 (W2) Average Rural No
Schools were accessed through a range of strategies, ranging from the research team
directly approaching schools, being “introduced” to schools through the NSPCC’s
school teams and from assistance through other organisations. In addition to the above
eight schools, three more were approached to take part but declined. In two cases
internal school issues were presented as the explanation. The third school, having
spent months in negotiations, stopped any correspondence without providing a reason.
Two of these schools were within the independent sector and one was a state school
situated in a very affluent area. Unfortunately, due to the time constraints of the
project and the very time-consuming access negotiations required for each school, we
were unable to pursue the inclusion of any further schools from the independent
sector. Thus, although we attempted to increase the number of participants from more
affluent backgrounds, we were unable to do so.
Survey participation rates
Pupil participation rates differed substantially between the eight schools. This was due
to two main reasons: the way that parental consent was gained and administration of
the fieldwork in schools. In six of the schools, agreement was gained to use opt-out
parental consent for young people. Thus, in these schools, parents were required to
©NSPCC 2009 14
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
send back participation forms by the given date only if they did not want their child to
participate. We received back only 19 opt-out forms from parents.
However, the administration of the survey in schools also significantly affected
participation rates. In four schools the survey was undertaken in scheduled lessons,
generally personal, social and health education classes (PSHE). In these schools
researchers directly asked young people if they would like to take part. Only a
minority of young people decided not to contribute and were given alternative work to
do. In these schools participation rates (as a percentage) were 94, 81, 79 and 71 per
cent. These percentages are based on comparing class lists with consent forms. This
does not take into account the fact that some young people were away from school,
undertaking other activities or in exams. Thus participation levels reflect how many
young people took part, but not how many declined to be involved, as some young
people were absent for other reasons.
In the two other schools, where opt-out parental consent was agreed, very different
administrative procedures were followed. Although researchers still initially
introduced the research, when the team returned to undertake the fieldwork they did
not secure direct access to pupils. Instead teachers in each class asked young people
on our behalf if they would like to take part. Those pupils who agreed to participate
were taken from their classroom to where the survey was taking place. However, in
many cases teachers were reluctant to encourage pupils to contribute, as this would
entail missing lessons and, in some instances, exam revision sessions. Consequently,
participation rates in these two schools (based on average class size as we were unable
to gain access to class lists) were reduced to 56 per cent and 52 per cent. We tried to
avoid busy times of the year, such as the run-up to exams but this was not always
possible given that fieldwork had to be completed within 12 months.
In the remaining two schools the education authority had a policy of opt-in parental
consent for all research with children and young people. Thus parents were required
to complete the consent form and return it to the research team in order for their child
to participate. Headteachers of both these schools stated that they often experienced
great difficulty in having forms returned by parents. Interestingly, both headteachers
said they felt opt-out consent would have been appropriate. In both cases the warnings
©NSPCC 2009 15
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
given were accurate. Roughly only a third of parents returned consent forms. Thus
significantly lower participation rates were recorded for these schools: 38 per cent and
31 per cent. No young people whose parents had agreed to their participation declined
to take part and many young people wanted to be involved but, for whatever reason,
did not have a signed parental consent form. We do not know if parents actively chose
to prevent their child’s participation or if this non-response simply reflected a much
more general pattern, as indicated by the school headteachers. Due to the low parental
response rate the education authority subsequently agreed to a request by the research
team, supported by a headteacher, to allow young people aged 16 years and over to
participate with opt-out parental consent. We, therefore, returned to one school
towards the end of the fieldwork to undertake the survey on this basis. The initial
participation rate with opt-in parental consent of 31 per cent increased to 83 per cent
with opt-out consent.
Overall, 1,377 young people completed the survey; 24 (2 per cent) of these were
spoiled or completed incorrectly, thus 1,353 questionnaire responses were analysed
(see section 1 for more detail of the survey sample breakdown). A total of 91
interviews (62 girls and 29 boys) were undertaken (see section 7 for more details).
Survey aims
A self-completion questionnaire was developed for young people aged between 13
and 16. The survey aimed to identify if respondents had ever experienced a range of
violent behaviours from their partners and/or ever acted in this way towards their
partners. Thus, the survey sought to understand, not only if young people had ever
been a target for or perpetrated partner violence, but also if and how these two
experiences were associated. A partner was defined in the research as anyone the
young person had been intimate with, ranging from a serious long-term boyfriend or
girlfriend, to a more casual partner or a one-off encounter. The main aims of the
survey were to document:
demographic characteristics of respondents
incidence rates for different types of violence/behaviours in partner
relationships
identification of the broad dynamics involved
©NSPCC 2009 16
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
impact on young people’s welfare
help-seeking behaviour and support
past experiences of child abuse/inter-parental violence and wider peer
violence.
Survey ethics and consent
Leaflets explaining the research aims, what was involved in participation, issues of
confidentially and anonymity, and who the researchers were, including a photograph,
were produced for the young people and parents. Due to the sensitive nature of the
research, the word “violence” was not used; instead the project was introduced more
broadly as a research study on teenage relationships and the concerns some young
people may have in relation to these. Consent leaflets were designed both for young
people and for parents. All young people who wanted to take part were required to sign
a consent form. The research was developed on the basis of opt-out parental consent. If
a parent did not want their child to take part, they were therefore required to complete
the consent form and return it, in the stamped addresses envelope provided, to the
research team by a specified date. However, one education authority insisted on opt-in
parental consent for the two schools within their area. Here, parents who wanted their
children to take part were required to complete the form and return it before their child
could participate. Some of the implications of these two systems of gaining parental
consent are presented under “Survey participation rates” on page 14.
Implementation of the survey
All surveys were confidential and anonymous. It was decided in both the YPAG and
the professional group that, unless participants were guaranteed anonymity, many
would be reluctant to answer such sensitive questions, especially concerning physical
and sexual violence. We asked young people their views regarding these issues at the
pilot stage of the research and received almost unanimous agreement in support of
total confidentiality and anonymity.
It is important to recognise the very sensitive nature of some of the questions
contained in the survey. The research team and members of the two advisory groups
were both very aware of the importance of ensuring that questions were asked in an
©NSPCC 2009 17
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
appropriate manner. However, language describing violence, and perhaps especially
sexual violence, can be problematic. For example, it was decided to use the phrase
“physically forced into sexual intercourse” rather than the word “rape”. Some
members of the adult advisory group raised initial concerns that the wording of
certain questions, especially this one, might lead some young people to minimise the
impact and consequences of their actions. Thus, if the word “rape” were included,
participants then would be more aware of the implications of their behaviour.
However, both advisory groups also recognised that few young people would answer
a question pertaining to rape, or indeed necessarily recognise their own experience of
violence in this way, a presumption that was subsequently upheld in the interview
data. This was an important issue requiring considerable thought. However, the
research team decided that unless we used language that was more oblique and did not
hold such direct connotations, we would severely limit young people’s ability or
willingness to participate fully in the survey.
Additionally, as participants for the second interview stage of the research were
directly recruited following the survey, researchers were unable to provide any direct
comments to pupils regarding the questions asked in the survey, as this may have
influenced participants’ subsequent interview responses. However, at the end of each
survey session all young people were provided with a leaflet containing information
on local and national sources of help.
We included in the sample a range of pupils within schools across year groups and
educational attainment. Sometimes choice of respondents was restricted due to school
or external priorities, such as tests or exams. Researchers introduced the study to
individual classes or, in the case of three schools, in the school assembly. Leaflets and
consent forms were given out to young people and any questions answered.
Researchers then returned to each school within three weeks to undertake the survey.
It was made clear to young people that participation was voluntary and they did not
have to take part if they did not wish. Two researchers were present in each session so
that young people who required reading support could be assisted. A “fun quiz” was
placed at the end of the survey so that children who had not experienced a
relationship, and therefore completed the survey sooner than others, could not be
identified by other pupils. This activity also provided a “wind-down” time for all
©NSPCC 2009 18
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
participants at the end of the session, as young people’s answers to the quiz were
compared. In some cases teachers remained in the classroom and were invited to
complete the quiz and compare answers, often to the amusement of their pupils.
As already discussed, the questionnaire included sensitive questions, which some
young people may have found disconcerting. Researchers took great care to ensure
that any participants who appeared to be affected by the questionnaire were discreetly
approached afterwards to see how they were feeling and if they wished for someone
in the school to talk with them. In addition, all researchers were available if any
young person wanted to discuss any of the issues further. No young people asked for
any additional help.
Aims of the qualitative interviews
The key aim of the second qualitative stage of the study was to enable young people
to articulate their experiences in greater depth and allow participants to wield greater
control over the focus and direction of the research (see Alldred 1998; Mayall 2002).
The main aims of the qualitative methodology were to:
engage young people to participate in “sensitive research”
enable a discussion of personal experiences of partner violence
explore the interpretations and meanings that young people ascribe to different
forms of partner violence
provide young people with greater control over the research interaction.
To fulfil these diverse aims, two complementary techniques were used: semi-
structured interviews and vignettes.
Semi-structured interviews
In-depth interviews enabled young people to identify, define and contextualise their
own personal experiences of partner violence. Within the interview, participants were
given considerable freedom to introduce relevant areas and factors that they viewed as
important in understanding these issues. Much methodological literature exists
concerning the use of qualitative semi-structured interview techniques, which was
drawn on within the research. We will not rehearse this here. Researchers used “active
©NSPCC 2009 19
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
listening” and “non-directional stance” (Whyte 1984), and conveyed a non-
judgemental attitude (Hill 1997). This was not always necessarily an easy stance to
maintain in interviews, especially when participants discussed their own use of
violence, often with little recognition of the impact of this on their partners.
Interviews took place in five of the eight schools, with a total of 91 interviews. In
each of these schools the qualitative interview followed several weeks after the
completion of the questionnaires. It was initially envisaged that young people would
self-select to take part in the interview by contacting the researchers, for example by
use of a self-addressed envelope or a text number, or through volunteering after
completing the questionnaire. All these strategies were found to be unsuitable at the
pilot stage of the research. Due to issues of confidentiality and anonymity,
questionnaires did not intentionally contain personal contact details, thus prohibiting
their use for selection of interview participants. We therefore decided to move to a
system where researchers asked a range of participants completing the questionnaires
in the classrooms if they would also like to take part in an interview. Researchers
observed to see which young people seemed to be engaging with the survey. They
then asked those young people if they would like to take part in the interview stage.
However, to ensure they were not perceived by their peers as being targeted for a
particular reason, we also invited a pupil sitting alongside them to participate. Thus it
appeared to the class that groups of pupils were asked rather than any individual. This
system worked well and enabled young people to participate without the need for
them to be seen to volunteer in front of their class. Young people received a £10 gift
voucher as a token of our appreciation for being interviewed; this also provided them
with a “valid” excuse for volunteering in front of their peers (Barter and Renold
2003). This selection method, therefore, inevitably entailed interviewing some young
people who stated that they had no experience of partner violence.
Nevertheless, these interviews proved important for a number of reasons. First, we
were able to identify protective factors in some young people’s lives, which seemed to
be associated with their positive relationship experiences. Second, although some
young people firmly stated at the beginning of the interview that they had not
experienced any such problems in their relationships, further discussions revealed a
more complicated picture involving subtle forms of control. These incidents had not
©NSPCC 2009 20
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
always been reported in their survey responses. Thus, if a researcher had not asked
these young people to participate, it is doubtful that they would have volunteered and
these more divergent experiences would have been lost. Consequently, the qualitative
dimension of the project provides evidence that, for certain forms of behaviour, the
quantitative survey findings may under-represent the incidence. This is an important
factor for interventions for young people. If young people do not recognise or define
their experiences as constituting partner exploitation or violence, they will be less
likely to engage with intervention programmes that focus exclusively on overt aspects
of teenage violence and control. Although some young people had no direct personal
experience of partner violence, many of them did have friends who had revealed their
own experiences to them. Thus, we were able to explore with these young people their
responses to these disclosures. This is an important area due to young people’s
reliance on peers as a source of assistance and support, identified in previous research
(Hird 2000; Ashley and Foshee 2005; Ocampo et al 2007). Section 7 of this report
also elaborates on these issues.
Some young people talked very openly about their experiences of partner violence.
However, in a number of interviews, researchers felt that young people were more
reticent and cautious with their replies. We often recalled that towards the end of
these discussions, the young people seemed to begin to feel more comfortable;
unfortunately in many instances the interview had to end due to time restrictions.
Therefore researchers felt that for some young people, a single interview was
insufficient to discuss such personal, and often hidden, experiences. This may also
have been influenced by the fact that all interviews took place in school. For some
participants the location may have been a barrier to talking openly about such
sensitive issues. We tried to ensure that all interview rooms were private and schools
had attempted to facilitate this. However, in some instances this was not always
possible. For example, interviews sometimes had to take place in a library, which
although separate was not private. Furthermore, in a small number of cases interviews
were interrupted by teachers and also occasionally other pupils. We are unable to
gauge the impact of this on young people’s ability to talk freely to us. Such issues
need to be taken into account in future research methodology in this area.
©NSPCC 2009 21
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
Young people were asked if they would like to be interviewed alone or with a friend.
In three cases young people preferred to be interviewed with their friends. Thus we
had one joint interview with two girls and a group interview involving three girls.
Only one joint interview with boys occurred. We do not know how these peer
dynamics influenced a young person’s ability to talk openly about their experiences.
However, these joint interviews did provide important insights into the ways in which
peer interactions can influence how partner violence is perceived and, in some cases,
normalised.
Interviews – ethics and consent
Issues of informed consent, anonymity and the balance between confidentiality and
ensuring the safety of children and young people were central components of the
research. An agreement was drawn up between the research team and each education
authority specifying procedural requirements, including managing potential disclosures
surrounding risk of serious harm, as well as ownership and storage of data.
Participant consent
Informed consent was sought from each young person. Before starting the interview
the researcher discussed with participants what they were consenting to, including
their right not to answer a question or to stop the interview at any stage,
confidentiality and the researcher’s obligation to report a risk of serious harm,
anonymity, storage of data, and how their responses would be used in the final report
and dissemination. Once the researcher was sure the young participant understood to
what they were consenting, they were asked to sign a consent form.
Within the interview researchers used their discretion to determine if the discussion
was becoming too stressful for the young person. In these circumstances the
researcher would ask the young person if they wished to discuss a less sensitive area,
introduce a vignette or take a break, either temporarily or permanently, depending on
the young person’s wishes.
Parental consent for interview
All parents of young people who agreed to be interviewed received an additional
interview leaflet and consent form. The leaflet explained the selection procedure by
©NSPCC 2009 22
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
which their child had been asked to participate, what was involved and the limitations
of confidentiality. Parents had to sign a consent form for their child to participate in
the interview stage of the research.
Vignettes
The recognition and discussion of partner violence by young participants was
facilitated within the interviews by the introduction of vignettes. Five vignettes were
used, each depicting a different situation involving partner violence. The vignettes
provided a range of benefits for the interviews. They enabled more subtle forms of
control and manipulation to be recognised by young people as being appropriate for
discussion, as shown in the following dialogue. All names have been changed to
ensure anonymity.
Interviewer introduces a vignette depicting a girl responding to her boyfriend’s violence
Interviewer: Has that ever happened to you, being in that sort of situation?
Emma: Yeah I’ve been in exactly that situation before.
Interviewer: So what did you do?
Emma: Pushed him away, told him to stop.
Interviewer: And did he take any notice?
Emma: Yeah.
Vignettes provided young people with the opportunity to decide if, and when, to
contribute their own experiences of violence to the discussion, giving them greater
control over the research interaction (see Barter and Renold 2000, 2003). For
example, as illustrated below, young people could stop personal discussion when it
became too sensitive by using the vignette (story b) as a way of changing the focus of
the discussion to something less personal:
Rebecca: I wasn’t even, I was like seeing him when I slept with him, like I don’t
know, I don’t mind, it wasn’t like, he never pressured me…
Interviewer: “Didn’t mind” is not quite the same as wanting to though?
Rebecca: But he didn’t pressure me to, it just happened, but since I’ve been going out
with him it hasn’t happened since…
Interviewer: Uum, OK.
Rebecca: But it’s hard.
Interviewer: Yeah…?
Rebecca: Yeah because he does do it all for me; shall we read story “b”?
©NSPCC 2009 23
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
The vignettes also enabled young people to respond to sensitive issues without having
to introduce the subject themselves or having to respond to a direct question regarding
their own experience of violence. Thus, vignettes provided a mechanism by which
participants could choose when, and if, they wished to describe their own experiences
rather than discuss the vignette scenario.
Although the vignettes provided a very useful tool with young people, they were not
used within this research as an independent data collection technique. Researchers did
not set out systematically to use each vignette within every interview. Indeed, once
young people felt comfortable speaking about their own experiences, the use of
vignettes became unnecessary. Thus, we have not analysed specifically young
people’s attitudinal responses to the vignette scenarios. A body of work has already
testified to the worrying tolerance that young people show towards relationship
violence generally (see Lacasse and Mendelson 2007; Silverman et al 2006; McCarry
2003), and we did not wish to repeat this here. Much less work in the UK has looked
at young people’s direct experiences of partner violence, and it is this gap in
knowledge that the research aimed to bridge.
Analysis
All survey responses were inputted onto an SPSS database for analysis. Due to the
large sample size, survey responses were primarily analysed using the Chi-Squared
test for non-parametric data, through SPSS version 14. A forward stepwise logistic
regression analysis was undertaken on significant associated factors identified through
the Chi-Squared analysis. This enabled the prediction of which factors were most
closely associated with both girls’ and boys’ (separately) experiences of partner
violence, both as recipients and perpetrators (see appendix II for a more detailed
explanation).
Interviews were all digitally recorded with participants’ permission and fully
transcribed and anonymised. NVivo8 software was used to facilitate qualitative
analysis. A detailed coding frame representing the dominant themes and patterns was
developed from initial reading of the transcripts. This coding frame was then refined
and further built on in NVivo. The constant-comparative method (Boeije 2002) was
used to interpret the qualitative findings. Analysis was informed by a standpoint
©NSPCC 2009 24
Barter et al Partner exploitation and violence in teenage intimate relationships Introduction
analytical framework (Harding 1993), which positions the young participant as the
“expert” in understanding her or his own experiences. However, wider social
processes of which young people may not be fully aware, for example gender
inequality, also shaped the analysis, enabling the interface between structure and
agency to be highlighted.
©NSPCC 2009 25
Barter et al Partner exploitation and violence in teenage intimate relationships Section 1
Section 1: Survey sample
Gender and age of participants
The survey of 1,353 young people was split evenly by gender, with 680 female and
669 male respondents (four replies omitted gender). Nearly three-quarters of
respondents were 15 years old or older (see figure 1). This reflected our sampling
framework, which prioritised older adolescents due to their increased opportunity of
experiencing some form of relationship. Nevertheless, we also wanted to include
younger respondents to ensure their experiences were represented. Gender was
approximately evenly distributed within each of the age groups.
Figure 1 Age of young people
291
569
140
343 Female
Male
Total
Age in years 16 or over151413
600
500
400
300
200
100
0
Nationality of participants
Approaching half of respondents (n=613) were English (see figure 2). A quarter of
young people were Welsh (n=317), one in eight respondents were Scottish (n=192)
and one in 10 described themselves as British (n=131). Only a minority gave
multiple-nationality (n=27) or chose “other” (n=60) as their answer. Four young
people were Irish. Gender was roughly equally distributed within each nationality
(largest difference was 3 per cent).
©NSPCC 2009 26
Barter et al Partner exploitation and violence in teenage intimate relationships Section 1
Figure 2 Nationality of young people
Ethnicity of participants
46%
24%
14%
10%
0%4% 2%
English
Welsh
Scottish
British
Irish
Other
Multi-nationality
Most respondents (77 per cent) were white (see figure 3). Asian/Asian British (Asian)
st minority ethnic group, representing just over one
nces
ral
respondents constituted the large
in eight young people (n=175). If we look at ethnic group by school breakdown, we
find that most black (n=40) and just under half of mixed ethnicity (n=24) young
people were all located within a single school (England 1). In addition, nearly all
Asian young people came from only two schools (n=166). Thus, we need to be
cautious when using ethnicity within our analysis as most minority ethnic young
people are located in a small number of schools and, consequently, their experie
may be more geographically distinct than those of white respondents. All three ru
schools had fewer than 10 respondents from minority ethnic groups. Slightly more
girls (3 per cent, n=30) than boys were white; otherwise gender was equally
distributed within each ethnic group.
©NSPCC 2009 27
Barter et al Partner exploitation and violence in teenage intimate relationships Section 1
Figure 3 Ethnicity of young people
Just over half of respondents stated they had no religion (n=695). Of respondents who
rs (n=429) stated they were Christian, 7 per cent
Religion of participants
1%
77%
13%
4%
4%
1%
White
Asian
Black
Mixed ethnicity
Chinese
Other
reported a religion, three-quarte
(n=91) were Muslim and 5 per cent (n=74) Sikh (see figure 4). A minority of young
people were Hindu (n=15) and 13 Jewish. Six young people had more than one
religion and 23 stated “other” as a category. Gender was equally distributed within
each of the categories.
©NSPCC 2009 28
Barter et al Partner exploitation and violence in teenage intimate relationships Section 1
Figure 4 Religion of young people
52%
32%
7%
5%
1%
1% 2% 0%
No religion
Christian
Muslim
Sikh
Hindu
Jewish
Other
Multi-religion
Looking at ethnic group and religion, a clear pattern emerges. The majority of white
participants (64 per cent) had no religion and a third said they were Christian. Most
Asian young people were Sikh (47 per cent) or Muslim (38 per cent), a minority
(n=10) were Hindu. The majority of black participants (75 per cent) were Christian,
with seven stating they had no religion. Over half (56 per cent) of mixed ethnicity
young people stated they were Christian, while nearly a third (31 per cent) reported
having no religion.
Family composition
Two-thirds of young people lived with both their natural parents (see table 2). A third
of respondents lived in either a single-parent household (n=250) or a reconstituted
family (n=224). Only 32 stated they lived in some “other” setting. Proportionally,
Asian young people were most likely to be in households with two biological parents
(86 per cent), while mixed ethnicity (45 per cent) and black (47 per cent) respondents
were least likely to be living with both their parents.
©NSPCC 2009 29
Barter et al Partner exploitation and violence in teenage intimate relationships Section 1
Table 2 Family compositions and gender
Both
parents Single-parent
family Reconstituted
family Other Total
Female
(count) 437 107 119 17 680
% within sample
Male
(count) 406 143 105 15 669
% within sample
Total
(count) 843 250 224 32 1,349
% within sample 63% 18% 17% 2% 100%
Disability
Three per cent (n=42) of young people self-reported some form of disability: 18 girls
and 24 boys. However, we did not ask young people to state what form or how severe
their disability was.
©NSPCC 2009 30
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
Section 2: Young people’s experiences of family and
peer violence
This section of the report examines young people’s wider experiences of violence.
Two distinct areas were explored within the survey: family violence and peer
violence. Both of these forms of violence have been identified in US research as
constituting a significant risk factor for experiencing teenage partner violence (Cyr et
al 2006; Collin-Vézina et al 2006; O’Keefe et al 1986; Wolfe et al 2001a; Wolf and
Foshee 2003; Whitfield et al 2003; Arriaga and Foshee 2004; Lavoie et al 2002).
Family violence
We asked young people to state if any adult in their house/family had ever used abuse
or violence in three ways: against them, against another child or young person, or
against another adult. We did not specify in the survey what constituted abuse or
violence by an adult. Hence, young people were left to determine how they defined
these concepts in relation to their own family experiences. Consequently, abuse or
violence by an adult may be emotional, physical and/or sexual.
Looking first at the initial category of family violence, 13 per cent of girls (n=91) and
9 per cent of boys (n=58) stated they had experienced violence from an adult within
their house or family. Thus, girls were significantly more likely to report experiencing
child maltreatment than were boys (x2 (1) = 6.997, p<.01). In contrast, a much smaller
proportion stated an adult had used violence against another child or young person.
Only 4 per cent of girls (n=26) and 2 per cent of boys (16) reported this. The highest
incidence of family violence concerned domestic violence where an adult in the
family or household had been violent towards another adult. Overall, one-fifth of girls
(n=128) and one in 10 boys (n=68) reported they had experienced domestic violence.
Again the gender difference was significant (x2 (1) = 19.185, p<.001), with twice as
many girls compared to boys reporting domestic violence in the home. We are unable
to conclude, with any certainty, why this gender difference exists. Such violence may
possibly be more hidden from adolescent boys compared to girls, due to the
instigator’s fear that boys may attempt to physically protect the victim. Girls may also
perceive behaviours to be more harmful, and therefore define them as constituting
©NSPCC 2009 31
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
violence, than do boys. Interestingly, many more girls in the interviews spoke about
family violence than did boys. In fact very few accounts of family violence were
forthcoming from boys.
To aid analysis, all three forms of family violence were combined into a single,
mutually exclusive category of family violence. By using this combined category we
find that nearly a quarter (22 per cent) of young people reported some form of family
violence from an adult (see figure 5). However, as illustrated above, this was
unevenly distributed. Girls were significantly more likely than boys to state they had
experienced some form of family violence: 29 per cent compared to 16 per cent (x2
(1) = 31.629, p<.001).
Figure 5 Family violence by gender
196
547
0
200
400
600
800
1,000
1,200
No Yes
Have you experienced any family violence?
Female
Male
Total
301
105
479
1,026
As explained under “Research aims and methodology” on page 11, the number of
questions we could ask in the survey was limited by the need to ensure that
contributing did not have a negative impact on young respondents. Due to these
constraints we did not ask respondents to identify who was perpetrating the violence.
However, by comparing household composition and family violence we can
determine in which family structures violence has been present. A significant
association was found between household structure and family violence (x2 (3) =
47.419, p = <.001). The lowest level of family violence (18 per cent) occurred in
households where both birth parents were present. This rose to 30 per cent for single-
©NSPCC 2009 32
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
parent households and 40 per cent for reconstituted families. However, we do not
know if the violence occurred prior to the family separation or following it. Most
young people in single and reconstituted family settings stated they lived with their
biological mother. Nearly a third (31 per cent) of young people who stated they lived
in some “other” form of household experienced family violence. We also asked young
people to say how their mother and father made them feel. A caveat is required
regarding how this question was worded. The survey question asked young people to
evaluate how, for example, their mother, stepmother or foster-mother made them feel.
We therefore cannot tell which of these parental figures young people were
evaluating. Nevertheless, overall, young people who had not experienced any family
violence provided more positive accounts of their parents than did those who had
been a victim of family violence. Furthermore, young people who had experienced
violence from a family member were much more likely to say that their father made
them feel sad compared to their mother. Thus, given the above finding and taking into
account the research literature on the dynamics of domestic violence, it is plausible to
assume that the majority of the domestic violence was perpetrated by a male adult
against a female victim, who was most likely the respondent’s mother. However, it is
important to recognise that the majority of young people within each of the family
compositions had no experience of family violence.
Ethnicity and religion
When ethnicity is taken into account a similar proportion of family violence was
experienced across most groups: between 22 per cent and 24 per cent. However, a
much higher percentage of young people from mixed ethnicity backgrounds (37 per
cent, n=18) stated they had experienced violence. Although numbers are small, this
does represent an important finding that would warrant further research. No
significant association was found for religion.
Location
Incidence rates for family violence were unevenly distributed across the eight schools.
Four schools within the sample significantly deviated from the average reporting level
of 24 per cent (x2(7) = 15.691,p<.05). Two schools (E4 and S2) reported higher
incidence rates (34 per cent and 29 per cent respectively), while two schools (E3, S1)
had significantly lower rates (18 per cent and 20 per cent respectively). Given the link
©NSPCC 2009 33
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
between social deprivation and domestic violence (Hird 2000), it may be assumed that
schools with higher levels of family violence would be located in more disadvantaged
areas. However, no such association was found within our analysis. Schools at both
ends of the family violence incidence spectrum were in low and high areas of social
deprivation. In addition, whether schools were in urban or rural settings was not
associated with family violence.
Peer violence
We did not ask young people if they had ever used violence with their peers. It was
felt that respondents had already been asked a range of sensitive questions about their
own behaviour and we did not wish to overburden participants. We were, however,
interested in measuring the level of intimidation and violence present in respondents’
wider peer group relationships. It was decided that asking about their peers’
behaviour, rather than their own, would be less intrusive for young participants.
Previous US research has suggested a link exists between experiencing and
perpetrating partner violence and having friends who routinely use intimidation and
aggression against peers (Roscoe and Callahan 1985; Bookwala et al 1992; O’Keefe
and Treister 1998; Arriaga and Foshee 2004).
Two survey questions sought to ascertain the level of peer violence within young
people’s lives. The first asked if their friends used aggression or abuse against their
partners. Overall, nearly three-quarters of young people stated that their friends did
not act in this way. A minority of 7 per cent (n=99) of young people, an equal number
of boys and girls, stated that their friends did use aggression with their partners. In a
third of cases respondents did not know if their friends used aggressive behaviour
with their partners.
The second question looked at peer violence more widely and asked if their friends
used aggression or intimidation against other young people. Again, the majority of
young people stated their friends did not use aggression in this way. However,
compared to the previous question, a higher affirmative response rate was found with
16 per cent (n=209) of young people, over double the proportion before, stating that
their friends did use aggression against wider peers. Again, little difference was found
©NSPCC 2009 34
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
between male and female responses, although a much greater percentage of males (28
per cent) than females (18 per cent) reported that they were unaware if their friends
acted in this way. Taking together both these areas of aggression, it seems that most
young people do not believe their friends used aggression. Nevertheless, it is also
apparent that many respondents did not know. This implies that young people are not
necessarily aware of their friends’ wider activities, especially in more deviant areas.
Ethnicity and religion
No significant associations were found between peer violence and ethnic group or
religion, although some interesting observations are evident from data relating to
violence against wider peers and ethnicity. Most white and minority ethnic young
people reported relatively similar levels of peer intimidation. White participants
reported the lowest level (15 per cent, n=146), followed by Asian (19 per cent, n=20)
and black (20 per cent, n=10) young people. However, as with family violence, mixed
ethnicity young people had a slightly increased incidence rate of 29 per cent (n=14).
But numbers are small and we need to be cautious in making inferences from these
statistics. Similar reporting levels for peer aggression were found across all religious
categories. Only a single group, Muslim young people, reported a slightly increased
level of peer aggression (26 per cent compared to an average rate of 16 per cent).
Age
For both forms of peer violence, as age increased so did reports of friends’ use of
aggression and intimidation. This is most evident in respect to aggression and
intimidation against other young people. While one in 10 (n=14) 13-year-olds
indicated that their friends used aggression or intimidation against their peers, this
rose to nearly a quarter (22 per cent, n=64) of the 16 plus age group.
Location
Due to low reporting levels, friends’ use of partner aggression cannot be looked at in
relation to school location. However, incidence rates for wider peer violence were
unevenly distributed across the eight schools. The average reporting rate for peer
intimidation was 15 per cent; within this, individual schools ranged from 9 per cent to
24 per cent. These differences did not seem to be related to social deprivation, as
schools at both ends of the peer violence spectrum were situated in disadvantaged
©NSPCC 2009 35
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
areas. This may reflect wider levels of bullying found in the schools, which in turn
may reflect how violence and intimidation are viewed within young people’s cultures
more widely. Unfortunately, we did not collect information from schools on the
incidence of bullying and cannot therefore make any further comparisons.
Family and peer violence
When reports of family violence and peer violence are compared, it is apparent that
experiencing violence in one area is associated with an increased risk of aggression in
another. While only 6 per cent (n=62) of young people who had not experienced
family violence said that their friends used aggression with their partners, this rose to
13 per cent of those who had experienced family violence. Similarly, although 13 per
cent (n=111) of those who had not experienced family violence reported they were
aware that their friends used intimidation or aggression, this rose to nearly a third (30
per cent, n=85) of young people who had experienced family violence. This was a
statistically significant difference (x2 (2) = 47.666, p<.001).
Summary points
Family violence
Thirteen per cent of girls and 9 per cent of boys experienced violence from an
adult family member.
Twenty per cent of girls and 10 per cent of boys experienced domestic
violence.
Overall 29 per cent of girls and 16 per cent of boys experienced some form of
family violence.
Peer violence
Seven per cent of the total sample, equal number of boys and girls, stated that
their friends used aggression with their partners.
Sixteen per cent of the total sample, equal number of boys and girls, reported
that their friends used aggression with other peers.
As age increased so did reports of peer aggression.
©NSPCC 2009 36
Barter et al Partner exploitation and violence in teenage intimate relationships Section 2
Family and peer violence
Young people who experienced violence in the family were also more likely to
state that their friends used violence.
©NSPCC 2009 37
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
Section 3: Young people’s relationships
This section concentrates on findings relating to young people’s own intimate
relationships. Before the research was undertaken, researchers provided a clear
explanation regarding the type of intimate relationships and encounters the survey
questions focused on. The researchers stressed that they were interested in both long-
term and more casual intimate relationships as well as one-off encounters, such as at a
party. Thus, the survey covered all forms of intimate relationships or encounters the
young people had ever experienced.
Overall, the vast majority of young people reported at least one relationship
experience (88 per cent, n=1185). We wanted to know if young people who reported a
relationship differed in any way from the 168 who had not. A range of factors were
identified and a Chi-Squared test for non-parametric data was used to determine if any
significant differences existed between the two groups.
Gender and age
No association was found between gender and the likelihood of ever having a
relationship (89 per cent of girls compared with 86 per cent of boys). In contrast, a
significant association was found between age and relationships (x2 (3) = 14.910,
p<.005). Thus, the likelihood of being in a relationship proportionally increased as the
age of respondents increased, from 84 per cent of 13-year-olds to 91 per cent of those
aged 15. However, and rather perplexingly, the oldest age group (16 years and over)
was lower, with only 82 per cent reporting some form of a relationship – the lowest
reporting level for all age groups. This trend remained constant for both boys and
girls. It would be expected that as age increased so would the proportion of
participants who had experienced some form of intimate relationship. We were unable
to identify any associated factors that may have accounted for this disparity, such as
ethnicity or religion (see below).
Ethnicity and religion
Ethnicity was significantly associated with a young person’s probability of being in a
relationship (x2 (3) = 91.379, p<.001). The vast majority of white (92 per cent), mixed
©NSPCC 2009 38
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
ethnicity (91 per cent) and black (86 per cent) respondents reported experiencing a
relationship. In comparison only two-thirds of Asian young people reported a
relationship experience.
As would be expected given the above finding, religion was also associated with the
likelihood of experiencing a relationship. Over 90 per cent of young people with no
religion or who stated they were Christians reported a relationship. However, for all
other religions relationship experiences were less frequent. Two-thirds of Sikh (n=57)
and Hindu respondents, and 62 per cent (n=56) of Muslim young people, reported a
relationship. As illustrated in the interview data, cultural and religious beliefs
concerning the inappropriateness of relationships before marriage may explain these
differences (see section 7).
Family violence
The probability of having experienced a relationship was significantly increased by
exposure to family violence (x2(1) = 22.480, p<.001). Young people who experienced
family violence were nearly four times as likely to have had a relationship than those
who had not. Additionally, those who had a history of family violence were more
likely to have experienced a relationship at an earlier age than young people who had
not. It may be that young people who have experienced family violence seek to
compensate for these negative events through forming intimate relationships outside
their families, perhaps from an early age. Thus, relationships may be perceived as a
means to receive the support and care that may be absent from their family situations.
Or the explanation may be more pragmatic: young people may seek to remove
themselves from the family home through the formation of a relationship.
Disability
Having a disability did not alter a young person’s probability of being in a
relationship, with the majority of disabled young people (85 per cent, n=37) reporting
a relationship experience.
©NSPCC 2009 39
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
Survey results for respondents who had been in a relationship
The remaining analysis in this section is undertaken only on the 1,185 respondents
who reported some form of a relationship.
Young people’s partners
The majority of young people (96 per cent) had partners only of the opposite sex. Fifty
respondents, 21 females and 29 males, reported a same-sex partner. Most male
respondents (59 per cent) stated that their partners had either been the same age as them
or younger. For boys, age of partner was unrelated to respondent’s age. Similarly, the
majority of boys within each ethnic group, except one, reported a relationship with a
same-age partner. However, only 44 per cent (n=12) of black boys had a same-age
partner, while 30 per cent, the highest rate for any group, reported an older partner. We
also wanted to know if experiencing violence within the family was associated with a
respondent’s choice of partner. Boys who had a history of family violence were more
likely to report a younger partner (27 per cent) than were boys with no history of violence
(16 per cent), and were slightly less likely to have a same-age partner. This association
was significant (x2 (2) = 9.170, p<.05). Perhaps boys who experience family violence are
more likely to choose a partner who, due to the age differential, is less likely to challenge
their authority within the relationship and thereby reduce the possibility of conflict.
In contrast, the majority of female respondents reported having an older partner (58
per cent), with 11 per cent of partners (n=68) being “much older”. We did not ask
young people in the questionnaire to state the actual age of all their partners, so we do
not know how respondents have assessed “much older”. As female respondents’ age
increased so did their likelihood of having an older partner. Just under half (47 per
cent) of 13-year-old girls had an older partner, increasing to 66 per cent of 15-year-
olds and 60 per cent of 16-year-olds. Only a minority of each age group reported a
younger partner, ranging from 3 per cent of 13-year-olds to 7 per cent of those aged
16 years and over. For girls, ethnic group was not significantly associated with age of
partner, although Asian girls were less likely than any other groups to have an older
partner (41 per cent compared to an average of 60 per cent).
©NSPCC 2009 40
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
A significant association was found between the age of a girl’s partner and experience
of family violence (x2 (2) = 8.610, p<.05), but differently to boys. While 56 per cent
(n=222) of girls who had not experienced family violence reported having an older
partner, this proportion increased to 69 per cent (n=126) for girls who reported family
violence. We have already seen that girls who experience family violence often
entered into intimate relationships at an earlier age than those who had not. The
interview findings show that girls frequently stated that boys of their own age were
uninterested in a serious relationship and wanted only sexual contact rather than
emotional intimacy. However, older boys were considered more emotionally mature
(see section 7). This may mean that girls who are seeking emotional support, due to
negative family experiences, may perceive an older male partner as being more able
to provide this within the relationship. Nevertheless, in reality, what was occurring for
both boys and girls was the recreation of a gendered power dynamic exaggerated by
the age disparity.
Within the survey we also wanted to know how, in general, young people felt about
the relationships they had experienced. Thus, young people were asked to rate how
their partners made them feel from one to five. We then categorised their answers for
analysis into “happy”, “neutral” and “sad”. The vast majority of young people said
that their partners made them feel happy. Again this was significantly associated with
gender (x2 (2) =12.156, p<.005). Interestingly, more boys (82 per cent) than girls (76
per cent) felt that their partner made them happy. Slightly more girls than boys (18 per
cent compared to 15 per cent) gave a neutral rating and 6 per cent of girls compared to
2 per cent of boys stated that their partners made them feel sad. Looking at the age of
partner and relationship evaluations, a worrying trend became apparent for girls. As
the age difference between partners increased so did girls’ negative evaluations of
their relationships. However, no such association was found with boys and age of
their partners.
Young people with same-sex partners
Respondents were asked if their partners had been male or female. We did not ask
young people to define their sexual orientation as, for example, heterosexual, gay,
lesbian or bisexual. It was felt that young people may not necessarily identify with
these definitions and may still be exploring and questioning their sexual identity. In
©NSPCC 2009 41
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
addition, given the high prevalence of homophobia in young people’s cultures (Rivers
and Cowie 2006; Rivers and Noret 2008), this seemed the most discreet way to ask
this question, given that young people may not be “out” at school. Also, it was felt
that using the above terminology may have encouraged some young people to respond
in a homophobic manner. Nevertheless, researchers still experienced several young
men making homophobic comments to this question. However, we believe this would
have been a greater problem if more explicit language had been used.
Fifty young people, 21 females and 29 males, reported a same-sex partner. Twenty-
five boys had only male partners and four boys reported both male and female
partners. Fifteen girls had only female partners and six reported both. Thus, 40 young
people reported only having a partner of the same sex, while 10 respondents had both
male and female partners. For these 10 young people we do not know if their
responses regarding their partners’ behaviour relates to their male or female partners.
Most respondents with same-sex partners were aged over 15 years (74 per cent). A
higher proportion of respondents with same-sex partners indicated that their partners
were older (54 per cent compared with 38 per cent of young people with opposite-sex
partners). For boys this is a distinct difference to those with female partners, who
mostly had same-age or younger partners. Slightly fewer respondents with same-sex
partners stated that their partner made them feel good (72 per cent compared with 79
per cent for the general sample).
Those with same-sex partners were also more likely to have experienced some form
of family violence (x2 (1) = 8.024, p<.005). While a quarter (24 per cent) of
respondents with opposite-sex partners reported family violence, this increased to 42
per cent for same-sex partners. Similarly, although 16 per cent of young people with
opposite-sex partners commented that their friends used intimidation, the equivalent
proportion for same-sex respondents was 25 per cent, although this difference was not
statistically significant. However, it is not known if these young people’s families or
friends were aware they had a same-sex partner. It is also important to note that we do
not know if the wider violence was experienced before or after these young people
had a same-sex relationship, and possibly, as a direct reaction to their choice of a
same-sex partner.
©NSPCC 2009 42
Barter et al Partner exploitation and violence in teenage intimate relationships Section 3
Summary points
Overall 88 per cent of young people reported some form of an intimate
relationship.
A lower percentage of Asian young people had been in an intimate
relationship.
Young people with experience of family violence were more likely to have
experienced a relationship, and more likely to have experienced one at an
earlier age than young people with no history of family violence.
Boys’ partners were generally either the same age or slightly younger
than them.
Girl’s partners were generally older; a minority had a “much older” partner.
Girls with a history of family violence had an increased likelihood of having
an older partner.
For girls, as the age of their partner increased so did their negative evaluations
concerning their relationships.
Four per cent of respondents reported a same-sex partner.
Young people with a same-sex partner were more likely to have experienced
family violence than had young people with an opposite-sex partner.
©NSPCC 2009 43
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Section 4: Recipients of teenage partner violence
This section looks at young people’s experiences of physical, emotional and sexual
violence within their relationships. All analysis is undertaken only on young people
who reported a relationship experience. Thus, when viewing the findings, especially
incidence rates, it is important to remember that young people who have not had a
relationship, and therefore cannot have experienced any violence, are excluded.
Physical partner violence
Incidence rates
We used two questions in the survey to determine the incidence of physical partner
violence. We firstly asked respondents if any of their partners had ever used physical
force, such as “pushing, slapping, hitting or holding you down” (see table 3).
Disconcertingly, overall, a quarter (n=150) of girls, reported some form of physical
violence from a partner. In comparison, slightly fewer boys (18 per cent, n=100)
stated their partner had been physically violent towards them. However, girls were
also much more likely to report that the physical violence had occurred more than
once, indicating that for girls this may represent a more established pattern of
victimisation than is experienced by boys.
Table 3 Have any of your partners ever used physical force such as pushing, slapping,
hitting or holding you down?
No Once A few times Often All the time
Female 450 47 84 15 4
Male 472 36 54 5 5
Total 922 83 138 20 9
79% 7% 12% 2% 1%
Respondents were then asked if their partners had ever used any more severe physical
force such as “punching, strangling, beating you up, hitting you with an object”. It is
important to note that we do not know how young people determined if their
experiences constituted severe violence. Consequently, one young person’s evaluation
of severity may be very different to another’s.
©NSPCC 2009 44
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Fewer young people reported this level of physical violence. Overall 8 per cent (n=90)
intimated severe physical violence (see table 4). Worryingly, one in nine girls (11 per
cent) claimed some form of severe physical violence from a partner. In comparison,
only 4 per cent of boys stated they had experienced severe physical partner violence.
In addition, girls were three times as likely as boys to have experienced repeated
severe violence from their partners.
Table 4 Have any of your partners ever used any more severe physical force such as
punching, strangling, beating you up or hitting you with an object?
No Once A few times Often All the time
Female 529 38 24 3 1
Male 547 14 7 1 2
Total 1,076 52 31 4 3
92% 5% 3%
A minority of young people had experienced only severe physical violence and not
reported lower forms. Thus, for statistical analysis both categories of physical partner
violence were combined into one mutually exclusive category. In addition, the range of
answers was reduced to “had” or “had not” experienced physical partner violence.
Unsurprisingly, given the above findings, a significant association was found between
gender and being a recipient of physical violence from a partner (x2 (1) =9.381, p<.005).
The incidence rates for physical violence found within our research are comparable to
those found within US research, where between a third and a half of both female and
male teenage respondents, reported physical aggression (O’Keefe et al 1986; Foshee
1996; Williams and Martinez 1999; Halpern et al 2001; Hickman et al 2004). Our
reported incidence rate for boys is substantially lower than that found within the
majority of US studies. However, our findings reflect previous research that has
clearly demonstrated that girls generally suffer more severe forms of physical
violence (Gamache 1991; Foshee 1996; Arriaga and Foshee 2004; Jackson et al 2000).
According to Wolfe et al (2001b), this makes physical violence a common event in
the lives of adolescents, rather than an isolated problem.
©NSPCC 2009 45
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Impact of violence
Although it is important to know how frequently young people experience violence, unless
we also understand what impact this has on them we are unable to develop an appropriate
response. We therefore asked young people to select from a list of options how the violence
made them feel. We grouped responses into two categories for analysis: “negative impact”
and “no impact”. The “negative impact” responses included scared/frightened,
angry/annoyed, humiliated and upset/unhappy. No impact responses consisted of loved/
protected, no effect and thought it was funny. Young people were asked to select all that
applied, thus respondents’ answers can be present within both categories.
An overwhelming gender division emerged from young people’s replies (see figure 6).
Over three-quarters of girls (n=115) who experienced physical violence emphasised that
it had a negative impact on them. This compares to just 14 per cent (n=15) of boys. Thus,
and in direct contrast to girls, 86 per cent of boys stated the violence had no impact. All
11 per cent of girls who reported severe physical violence perceived a negative impact.
Furthermore, not only did the majority of girls report a detrimental impact, but also most
identified more than one negative impact on their wellbeing. The most commonly
reported effects were to feel scared/frightened and/or upset/unhappy.
Figure 6 Impact of physical partner violence by gender (per cent)
0
20
40
60
80
100
Female Male
86
14
24
76 No impact
Impact
©NSPCC 2009 46
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Physical violence dynamics
We asked young people to report how many relationships had contained physical
violence. In two-thirds of cases, girls and boys both expressed that the physical
violence had occurred in only one of their relationships; a third reported the violence
happened in a few. Only a minority of girls (n=5) and boys (n=8) replied that it
happened in many or all of their relationships.
We were also interested in finding out if the violence young people experienced
changed over time. We therefore asked if the behaviours stopped, stayed the same or
got worse. With regards to the behaviour stopping, we do not know if this was due to
the relationship ending or a discontinuation of the behaviour itself. Again, gender
divisions emerged in young people’s responses. The violence stopped for the majority
of boys (61 per cent), this percentage being lower for girls at only 42 per cent. For
girls, the violence was much more likely to stay the same (38 per cent) and, for one in
five, it escalated. A third of boys reported the violence remaining the same and only
six said it increased. We did not ask young people if they ended the relationship
because of the violence. If the violence stayed the same or increased, it is probable
that the young person had remained with their partner (although we recognise that the
violence may have continued after the relationship ended – see the interview findings
in section 7). We can extrapolate from these figures that the majority of girls
remained within a relationship after the violence occurred, reinforcing previous
research findings (Sugarman and Hotaling 1989; Bergman 1992; Jackson et al 2000).
This has led commentators to conclude that, although physical violence is associated
with emotional trauma and fear, in itself it is insufficient to terminate a relationship
(Henton et al 1983; Roscoe and Callahan 1985). It may be that a girl’s desire to have a
boyfriend and the social acceptance this brings outweigh their desire to leave violent
partners (Hird and Jackson 2001; Banister et al 2003). These issues are explored
further in section 7.
When incidence rates for physical partner violence are looked at in relation to impact, it
is clear that for girls physical violence from a partner represents a significant social
problem, which impacts on their wellbeing. However, boys’ evaluations concerning the
impact of physical partner violence are very different. Due to discourses surrounding
“hard” masculinities (Frosh et al 2002), it may be argued that it is especially difficult for
©NSPCC 2009 47
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
boys to acknowledge that violence, and perhaps especially violence from a female
partner, has negatively affected them. Boys may be forced to hide their vulnerability
due to their perceived need to adhere to a hegemonic masculinity where feelings,
especially vulnerability, are restricted to female, and therefore inferior, traits.
However, if this were true it would seem inconsistent to admit to being a victim of
physical violence but then to deny any impact. It would perhaps be more coherent to
deny both, especially as violence may itself imply a certain lack of control over a
girlfriend, something that again would be inconsistent with a hegemonic masculinity.
This issue is looked at in greater depth in the interview findings (see section 7),
although little evidence is present to support a hidden impact theory. In addition,
interview data in this area is supplemented by discussions from the young people’s
advisory group, which provide important insights into this debate.
Help-seeking
We asked young people who reported physical violence if they had told anyone what
had happened. A similar pattern for help-seeking emerged for all young people
irrespective of gender, age or ethnic group. Interestingly, the most frequent, indeed
only, source of advice and support accessed by young people was friends (see figure
7). More girls (54 per cent, n=85) than boys (34 per cent, n=37) sought help from a
friend. Perhaps for boys this reflected that they had not reported any negative impact
from their experience. It may also indicate a more general pattern concerning the
“boundaries” of male peer friendships, discussed in the interview data. All forms of
adult support were used very infrequently: only 8 per cent of girls and 7 per cent of
boys told a parent or carer. Similar proportions informed a sibling. Thus, the majority
of young people did not feel able to speak to their families, including parents, about
this very important issue. Only seven girls and four boys told an adult who was not a
parent or carer. As we shall see, this pattern of help-seeking is repeated for each form
of partner violence experienced by young people. Previous research also shows the
reluctance of young people to approach adults for assistance in this area (Western
Australia Crime Research Centre 2001; Watson et al 2001; Ocampo et al 2007). It is
of great concern that the vast majority of young people feel unable or are unwilling to
talk with their parents or other adults about these very important issues.
©NSPCC 2009 48
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Figure 7 Who young people told about physical partner violence (per cent)
5
53
3
3
37
7
8
8
7
78
87
73
35
5
4
43
3
6
64
4
0
10
20
30
40
50
60
70
Friends Parents Brother/
sister
Female
Male
No one Other adult
Associated factors for experiencing physical partner violence
Having identified the incidence rates for physical violence in young people’s
relationships, our analysis moved on to consider which associated factors may help us
to understand if any groups are particularly vulnerable to experiencing this form of
victimisation. It is important to remember that the presence of such factors in the lives
of young people does not necessarily mean that they will experience partner violence.
Their presence does however indicate that these young people may be more
vulnerable to violence. If and how these associated factors increase susceptibility to
violence is mediated through, among other factors, young people’s own actions and
perceptions. The interview data provides important insights into these processes (see
section 7).
Age
It may be presumed that, in general terms, as a respondent’s age increased so did their
level of relationship experiences: this would be reflected, due to a cumulative effect,
in the incidence rates for violence. However, as we shall see, this assumption does not
hold true for all forms of teenage partner violence. Thus, although age was
significantly associated with physical partner violence for boys (x2 (3) = 11.230,
p<.05), it was not for girls. Boys’ experiences of physical violence generally
©NSPCC 2009 49
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
increased as age increased, peaking at the 16 years and over age group, with 28 per
cent of this group reporting physical victimisation. In contrast, incidence rates for
girls were more evenly spread across the age range, from 21 per cent for 13-year-olds
to 26 per cent of girls aged 16 and over. Consequently, girls as young as 13 were as
likely as those aged 16 to have experienced physical violence from their partners.
Indeed, looking specifically at only severe forms of partner violence, 13-year-old girls
were slightly more likely than any other age group to report experiencing severe
violence. This finding is difficult to explain given that, if a 13-year-old girl reported
severe violence, this should also (if girls are answering in relation to all their
experiences) be included in the responses of older participants. There are a number of
possible explanations for this incongruity. Maybe older girls are failing to report
experiences that occurred a few years earlier and are restricting their responses to
more recent relationships. Another possibility is that as girls get older they redefine
previous incidents of violence as being less severe. Or it may be that this youngest age
range did experience severe violence at 13 while older participants did not.
Whatever the explanation for this discrepancy, it is of grave concern that girls this
young are experiencing violent relationships, something their younger male
equivalents appear protected from. This gender divide may reflect the fact that the
girls are more likely to enter into more serious forms of relationships at an earlier age
compared to boys, often with an older partner. Some of the associated factors that
may help to illuminate this difference are explored next.
Age of partner
Having an older partner was a significant risk factor for experiencing physical partner
violence (x2 (3) = 35.841, p<.001). A distinct gendered pattern emerges when we look
more closely at age of partners and physical violence. We have already reported that
girls were much more likely to have an older partner than were boys. However, when
we look at age of partner and girls’ exposure to physical partner violence, a worrying
association becomes evident. Using both levels of physical violence combined, 14 per
cent (n=29) of girls with a same-age partner, 23 per cent (n=64) with a slightly older
partner and 70 per cent (n=47) of girls who had a much older partner reported
physical violence. Taking just the severe level of physical violence, no girls with a
younger partner and only 6 per cent of girls with a same-age partner experienced
©NSPCC 2009 50
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
severe physical violence. This proportionally increases to 16 per cent for girls with an
older partner. Therefore, over two-thirds of girls with a much older partner
experienced physical violence, while nearly one-fifth also experienced severe
violence. As previously stated, we do not know how girls assessed their partner’s age
as being “much older”. As boys aged 16 and over generally reported their partner as
being the same age or younger, we can assume, at least in relation to the older girls in
the sample, that many of these “much older” male partners were past compulsory
school age. Thus, girls with older partners seem particularly vulnerable to physical
violence in their relationships. Moreover, girls with “much older” partners are more
likely than not to experience physical violence. In contrast, boys’ vulnerability to
physical violence seems to be the reverse. Over a third of boys with a slightly younger
partner reported being a victim of physical violence; this reduced to around 12 per
cent for both same-age and slightly older partners.
Ethnicity and religion
Although statistically ethnic group was not associated with physical partner violence,
some noteworthy distinctions did emerge in relation to ethnicity and gender. As
numbers are small, caution needs to be shown in making any generalisations. Black
girls reported a lower rate of physical victimisation (9 per cent) than any other group,
where comparative incidence levels were around 25 per cent. Conversely, incidence
rates for boys reflected a different pattern as black and Asian male respondents were
most likely to report victimisation. In contrast, although religion was not significantly
associated with physical victimisation for girls, it was for boys (x2 (3) = 13.529,
p<.005). Most girls, irrespective of religion, reported similar levels of violence. The
average reporting levels for boys was 17 per cent. Two religious groups, Muslims (39
per cent, n=12) and Sikhs (24 per cent, n=8), reported increased rates of physical
victimisation. Due to survey limitations, we did not ask participants to define their
partner’s ethnicity or religion. It is therefore not possible to fully comprehend how
these tentative patterns surrounding ethnicity and religion are experienced. However,
insights were gained through young people’s interviews, which allowed a more
detailed understanding to emerge (see section 7).
©NSPCC 2009 51
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Family and peer violence
The vast majority of young people (83 per cent) who had not experienced family
violence also reported no experience of physical partner victimisation. In comparison,
36 per cent of young people who had experienced family violence also reported
experiencing physical partner violence, which represents a significant difference (x2
(1) = 46.921, p<.001). We then wanted to investigate if severe physical partner
violence was also disproportionally experienced by those who reported family
violence. While 5 per cent of respondents who had no history of family violence
experienced severe partner violence, this increased to 16 per cent for those with a
history of family maltreatment. In addition, young people with experience of family
violence were also more likely to experience physical partner violence on a more
frequent basis than young people who had no history of violence from a family adult.
Previous research has clearly identified that family violence represents a significant
risk factor for partner violence victimisation within teenage relationships (Foshee et al
2005; Wolf and Foshee 2003; Cyr et al 2006). Although explanations concerning this
association differ within the US literature, many commentators use theories based on
the intergenerational transmission of violence: in essence arguing that children and
young people who are exposed to family violence will themselves repeat these
experiences within their own relationships. Further elaborations surround the process
of “normalisation” of violence and the acceptance of such behaviours by young
people as a normative aspect of all intimate relationships.
Although these theoretical concepts provide some level of insight into this complex
phenomenon, they cannot in isolation explain how these factors are associated (Barter
2009). We should not presume that all young people see violence as a normative
experience or that they seek to recreate their parents’ violent relationships. Research
on young people’s experiences of domestic violence clearly shows that many children
and young people do not view their experiences in this way (McGee 2000; Mullender
et al 2002; Humphreys and Stanley 2006). Similarly, in the interviews girls rarely
stated that they viewed their experiences of domestic violence as acceptable or a
normal aspect of intimate relationships. Many provided more complex explanations
regarding how family violence had impacted on their subsequent experiences of
partner violence (see section 7).
©NSPCC 2009 52
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Looking at peer violence, a significant association was also found between having
friends who use intimidation and respondents’ experiences of physical victimisation
(x2 (2) = 40.258, p<.001). While 81 per cent (n=569) of respondents whose friends
did not use intimidation also did not report physical partner violence, this decreased to
60 per cent (n=115) for young people whose peers used intimidation. In addition, the
incidence of severe violence also increased for respondents with aggressive peers,
from 6 per cent (n=43) to 17 per cent (n=32). As with family violence, respondents
whose friends used intimidation were also more likely to report that the victimisation
occurred more frequently. Thus, young people with peer groups characterised by the
use of intimidation were more susceptible to being victimised in their intimate
relationships. This may perhaps be due to teenagers frequently forming intimate
relationships with young people who they already knew or whom they met through
their wider peer networks.
Location
A significant association was found between school sites and incidence levels for
physical violence (x2 (7) = 46.802, p<.001). Three schools (E2, E3 and S1)
demonstrated a significantly lower reporting level for physical violence compared to
other schools in the sample. Incidence rates in these schools ranged from 4 per cent to
14 per cent, compared with an average rate of 21 per cent. Interestingly, both of the
schools within the most deprived areas were located within this group. So, children
from very disadvantaged areas seemed to hold no greater risk of experiencing
physical violence than those from less deprived areas. This finding is contrary to
previous UK research, which identifies social deprivation or “class” as representing a
significant risk factor for teenage physical partner violence (Hird 2000). It may be that
our gauge, based on free schools meals, is too rudimentary. We did not ask young
people directly about their parents’ social class. This was due to the complexities of
gaining an accurate reflection of social class from young people and the extensive
range of questions required (Cawson et al 2000). Thus, levels of social deprivation in
individual schools may differ substantially, and such variations would not be reflected
in our analysis. Alternatively, it may be that social disadvantage is not directly linked
to physical partner violence.
©NSPCC 2009 53
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Same-sex partner
Having a same-sex partner significantly increased the risk of experiencing physical
partner violence (x2 (1) = 14.561, p<.001). While one in five (n=229) young people
with opposite-sex partners reported some form of physical partner violence, this
proportion doubled to 44 per cent (n=21) for participants with same-sex partners. The
question exploring only the most severe forms of physical violence showed an even
greater proportional difference between the two groups. Whereas 7 per cent (n=78) of
participants with opposite-sex partners experienced severe physical violence, this
increased to nearly a quarter (23 per cent, n=11) of young people with same-sex
partners. A slightly higher percentage of young people with same-sex partners (57 per
cent) reported being negatively affected by the violence compared to young people
with heterosexual partners (48 per cent), this not being a statistically significant
difference. Although the numbers of respondents with a same-sex partner are small,
these findings do indicate an apparent heightened risk of physical victimisation within
teenage same-sex relationships. Little research addresses this area of victimisation.
However, UK work by Donovan et al (2006) on domestic violence and adults with a
same-sex partner showed that 40 per cent had experienced some form of physical
violence from their partner, and that adults under the age of 25 were most likely to
report physical partner violence.
Disability
A slightly higher percentage of disabled young people (28 per cent, n=10) than non-
disabled young people (21 per cent, n=239) reported physical violence from their
partner, although this was not statistically significant.
Summary points
Twenty-five per cent of girls and 18 per cent of boys reported some form of
physical partner violence.
Eleven per cent of girls and 4 per cent of boys reported severe physical violence.
Seventy-six per cent of girls and 14 per cent of boys stated that the physical
violence had negatively impacted on their wellbeing.
Girls were more likely to say that the physical violence was repeated and also
that it either remained at the same level or worsened.
©NSPCC 2009 54
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
The majority of young people either told a friend or told no one about the
violence; only a minority informed an adult.
Boys’ vulnerability to physical violence increased as they became older;
in contrast, girls’ exposure to violent victimisation was not associated with
their age.
Having an older partner, especially a “much older” partner, was a risk factor
for girls.
Experiences of family violence and wider peer intimidation were risk factors
associated with physical partner violence for both girls and boys.
Young people with a same-sex partner were more likely to report physical
violence than were those in heterosexual relationships.
Emotional partner violence
Emotional forms of violence are possibly the most difficult to ascertain, due to the
wide range of behaviours that may constitute victimisation. Stark (2007), based on his
research with adults, argues that what he terms “coercive control” is the most
prevalent form of domestic violence, as it underpins both physical and sexual forms of
intimate violence, but is often the most hidden form of abuse. Stark argues that this is
due to the individualised form this abuse takes, with perpetrators targeting specific
behaviour at their victims, which becomes meaningful only when placed within the
wider context of an abusive history.
Consequently, the complexity of emotional violence and the wide range of behaviours
that can constitute it, make it difficult to adequately determine in a survey. On
reviewing previous research in the area, and following advice from both our advisory
groups, eight questions were used in our survey to ascertain the prevalence of this form
of violence in young people’s relationships. Each of the eight questions was designed to
assess a particular aspect of emotional violence: from harming a young person’s self-
esteem through ridiculing them, making negative remarks and surveillance, to
controlling behaviour, including using threats of violence (see table 5).
The survey questions on emotional violence account for the highest number of missed
responses from young people (n=187). All eight questions demonstrated a very
©NSPCC 2009 55
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
similar non-response rate (on average 2 per cent): young people were not
systematically declining to respond to any specific question.
©NSPCC 2009 56
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Table 5 Young people’s experiences of emotional violence
Have any of your partners ever? Never Once A few
times Often All the
time Total
Made fun of you
Girls
Boys
315
(53%)
397
(70%)
90
(15%)
60
(11%)
159
(27%)
87
(15%)
21
(4%)
16
(3%)
12
(2%)
7
(1%)
597
(100%)
567
(100%)
Shouted at you/screamed in your
face/called you hurtful names
Girls
Boys
378
(64%)
449
(79%)
83
(14%)
54
(9.5%)
109
(18%)
54
(9.5%)
19
(3%)
7
(1%)
5
(1%)
5
(1%)
594
(100%)
569
(100%)
Said negative things about your
appearance/body/friends/family
Girls
Boys
385
(64%)
478
(85%)
87
(15%)
44
(8%)
103
(17%)
31
(5%)
19
(3%)
6
(1%)
3
(1%)
5
(1%)
597
(100%)
564
(100%)
Threatened to hurt you physically
unless you did what they wanted
Girls
Boys
525
(89%)
546
(96%)
38
(6%)
14
(3%)
22
(4%)
6
(1%)
5
(1%)
1
(-)
2
(-)
2
(-)
592
(100%)
569
(100%)
Told you who you could see and
where you could go
Girls
Boys
417
(70%)
493
(87%)
59
(10%)
30
(5%)
77
(13%)
26
(5%)
29
(5%)
7
(1%)
15
(3%)
11
(2%)
597
(100%)
567
(100%)
Constantly checked up on what you
were doing, eg by phone or texts
Girls
Boys
344
(58%)
401
(71%)
67
(11%)
45
(8%)
116
(19%)
73
(13%)
49
(8%)
30
(5%)
22
(4%)
19
(3%)
598
(100%)
568
(100%)
Used private information to make
you do something
Girls
Boys
539
(91%
537
(94%)
37
(6%)
17
(3%)
17
(3%)
12
(2%)
1
(-)
2
(-)
0
1
(-)
594
(100%)
569
(100%)
Used mobile phones or the internet
to humiliate or threaten you
Girls
Boys
528
(88%)
546
(96%)
42
(7%)
14
(2%)
28
(5%)
5
(1%)
1
(-)
1
(-)
0
3
(-)
599
(100%)
569
(100%)
Note: Percentages may not total 100% due to rounding.
©NSPCC 2009 57
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Responses to all eight questions were combined to provide an overall incidence rate
for emotional violence. This showed that three-quarters of girls (n=428) and half of
boys (n=289) experienced some kind of emotional violence from their partner, a
statistically significant gender difference (x2 (1) = 50.662, p<.001). It is important to
remember that the above figures include behaviours that occurred only once, such as
being shouted at or made fun of. Consequently, such isolated acts may not be viewed
by young people as constituting a problem. This observation is upheld within our
analysis concerning the impact of the emotional violence, which is presented later in
this section. Nevertheless, when young people’s responses were compared across the
range of questions, it became apparent that young people did not usually experience
one aspect of emotional violence in isolation. Most commonly, young people who
experienced emotional violence reported experiencing between two and three aspects.
However, girls generally reported experiencing more forms of emotional violence
than did boys. The very high incidence rate for emotional violence found within our
research reflects previous findings in this area (Bergman 1992; Jezl et al 1996; Hird
2000; Halpern et al 2001; Collin-Vézina 2006; Sears et al 2006; Sears et al 2007).
Some studies have identified rates as high as 90 per cent (Jezl et al 1996).
By examining responses to the questions individually, a gender divide became more
apparent. For each of the eight components of emotional violence, a higher proportion
of girls than boys reported victimisation. However, looking at which types of
emotional violence young people were most likely to experience, a similar pattern
emerged for both girls and boys. The most commonly experienced form of emotional
violence, irrespective of gender, was being made fun of: nearly half (47 per cent) of
girls and a third (30 per cent) of boys reported this. The second most frequently
reported behaviour was constantly being checked up on by partners: 42 per cent of
girls and 29 per cent of boys divulged this form of control. However, analysis of overt
forms of controlling behaviour, determined by asking young people if their partners
ever told them whom they could see and where they could go, produced a more
distinct gender divide. A third of girls reported experiencing this, compared to just
over one in eight boys (13 per cent). Similarly, more direct forms of emotional
violence were also more prominent for girls. Just over a third of girls (36 per cent)
reported their partners had shouted at them, screamed in their face or called them
hurtful names. A similar proportion of girls (35 per cent) also stated that their partners
©NSPCC 2009 58
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
said negative things about their appearance, body, friends or family. In comparison,
one in five boys reported being shouted at and only 15 per cent experienced negative
comments about themselves or their families. Less prevalent forms of violence
involved the use of new technologies. Overall, 12 per cent of girls and 4 per cent of
boys said that their partners had used mobile phones or the internet to humiliate and
threaten them. The low response rate to this question was surprising when viewed in
the context of our interview findings. Many girls stated that online communication
technologies were an important component of teenage relationships generally.
However, for those young people who were in a violent relationship, such
technologies provided an extra mechanism by which partners could exert control (see
section 7). It may be that the survey question was worded too imprecisely and,
therefore, did not adequately capture the subtle ways in which new technologies are
used by partners to control and manipulate.
Emotional violence also included the use of threats to intimidate partners, although
girls were more often the recipients of such tactics than were boys. One in 11 girls (9
per cent) and 6 per cent of boys stated that their partners had threatened to use
personal information to make them do things against their wishes. However, the
gender discrepancy increased when threats concerned physical violence: 11 per cent
of girls and 4 per cent of boys reported being physically threatened by a partner unless
they did what they wanted.
Thus, we can see that girls are more likely to experience emotional violence than are
boys, more frequently experience more than one form of emotional violence, and are
more likely to experience direct intimidation and control. This gender distinction
continues when we observe repeat victimisation. A third of girls stated they had been
made fun of at least a few times, and a similar proportion alleged their partners
frequently checked up on what they were doing. Around one in five girls reported that
partners often shouted at them, said negative things about them or their families,
and/or told them whom they could see or where they could go. Boys also reported
behaviours occurring repeatedly, although at a much more reduced rate. Primarily,
only two forms of emotional violence were experienced with any frequency by boys.
A fifth of boys (21 per cent) stated that their partners had frequently checked up on
their movements by phone or text, and a similar proportion (19 per cent) said they had
©NSPCC 2009 59
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
been made fun of more than once. For all other aspects of emotional violence, only a
very small minority of male respondents reported repeat victimisation. A few previous
studies have also identified girls’ increased reports of more severe forms of emotional
violence. For example, Sanders (2003) looked at behaviours deemed controlling and
isolating and found that nearly a quarter of her sample of girls experienced what she
terms “severe emotional violence”.
Impact
As discussed earlier in this section it may be that some forms of emotional violence, if
experienced in isolation, may have minimal impact on young people’s wellbeing. This
contention is partly upheld by our analysis, where 69 per cent (n=379) of girls and
nearly all boys (94 per cent, n=471), who experienced emotional violence stated that
it had no impact. If we compare this result to the impact of physical violence, and
indeed later in relation to sexual violence, this represents the only form of partner
violence where the majority of recipients, irrespective of gender, considered it had no
effect on their welfare. Young participants who stated a negative impact (n=202) were
more likely to report more forms of emotional violence, which occurred with greater
regularity, than young people who were unaffected.
Only 6 per cent of boys, compared to a third of girls, claimed that they were
negatively affected by the emotional violence they experienced. This gendered impact
disparity upholds Stark’s (2007) contention that coercive control, which many of our
components of emotional violence reflect, is made meaningful only when placed
within a gendered power understanding of intimate violence. Thus, although girls had
used emotional violence, without it being underpinned by other forms of inequality
and power, their attempts were rendered largely ineffectual.
Emotional violence dynamics
Reflecting the findings for physical violence, the majority of girls (n=251) and boys
(n=159) stated that they had experienced emotional violence in only a single
relationship. Nevertheless, roughly a third of girls (n=129) and a similar proportion of
boys (n=68) who experienced emotional violence divulged that this had occurred in a
few relationships.
©NSPCC 2009 60
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
However, a more prominent gender divide occurred with regards to how the violence
changed over time. Again this reflects the earlier findings concerning physical partner
violence. For two-thirds of boys the violence quickly stopped, but only half of the
girls reported this. Where the violence did not stop, most boys stated it remained the
same. In contrast, girls were as likely to report that the emotional violence either
remained at the same level (43 per cent) or increased (6 per cent).
Help-seeking
Compared to other forms of partner violence, a greater proportion of young people
told someone about the emotional victimisation (see figure 8). Again, peers were the
most prominent source of assistance: 57 per cent of girls (n=220) and 38 per cent of
boys (n=89) told a friend. All other forms of support were used infrequently.
Figure 8 Who young people told about emotional violence (per cent)
5
57
7
3
38
8
9
9
5
58
87
72
23
3
4
40
0
6
60
0
60
Female
50 Male
40
30
20
10
0
Associated factors for experiencing emotional partner violence
Age
Age was significantly associated with experiencing emotional violence both for girls
(x2 (3) = 18.162, p<.001) and for boys (x (3) = 19.420, p<.001). For all young people,
as age increased so did the likelihood of having experienced emotional violence.
Friends Other adult
Brother/
sister
Parents No one
©NSPCC 2009 61
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
However, some gender distinctions were present. For girls, incidence levels increased
as girls became older, from 57 per cent at 13 years old to a peak of 80 per cent at 15,
but then dropped slightly to 65 per cent for 16 years old and over. As with severe
physical violence, it is not possible to say with any certainty why this decrease
occurred, although the previous suppositions are applicable. In comparison, younger
boys (aged 13 and 14 years old) reported lower incidence rates than did younger girls
(on average 42 per cent), which increased to half of all 15 year olds. However, for the
oldest age range (16 years and over), boys reported an increased incidence rate of 68
per cent. This represents the highest rate for of any age group, irrespective of gender.
Ethnicity and religion
No association was found between experiencing emotional violence and ethnic group
or religion for boys or girls. Interestingly, only half of Muslim girls reported
emotional abuse compared to an average incidence rate of 71 per cent.
Age of partner
Age of partner was significantly associated with experiencing emotional violence for
girls (x2 (2) = 34.132, p<.001) and boys (x2 (2) = 17.064, p<.001), although for
opposite reasons. Three-quarters (n=20) of girls with a younger partner and 58 per
cent (n=114) with same-age partners reported emotional violence. Four-fifths (n=232)
of female respondents with an older partner experienced emotional violence. In
addition, girls with an older partner were more likely to experience multiple forms of
emotional abuse, and to experience them more often, than those with a younger or
same-age partner. When the above findings are linked to impact, it becomes apparent
that negative impacts for girls are more often associated with older partners.
In contrast, for boys, having a younger partner was associated with the highest risk of
experiencing emotional violence. Over three-quarters (68 per cent, n=60) of boys with
younger partners reported emotional violence, compared to 56 per cent (n=46) with an
older partner and 44 per cent (n=156) with a same-age partner. Again, when viewing
these figures, it is important to remember that for most boys no negative impact
resulted from their experiences.
©NSPCC 2009 62
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Family and peer violence
Just over half (57 per cent) of young people who had not experienced family violence
reported emotional partner victimisation. However, for young people with a history of
family violence, incidence levels increased to three-quarters (77 per cent), constituting
a significant difference (x2 (1) = 39.444, p<.001). Similarly, an association was found
for peer violence (x2 (2) = 28.148, p<.001). While 58 per cent of young people whose
friends did not use violence stated that they had experienced emotional violence, this
rose to 79 per cent of young people with aggressive peer networks.
Location
Across all eight schools the average rate of experiencing emotional violence was 65
per cent. However, individual schools within this ranged from between 51 per cent to
75 per cent. This constituted a significant association (x2 (7) = 26.441, p<.001). Three
schools (E2, E3, S1) had significantly lower incidence rates of between 51 per cent to
54 per cent. Two of these schools were in areas of very high deprivation. Interestingly
these are the same three schools that also had the lowest rates of physical partner
violence. In contrast, two schools (E4, W2) had significantly higher incidence rates
(69 per cent and 75 per cent respectively). These two schools were situated in areas of
average and low social deprivation. Thus, our findings seem to imply that social
deprivation is linked to reduced levels of emotional partner violence. The possible
caveats regarding our findings in this area have already been provided (under
“Location” on page 53).
Same-sex partner
As with physical violence, having a same-sex partner was associated with a
significantly higher risk of experiencing emotional partner violence (x2 (1) = 9.779,
p<.003). While 60 per cent of respondents with opposite-sex partners (n=674)
reported emotional abuse, this compares with 83 per cent (n=40) of participants with
same-sex partners. In addition, more young people with a same-sex partner (38 per
cent) reported being negatively affected by the violence compared to young people
with heterosexual sex partners (18 per cent), this being a statistically significant
difference (x2 (1) = 9.779, p<.005). Donovan et al (2006) also found similar rates,
with 78 per cent of their adult sample reporting some form of emotional violence from
©NSPCC 2009 63
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
a same-sex partner, with respondents under the age 25 being most likely to report
victimisation.
Disability
A slightly lower percentage of young people with a disability (57 per cent, n=20) than
non-disabled young people (62 per cent, n=693) reported emotional victimisation
from their partner, although not statistically significant. This is perhaps a slightly
surprising finding given the wide body of literature testifying to the increased
incidence of bullying and peer violence directed at disabled young people (Mishna
2003; Smith and Tippett 2007).
Summary points
Seventy-two per cent of girls and 51 per cent of boys reported some form of
emotional partner violence.
The majority of these young people reported more than one form of emotional
violence.
Most commonly reported forms of emotional violence, irrespective of gender,
were “being made fun of” and “constantly being checked up on by partner”.
More direct or overt forms of abuse were reported more frequently by girls
than boys.
Thirty-one per cent of girls and 6 per cent of boys stated that the emotional
violence had negatively impacted on their wellbeing.
Girls were more likely to say the violence occurred repeatedly and more likely
to perceive that it remained the same or got worse.
Most told either a friend or no one about the emotional violence; few told an
adult.
As age increased so did young people’s reports of emotional violence.
Having an older partner, especially a “much older” partner, was an associated
factor for girls; in contrast, boys with a younger partner were most at risk.
Experiences of family violence and wider peer intimidation were factors
associated with emotional violence, irrespective of gender.
©NSPCC 2009 64
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Young people with a same-sex partner were more likely to report emotional
violence than did those with an opposite-sex partner, and more likely to report
a negative impact.
Sexual partner violence
Four questions were used in the survey to gauge the incidence of sexual violence in
young people’s relationships. The questions were designed to reflect the role that both
“pressure” or coercive control and physical force can play in perpetrating sexual
violence. In addition, two levels of sexual violence were investigated. Firstly, a wide
definition of sexual violence was utilised, upon which basis respondents were asked if
they had ever been “pressured” or physically forced “to do something sexual, such as
kissing, touching or something else”. This was followed by a more restricted
definition, which focused on ever being pressured or forced into “having sexual
intercourse”. These two levels of sexual violence were used, as previous US research
indicates that different risk factors are associated with these two levels of violence,
and specifically that sexual violence involving physical force and intercourse has
significantly more impact than other forms (Kaestle and Halpern 2005).
Girls were significantly more likely than boys to experience sexual violence (x2 (1)
=34.026, p<.001). Combining all reported incidents, 31 per cent (n=185) of girls,
compared to 16 per cent (n=93) of boys, experienced sexual violence from their partner.
Thus, one in three girls had experienced sexual violence. Breaking this down into the
specific questions asked, in the case of both genders coercive control or “pressure” was
used much more frequently by partners than physical force (see tables 6–9).
Just over a quarter of girls (27 per cent, n=162), stated that they felt pressured into
doing something sexual against their wishes (see table 6). For the majority this had
happened only once or a few times. However, for a small minority (n=9) this had been
a regular occurrence. More than one girl in eight (13 per cent) had been physically
forced into doing something sexual, but for most this was an isolated incident (see
table 7). For 31 girls (5 per cent) the victimisation was occurring on a more systematic
basis. In relation to sexual intercourse, 16 per cent (n=93) of girls reported that they
had been pressured into intercourse (see table 8) and 6 per cent (n=35) stated they had
©NSPCC 2009 65
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
been physically forced (see table 9). Again, although for many this represented an
isolated experience, 6 per cent had regularly been pressured into sexual intercourse
and five girls reported that physical force had been reported.
Boys reported much lower levels of sexual violence, although some anomalies
existed, which we will address shortly. Overall, 15 per cent of boys said they had been
pressured to undertake something sexual against their wishes, with 6 per cent stating
physical force had been used. The same proportion of males (6 per cent) stated that
their partner had pressured them into sexual intercourse and 3 per cent reported
physical force had been used. In contrast to the girls, the majority of boys remarked
that the violence had occurred more than once. A minority of boys claimed they were
being pressured into some form of sexual act frequently. We will return to discuss
some problematic issues relating to boys’ rates of sexual victimisation later in this
section (see ”Note on boys’ experiences of sexual violence” on page 71).
Table 6 Have any of your partners ever pressured you into kissing, touching or
something else?
Never Once A few times Often All the time
Female 438 78 75 7 2
Male 484 33 35 7 9
Total 922 111 110 14 11
79% 10% 9% 1% 1%
Table 7 Have any of your partners physically forced you into kissing, touching or
something else?
Never Once A few times Often All the time
Female 516 49 27 3 1
Male 532 12 15 7 1
Total 1,048 61 42 10 2
90% 5% 4% 1%
©NSPCC 2009 66
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Table 8 Have any of your partners ever pressured you into having sexual intercourse?
Never Once A few times Often All the time
Female 505 57 34 1 1
Male 532 13 14 2 6
Total 1,037 70 48 3 7
89% 6% 4% 1%
Table 9 Have any of your partners physically forced you into having sexual intercourse?
Never Once A few times Often All the time
Female 562 30 4 0 1
Male 549 9 8 1 1
Total 1,111 39 12 1 2
95% 3% 1%
Estimates of sexual coercion and violence within previous research range from 4 per
cent to as high as 78 per cent, although most studies consistently show, as does ours,
that girls are most likely to be victims and males the perpetrators (Lane and
Gwartney-Gibbs 1985; Roscoe and Callahan 1985; Muehlenhard and Linton 1987;
Gamache 1991; Silverman et al 2001; Ackard et al 2003). This wide variation reflects
in part a definitional problem of what encompasses sexually aggressive acts. For
example, Muehlenhard and Linton (1987) found that 15 per cent of their sample had
been rape victims, although nearly 80 per cent had experienced some form of
unwanted sexual activity from their boyfriends, mostly forced kissing and touching.
Others report levels more consistent with our own findings. For example, Ackard
(2003) found that 4 per cent of adolescent girls reported being physically forced into
sexual contact against their will in a relationship or during a date. Our research sought
to overcome this methodological dilemma by including a range of survey questions on
specific aspects of sexual violence, and analysing responses both on an individual
basis and as a combined category of sexual violence.
Impact
The vast majority of girls (70 per cent) who were sexually victimised emphasised that
this had a negative impact on their welfare (see figure 9). In direct contrast, 87 per
cent of boys reported no adverse effect. In fact only 12 boys reported a negative
outcome. Thus, from these findings it seems conclusive that partner sexual violence
©NSPCC 2009 67
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
represents a problem for girls, while boys report being unaffected. As before, boys’
inability to express their vulnerability may be relevant. Within masculinity discourses,
boys are often pressured to present an image of a controlling and pervasive sexuality.
Therefore, to contradict this through an acknowledgement of sexual victimisation may
be difficult. Boys could, as discussed previously, minimise the impact of their
experiences. Much literature exists on sexual violence against women, which
identifies the increased impact of sexual violence when the perpetrator is a partner
(Easteal and McOrmond-Plummer 2006). This is primarily due to the violence also
involving an indisputable betrayal of trust. These very sensitive issues are explored in
more depth in the interview findings in section 7.
Figure 9 Impact of sexual violence (per cent)
0
20
40
60
80
100
Female Male
Impact
No impact
87
13
30
70
Sexual violence dynamics
Three-quarters (n=146) of girls stated the sexual violence occurred in only a single
relationship. However, for a quarter (n=43) the violence had happened in a few
relationships. In contrast, boys reported a higher level of sexual victimisation across
relationships. Fewer boys, nearly half (n=47), reported that the violence had occurred
in only a single relationship. A similar proportion of boys, a third (n=31), stated that it
had happened in a few relationships. However, and in contrast to girls, 20 per cent of
boys reported sexual violence in many or all of their relationships. This represents the
highest frequency rate for all forms of violence irrespective of gender. We return to
this finding later in this section.
©NSPCC 2009 68
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Moving on to look at how the sexual violence changed over time, a slightly different
picture emerges compared to physical and emotional forms of violence. For the first
time a similar proportion of girls and boys stated that the violence stopped. Previously
a much higher proportion of boys compared to girls reported that the violence ceased.
Only 44 per cent (n=43) of boys said that the violence stopped, just over half (51 per
cent, n=49) said it stayed the same and five male respondents claimed that it
worsened. Proportionally, slightly more girls (46 per cent, n=89) said the violence
stopped. Over a third (39 per cent, n=70) said it remained the same. Girls (n=28) were
more likely than boys to report an escalation in sexual violence within their
relationships.
Help-seeking
Reflecting the established pattern for help-seeking, peers were the only substantial
area of support used by young people. Nearly half of girls (48 per cent, n=92) and 44
per cent of boys (n=42) informed a friend about their sexual victimisation. Siblings
were used by 8 per cent of girls and 4 per cent of boys. Only a very small minority of
young people told an adult.
Associated factors for experiencing sexual partner violence
Age
Age was not associated with sexual violence for either gender. Incidence rates for
sexual violence were equally represented across the whole age range. This means, as
before, that younger girls and boys were as likely to report sexual violence as did
older participants. Given the impact on girls, this must be viewed as a very worrying
finding and one that has important safeguarding implications – these are discussed
later in section 8. However, more girls in the oldest age group (16 years old and over)
reported being physically forced into sexual intercourse than were any other age
range: 11 per cent compared to an average of 5 per cent. It may be that as girls get
older they are more able to resist sexual pressure and therefore some boys resort to
physical force. For boys, reporting of sexual violence peaked for two age groups: 14
and 15-year-olds.
©NSPCC 2009 69
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Age of partner
As with previous forms of victimisation, a significant association existed between age
of partner and sexual violence for girls (x2 (2) = 27.315, p<.001), although not for
boys. Three-quarters (76 per cent) of all incidents of sexual violence for girls occurred
with an older partner. No fewer than three-quarters of girls who had a “much older”
partner reported sexual violence. In contrast, boys marginally experienced greater
rates of sexual violence from younger-age partners.
Ethnicity and religion
Ethnicity and religion were not significantly associated with experiencing sexual
violence for girls. However, for boys, ethnicity was associated with sexual
victimisation (x2 (3) = 29.413, p<.001), as was religion (x2 (3) = 33.327, p<.001).
White boys reported the lowest level of sexual violence (13 per cent), followed by
black (18 per cent) and mixed ethnicity (24 per cent) young males. In comparison, an
incidence rate of 39 per cent was found for Asian male respondents. This represents
the highest incidence rate for all ethnic groups irrespective of gender. Introducing
young people’s religion provides some further insights. Compared to the average
reporting rate of 16 per cent, two religious groups reported significantly higher levels
of sexual victimisation. Nearly half of Sikh boys (47 per cent) and a third of Muslim
males (35 per cent) stated that their partner had been sexually violent towards them. A
possible explanation concerning these findings, specifically how authentic they may
be, is presented towards the conclusion of this section. We return later to the validity
of this.
Family and peer violence
As with physical and emotional partner violence, a significant association was found
between family violence and sexual partner victimisation (x2 (1) = 56.447, p<.001).
Just under a fifth (18 per cent) of young people with no family violence reported
sexual partner violence; this proportion doubled to 40 per cent for those who had
experienced some form of family violence. The inclusion of gender in the analysis
showed that, although a quarter of girls (25 per cent) who had not experienced
violence in the family experienced sexual violence, this increased to 45 per cent for
girls with a history of family maltreatment. For boys, the proportional increase was
even greater: from 13 per cent with no history to 30 per cent of boys who experienced
©NSPCC 2009 70
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
family violence. Likewise, having friendship networks characterised by aggression
was also significantly associated with an increased risk of sexual violence (x2 (2) =
16.993, p<. 001). While a fifth of young people with non-aggressive peers reported
sexual violence, this incidence rate increased to over a third (36 per cent) of young
people who reported aggressive friendship networks.
Location
Reporting of sexual violence was unevenly distributed across the eight school sites,
these differences being significant (x2 (7) = 34.547, p<.001). Looking first at girls,
most schools exhibited an incidence rate of between 14 per cent and 28 per cent.
Three schools had significantly higher incidence rates (E4, W1 and W2). In one
school nearly half of all female respondents (49 per cent) reported some form of
sexual violence. The remaining two schools had an incidence rate of 40 per cent.
Pupils from these schools were all white, the group reporting the highest level of
sexual violence. It is also important to note that these three schools constituted all of
the rural schools in our sample and included both of the Welsh schools. Two were in
areas of average deprivation and one in the least deprived sample area.
For boys, incidence rates varied from between 9 per cent and 22 per cent. One school
(E1) had a markedly higher rate for sexual violence, this being 37 per cent. This was
mostly attributed to the large percentage of Asian boys attending this school who
reported sexual victimisation. Thus, for boys, it appears that this discrepancy can be
explained by ethnicity, rather than geographical location.
Note on boys’ experiences of sexual violence
Taking all of the above findings into consideration, it appears that a very high
proportion of Asian boys in the same age group from a single school accounted for a
high proportion of all sexual violence reported by boys, especially repeated forced
sexual violence. Thus, either the school had a specific problem with teenage female
sexual perpetrators, or this group of boys did not complete the form completely
truthfully. Looking at our fieldwork notes from this school, the researchers noted a
great deal of laughing and talking between male respondents, even though researchers
and female respondents attempted to minimise this disruption.
©NSPCC 2009 71
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Same-sex partners
As with physical and emotional violence, having a same-sex partner was associated
with a significantly higher risk of experiencing sexual partner violence (x2 (1) =
7.908, p<.005). While nearly a quarter (23 per cent, n=258) of young people with
opposite-sex partners reported some form of sexual victimisation, this rose to 41 per
cent of participants with same-sex partners. However, no significant difference
between the two groups existed in relation to impact. Again, work by Donovan and
colleagues (2006) found that adults with same-sex partners also reported significant
rates of sexual violence, with respondents under the age of 25 being most at risk.
Looking at same-sex sexual victimisation more closely, some important patterns
emerge in relation to gender, although these differences are not statistically significant
due to the small number of participants with same-sex partners. For the young people
with both male and female partners, it is not possible to know which partner
committed the violence. However, what is apparent is that having a same-sex partner
only can increase vulnerability to sexual violence, even if the same-sex partner is not
perpetrating the violence.
Disability
No significant association was found between disability and sexual violence.
Summary points
Thirty-one per cent of girls and 16 per cent of boys reported some form of
sexual partner violence.
The majority were single incidents; however, for a minority of young people
sexual violence was a more regular feature of their relationships.
Seventy per cent of girls and 13 per cent of boys stated that the sexual
violence had negatively impacted on their welfare.
Some anomalies exist with regards to boys’ responses, which may call into
question the validity of some of the findings on male sexual victimisation.
Most young people informed a friend; all other avenues of disclosure were
used infrequently.
©NSPCC 2009 72
Barter et al Partner exploitation and violence in teenage intimate relationships Section 4
Respondents’ age was not associated with incidence rates for either girls or
boys. However, for girls, as age increased so did reports of being physically
forced into sexual intercourse (rape).
Seventy-eight per cent of all incidents of sexual violence against girls
occurred with an older partner and 75 per cent of all girls with a “much older”
partner experienced sexual violence.
Experiences of family violence and wider peer intimidation were associated
with increased rates of sexual violence for both girls and boys.
Young people with a same-sex partner were more likely to report sexual
violence than those with an opposite-sex partner.
©NSPCC 2009 73
Barter et al Partner exploitation and violence in teenage intimate relationships Section 5
Section 5: Instigation of teenage partner violence
Overall, young people reported significantly less involvement in instigating partner
violence compared to receiving it. This pattern held true for all three forms of partner
violence, irrespective of gender.
Physical partner violence
More girls reported using physical violence against their partner than did boys; this
represented a significant difference (x2 (1) = 60.804, p<.001). A quarter (n=148) of
girls compared to 8 per cent (n=44) of boys stated that they had used some form of
physical violence against their partner. Looking first at less severe physical violence
(see table 10), the vast majority of girls (89 per cent) reporting the use of physical
violence had used it once or a few times. Only a few (11 per cent) used it more
frequently. Similarly, the small proportion of boys who admitted using physical
violence also generally used it infrequently (83 per cent).
Table 10 Have you ever used physical force against your partner?
Never Once A few times Often All the time
Female 455 66 64 9 7
Male 526 9 27 4 3
Total 981 75 91 13 10
84% 6% 8% 1% 1%
In relation to severe physical violence, only a very small minority (3 per cent) of
young people indicated they had used this against a partner (see table 11). Again,
more girls reported using severe violence: 5 per cent (n=30) compared to 2 per cent
(n=12) of boys. Generally, the violence occurred either once or a few times; only
three young people reported using it frequently.
©NSPCC 2009 74
Barter et al Partner exploitation and violence in teenage intimate relationships Section 5
Table 11 Have you ever used more severe physical force against your partner?
No Once A few times Often All the time
Female 569 19 9 1 1
Male 556 5 6 0 1
Total 1,125 24 15 1 2
97% 2% 1%
Reasons for violence
We asked young people to say why they had behaved in this way. A range of possible
reasons were presented in the survey and young people were asked to tick all that
applied. Due to the small sample size all responses were grouped into three
categories: negative reasons, messing around and other. Negative reasons for violence
included: to hurt them, due to their behaviour, jealousy, to impress others, to get what
I wanted, anger, to humiliate them and drinking/drugs. As young people could
provide multiple responses, counts add up to more than the number of respondents. A
similar proportion of girls reported that they had acted in this way due to a negative
reason (45 per cent, n=67) or due to messing around (43 per cent, n=64). Only 18 per
cent gave “other” as a reason. In contrast, boys were much more likely to say that
their behaviour was due to messing around. Over half of boys (56 per cent, n=27)
gave this response, a third stated it was due to a negative reason (n=160) and five
stated “other”. Similarly, Foshee et al (2007) identified in their research that in many
of the young people’s accounts, and in a slightly higher proportion of boys’ accounts,
violence was often perceived as a playful and accepted aspect of relationship
behaviour. However, behaviour that is viewed as playful by one person may not
necessarily be seen the same way by their partner. Young people who view physical