2011. Vol.2, No.9, 941-947
Copyright © 2011 SciRes. DOI:10.4236/psych.2011.29142
Family Background and Environment, Psychological Distress,
and Juvenile Delinquency
Tony Cassidy
University of Ulster, Coleraine, UK.
Received August 31st, revised September 30th, 2011; accepted October 28th, 2011.
The relationship between youth offending and family background is still unclear in the literature. This study ex-
plored the role of family factors and psychological distress in relation to delinquency and youth offending to try
and explicate the relative importance of family structure, family relations, and psychological distress. The study
used the Brief Symptom Inventory, the Family Environment Scale, and the Delinquency Scale in a structured
interview format to measure psychological distress, family structure and relations, and levels of youth offending,
in 219 older children and adolescents aged between 12 - 17 years living in areas associated with high levels of
youth offending in the UK. Analysis involved correlations, hierarchical multiple regression and analysis of
variance. Family relations were the best predictors of delinquency and were also correlated with psychological
distress. The relationship between delinquency and psychological distress indicated that participants with more
psychological distress were less likely to be involved in criminal behaviour. The study supports the conclusion
that youth offending and psychological distress are both influenced by a range of factors in the family, but may
be unrelated to each other.
Keywords: Juvenile Delinquen cy, Family Background, Family Structure, Family Relations, Me nta l Hea lth
Young offenders are those law breakers who fall between a
minimum age of 10 years old and who are still under 18 years
old, and are dealt with by the juvenile justice system in most
countries. The concept of a juvenile delinquent, which includes
behaviours that would not be considered unlawful and is there-
fore broader than the concept of young offender, evolved to-
wards the end of the 19th Century to distinguish between adults
and children in the criminal justice system. It originates from a
quasi-medical model (Blackburn, 1993) and this model of the
young offender as ill and in need of intervention has had a ma-
jor impact on the treatment of young offenders. As an explana-
tion for offending behaviour however it has been controversial
and is still the subject of debate in the literature. Hirsch (1937)
in his book “Dynamic Causes of Juvenile Crime posited 4
categories of causation; heredity, environment, accident, and
genius. By genius he seems to suggest that the absence of social
intelligence is associated with crime, and by accident he meant
anything that cannot be included under heredity, environment
or genius. He concluded that delinquency is unlikely to be
caused by any single factor, and in his more detailed analysis he
identified, broken homes, family position, defective intelligence,
and enuresis as important causal factors.
Research on parental separation and delinquency is not new
(e.g. Glueck & Glueck, 1950), however findings have been
equivocal. Fergusson, Horwood and Lynkskey (1994) found
little effect of parental separation while Lindner, Stanley-Hagen
and Cavanaugh Brown (1992) found a significant increase in
delinquency following divorce. However the latter finding more
closely reflects the lay view often reflected in media coverage
which has been vociferous in linking delinquency and non-
traditional families. More recently Mack et al. (2007) found that
family type had no predictive relationship with delinquency and
the only significant effect was for maternal attachment. What
the debate often fails to do, apart from accurately reflecting the
research evidence, is to distinguish between the various forms
of non-traditional families. For example the term single parent
family is used as a generic term to describe family structures
that have derived in various ways, separation, divorce, non-
married partners who have split up, and parents who never had
a long term live in partner. In addition factors to do with rela-
tionships within the family, parent-child relationships, and par-
enting behaviour are often ignored. Factors such as increased
conflict and reduced cohesion (Haapasalo & Tremblay, 1994),
and increased pressure and demands on children (Hetherington,
Cox, & Cox, 1985), are associated with family break up. How-
ever divorce can either increase or decrease these problems
which may partly explain the equivocal findings. In a longitu-
dinal study Pagani et al. (1998) suggest that boys who experi-
ence remarriage between the ages of 11 - 15 were most at risk
for delinquency, however this was partly explained through
reduced expressiveness in the family. Smith and Farrington
(2004) in a complex study across three generations found con-
tinuity in antisocial behaviour which was partly mediated by
parenting behaviour. Parental conflict and authoritarian parent-
ing were related to antisocial behaviour in children. Farrington
(2005) in a review of research identifies a long list of psycho-
social risk factors for antisocial behaviour including poor pa-
rental relationships and parenting, and peer, neighbourhood and
school context. He proposes the Integrated Cognitive Antisocial
Potential (ICAP) theory which describes a process of person-
situation interaction largely based on a social learning model of
antisocial behaviour. There seems to be strong evidence that
antisocial behaviour and delinquency has a complex, causal eti-
ology based in social learning and involving a range of factors
in the fami ly (Stout hamer-Loe ber & Loeber, 1988; Stouthamer-
Loeber et al., 2002a, 2002b; Tremblay et al., 1992; Farrigton,
2005). Identifying the predictive correlates at a sufficient level
of specificity enables the design and implementation of preven-
tive strategies.
There is a demonstrable sex difference in delinquency with
boys more frequently involved than girls (Fitzgerald, 2003). In
addition girls who do become delinquent are less likely to
demonstrate aggressive or violent behaviour. It appears that
girls who are bullied and have lower commitment to school are
more susceptible to delinquency while the same factors do not
differentiate delinquent versus non-delinquent boys (Fitzgerald,
2003). This is important for sex based interventions.
In addition to the non-traditional family myth there is a per-
vasive belief represented in the media that delinquency is
closely linked with mental health problems. The evidence is
again equivocal. Nader, Singleton and Meltzer (2003) suggest
that the prevalence of psychotic and affective disorders among
young offenders in the UK is low (8% - 10% for males and 9%
for females) but the prevalence for neurotic disorders was high
(41% for males and 67% for women). The same study shows a
high prevalence of personality disorder but since a key element
of the definition of personality disorder is aggressive or anti-
social behaviour it seems to add little to our understanding of
the problem. In much of the literature the terms mental illness
are used reflecting a medical model perspective which is prob-
lematic for explanations of offending behaviour since the lit-
erature largely supports a social learning explanation (Farring-
ton, 2005; Smith & Farrington, 2004; Mack et al., 2007). In
order to take a stance on this the current author prefers to use
the terms psychological distress.
Assumptions about the heritability of juvenile crime are one
of the illogical conclusions drawn from the general belief about
family links to crime. It is generally accepted that early claims
about the heritability of delinquency such as Hirsch (1937)
were more a function of Nativistic bias in the theorist than any-
thing in the evidence base presented. There have been many
critiques of the limitations of behavioural genetics particularly
in regard to crime (Taylor, 2001) yet despite the social, psy-
chological and environmental confounds in their studies some
authors still choose to argue a genetic explanation (Kakar, 2005;
Tuvblad, Grann & Lichtenstein, 2006). Moffitt (2005) argues
that behavioural genetic methods are necessary to move re-
search on antisocial behaviour beyond the identification of risk
factors to the establishment of causality. Theorists adopting this
approach are will ing to a d mit that per haps 50% - 6 0% of t he va ri-
ance in antisocial behaviour may have environmental causes
based on concordance estimates of 40% - 50% for genetic shar-
ing. This assumes that the concordance estimates prove a ge-
netic influence an assumption that is by no means accepted by
the majority of researchers. In effect the introduction of a ge-
netic argument in the field seems to be a distraction from ef-
forts of identify and intervene in preventing antisocial behav-
iour. Perhaps this “genetic” component can be better explained
through social learning as demonstrated in Smith and Farring-
ton’s (2004) intergenerational study. The presence of an antiso-
cial relative as a social learning model would seem to offer a
more reasonable explanation for increased prevalence of delin-
quency than the presence of some inherited disposition (Carey,
Juvenile delinquency is a major problem in most societies
and those involved in the justice system face the ongoing di-
lemma of designing and implementing interventions in the
context of erroneous media opinion about causation and a con-
fusing evidence base. There is still a lack of clarity about the
role of the family and mental health in causing and maintaining
juvenile delinquency. The current study aims to explore the
relationship between factors of family structure, family rela-
tions, psychological distress and juvenile delinquency.
The specific predictions are:
1) Family background and family relations will be related to
both psychological distress and juvenile delinquency, but
through different pathways.
2) Psychological distress will not be significantly related to
juvenile delinquency.
The study used a quasi-experimental survey design with
structured interview and questionnaire data collection tech-
niques to investigate the relationship between family back-
ground, family relations, psychological distress and juvenile
delinquency in a sample of 219 participants living in areas with
high levels of reported crime.
These were 219 young people (147 males and 72 females)
aged between 13 - 18 years. Of these 62 were aged 13 - 14
years, 84 were 15 - 16 years, and 73 were 17 - 18 years. In
terms of home background, 94 had both parents still together,
52 had been born into a single parent situation, and 73 had ex-
perienced a family break up (69 because parents were separated
or divorced, and 4 because one parent had died). Sampling was
a mix of purposive and snowball techniques through targeting
youth centres in a target area.
A structured interview format was used to obtain information
about age, sex, parental situation (single parent, separated, di-
vorced, widowed), age at which family break up occurred, cur-
rent family structure (both natural parents, single parent, one
parent plus partner), number of brothers or sisters, birth order,
and if they had any relative who had been involved in criminal
activity. In addition three standardised questionnaires were used
to measure, 1) family environment, 2) delinquency, and 3) psy-
chological distress.
1) Family Environment Scale (Moos & Moos, 1986). This
is a 90 item scale which measures 10 first order factors of fam-
ily environment, cohesion (Alpha = .86), expressiveness (Alpha
= .82), conflict (Alpha = .87), independence (Alpha = .84),
achievement orientation (Alpha = .89), intellectual-cultural ori-
entation (Alpha = .80), active-recreational orientation (Alpha
= .83), moral-religious orientation (Alpha = .81), organisation
(Alpha = .92), and control (Alpha = .90). The scales are scored
so that a higher score indicates more experience of the specific
factor within the family. The 10 first order factors can be
grouped into 3 second order factors, 1) relationships (cohesion,
expressiveness, and conflict), 2) personal growth (independence,
achievement orientation, intellectual-cultural orientation, ac-
tive-recreational orientation, and moral-religious orientation)
and 3) systems maintenance (organisation and control).
2) The Delinquency Scale (LeBlanc & Tremblay, 1988):
This is a 27 item scale measuring 4 dimensions of delinquent
behaviour, physical aggression, stealing, vandalism, and alco-
hol and drug use. The scale can also be used as a single meas-
ure of delinquency. The scale has been used in a number of
studies (Haapasalo & Tremblay, 1994; Pagani et al., 1998) and
has shown good reliability and validity. The Cronbach Alpha
scores in this study were, physical aggression (Alpha = .86),
stealing (Alpha = .88), vandalism (Alpha = .91), alcohol/drug
use (Alpha = .89), total scale (Alpha = .87).
3) The Brief Symptom Inventory (Derogatis, 1993): This
is a 53 item scale derived from the Symptom Checklist 90
(Derogatis, 1977) which measures 10 dimensions of symptoms
of psychopathology, somatisation (Alpha = .92), obsessive-
compulsive (Alpha = .94), interpersonal sensitivity (Alpha
= .97), depression (Alpha = .89), anxiety (Alpha = .88), hostil-
ity (Alpha = .85), phobic anxiety (Alpha = .92), paranoid idea-
tion (Alpha =. 84), psychoticism (Alpha = .87), and suicide
ideation (Alpha = .86). The scale gives three summary indexes,
a) General Severity Index (GSI)—a weighted frequency score
based on the sum of ratings on each symptom, b) Positive
Symptom Total (PST)a frequency count of the number of
symptoms, and c) Positive Symptom Distress Index (PSDI)a
score reflecting the intensity of the distress. Previous studies
have used measures such as the General Health Questionnaire
(GHQ) (Goldberg, 1972) which produces an overall measure of
psychological distress based on a combination of symptoms of
depression, anxiety and somatisation. This excludes psychotic
symptoms and symptoms of hostility and interpersonal sensi-
tivity which might be more likely to be associated with crimi-
nality. In the current study the authors decided to use the BSI
because it includes a wider range of symptoms and is therefore
more comprehensive than the GHQ but still allows a composite
measure of psychological distress.
Participants were accessed via a number of youth centres
following approval from youth leaders and local councils. Ini-
tially youths attending centres were approached by youth lead-
ers and asked to participate. Snowball sampling was then used
to try and reach those who did not attend or were irregular at-
tendees. The initial target number was 500, however only 288
participants were accessed and provided with questionnaires.
Of these 219 completed questionnaires were collected, a re-
sponse rate of 76%. This is a difficult population to sample and
it is difficult to say if those who did not respond represent a
specific sub-sample. This may affect the ability to generalise
findings though the fact that approximately one third of the
sample were obtained through the use of snowball sampling
should reduce the problem. All participants were briefed on the
study requirements via an information sheet and asked for their
voluntary participation. It was explained that data would remain
anonymous and be treated in confidence and that they could
withdraw at any time. All data was entered into SPSS for
analysis. Analysis involved one way analysis of variance
(Anova) was used to test for main effects on the independent
variables. Pearson Correlations to test for relationships and
Structural Equation Modelling was used to test the fit of a Path
Anova was used to test for main effects on age, sex, family
structure, and the presence of another family member who had
been involved in crime.
There were significant age effects for physical aggression
(F(2, 216) = 11.47, p < .001), stealing (F(2, 216) = 8.54, p < .001),
vandalism (F(2, 216) = 10.43, p < .001), alcohol/drug use (F(2, 216)
= 43.16, p < .001), total delinquency (F(2, 216) = 21.54, p < .001),
family relations (F(2, 216) = 3.75, p < .05), personal growth
(F(2, 216) = 6.03, p < .01), positive symptom index (F(2, 216)
= 4.41, p < .01), overall distress (F(2, 216) = 3.09, p < .05), and
general severity index (F(2, 216) = 3.22, p < .05). The means
and standard deviations are shown in Table 1.
There were significant main effects for sex on physical ag-
gression (F(2, 216) = 30.51, p < . 001), stealin g ( F(2, 216) = 6 .59 ,
p < .01), and total delinquency (F(2, 216) = 11.12, p < .001).
In terms of family structure participants were divided into
three groups, those whose original family was still intact, those
who had always be en in a single parent family, and those whose
parents had separated or were divorced. On this variable there
were significant main effects for stealing (F(2, 216) = 6.45, p
< .01), alcohol/drug use (F(2, 216) = 3.89, p < .05), family rela-
tions (F(2, 216) = 8.78, p < .001), positive symptom index (F(2,
216) = 5.54, p < .01), overall distress (F(2, 216) = 7.82, p
< .001), and general severity index (F(2, 216) = 8.19, p < .001).
Of the total participants, 60 reported that at least one member
of their family had been involved in crime. Using this as an in-
dependent variable, significant main effects were found on physic-
cal aggression (F(2, 216) = 30.53, p < .001), stealing (F(2, 216)
= 27.37, p < .001), vandalism (F(2, 216) = 53.31, p < .001),
Table 1.
Means and standard deviations for d elinquency, family environment and ps y c ho l og i c al d is t ress across age, sex, and family structure.
13 - 14 yr
N = 62 15 - 16 yr
N = 84 17+ years
N = 73 Males
N = 147Females
N = 72 Intact family
N = 94 Single family
N = 52
N = 73
No family
N = 159
crime N = 60
Mean (Sd) Mean (Sd) Mean (Sd)Mean (Sd)Mean (Sd)Mean (Sd)Mean (Sd)Mean (Sd) Mean (Sd)Mean (Sd)
Physical a ggression 10.5 (2.9) 12.6 (3.6) 13.3 (3.7)13.1 (3.4)10.4 (3.4)11.9 (2.9)12.9 ( 4.0)12.1 (4.0) 11.4 (3.4)14.3 (3.4)
Stealing 11.6 (4.3) 13.9 (4.2) 14.3 (3.7)13.9 (3.9)12.4 (4.4)12.6 (3.9 )12.9 (3.6)14.8 (4.5) 12.5 (3.6)15.7 (4.8)
Vandalism 7.6 (2.1) 9.7 (2.9) 9.2 (3.0)9.1 (2.8)8.7 (2.9) 9.2 (2.8) 8.3 (2.9) 9.1 (2.9) 8.2 (2.3) 11.0 (3.1)
Alcohol/Drug 5.3 (2.6) 8.4 (2.3) 9.1 (2.5)7.9 (2.7)7.3 (3.1) 7.7 (2.8) 6.9 (2.9) 8.3 (2.9) 7.3 (2.9) 8.9 (2.4)
Total Delinquency 35.1 (10.2) 44.5 (10.4) 45.8 (10. 2 )44.0 (10.7)38.8 (11.4)41.5 (10.1)41.0 (10.8)44.3 (12.5) 39.4 (10.3)49.9 (9.8)
Family re lations 16.4 (5.3) 18.9 (6.9 ) 18.8 (5.7)18.2 (6.1)18.1 (6.4)19.9 (6.9)18.4 (4.4)15.4 (6.8) 22.4 (6.3)16.6 (5.3)
Systems maintenance 10.9 (4.8) 12.7 (5.4) 12.5 (5.7)12.3 (4.7)11.8 (6.5)11.9 (4.4)13. 2 (5.0)11.6 (6.6) 13.6 (4.5)11.6 (5.6)
Personal growth 14.5 (6.2) 17.4 (7.5) 18.7 (7.4)17.1 (7.4)16.9 (7.0)16.3 (5.9)18.5 (6.9)16.9 (8.9) 19.9 (7.7)15.9 (6.8)
Positive Symptom index 7.7 (2.1) 6.4 (3.2) 7.0 (2.5)6.9 (2.8)7.1 (2.7) 7.0 (2.1) 7.9 (3.0) 6.3 (2.8) 6.7 (3.1) 7.6 (2.3)
Distress index 4.6 (3.9) 3.3 (3.5) 3.3 (2.9)3.6 (3.4)3.8 (3.8) 3.2 (3.5) 5.3 (3.7) 3.0 (3.0) 3.9 (2.1) 4.3 (3.7)
General Severity Index .80 (.68) .58 (.60) .56 (.51).62 (.59).66 (. 64) .57 (.60) .92 (.65) .52 (.51) .69 (.57) .75 (.64)
alcohol/drug use (F(2, 216) = 15.81, p < .001), total delin-
quency (F(2, 216) = 46.13, p < .001), family relations (F(2, 216)
= 46.09, p < .001), systems maintenance (F(2, 216) = 6.52, p
< .01), and personal growth (F(2, 216) = 14.17, p < .001).
The relationship between delinquency, family environment
and psychological distress was explored using Pearson correla-
tions and is shown in Table 2. Total delinquency, physical ag-
gression, vandalism, stealing, and alcohol/drug use were nega-
tively correlated with the family environment dimensions of
family relations, systems maintenance and growth, and also
negatively correlated with the three overall measures of psy-
chological distress from the BSI.
The next stage in analysis was to try and identify the signifi-
cant predictors of delinquency and psychological distress using
hierarchical multiple regression analysis (HMRA). This is
shown in Table 3. All categorical variables were re-coded and
entered as dummy variables in the regression.
The significant predictors of total delinquency were the fam-
ily environment dimensions of personal growth and systems
maintenance, and the general severity of symptoms index from
the BSI which were negative indicators, having a member of
family involved in crime which was a positive indicator, and
sex and age, accounting for a total of 47% of the variance.
When the separate dimensions of delinquency were considered
separately, physical aggression was negatively predicted by
personal growth, general severity index, sex and family size,
and positively indicated by having a family member involved in
crime and age, accounting for 47% of the variance. For stealing,
personal growth, family relations, systems maintenance, and
family size were the negative predictors, and birth order, age,
family structure (intact, single parent or divorced), and having a
family member involved in crime were positive predictors,
accounting for 41% of the variance. For vandalism, having a
family member involved in crime was the only positive predic-
tor, and family relations and personal growth were negative
predictors accounting for 30% of the variance. Age and birth
order were the positive predictors of alcohol/drug use and sys-
tems maintenance and personal growth were negative predic-
tors, accounting for 50% of the variance. Psychological distress
as measured by the general severity index of the BSI was nega-
tively predicted by family relations and birth order, and posi-
tively predicted by alcohol/drug use and having a family mem-
ber involved in crime, accounting for 44% of the variance. Fi-
nally family relations was negatively predicted by vandalism,
stealing, and having a family member involved in crime, and
positively predicted by family size, accounting for 30% of the
The AMOS Structural Equation Modelling software was then
used to test a path model of the predictors of delinquency and
psychological distress and the model shown in Figure 1 was the
best fit for the data on a range of indicators. The chi-square was
non-significant, the Goodness of Fit Index (GFI) was 0.99, the
Comparative Fit Index (CFI) was 0.99, the Normed Fit Index
(NFI) was 0.99, and the root mean square error of approxima-
tion (RMSEA) was 0.03 all evidence of how closely the model
fits the data.
As with previous research this study shows that delinquency
in total and in its various forms is subject to age effects. Older
children were more likely to have had more involvement in all
types of delinquent behaviour than their younger peers. There
were also significant age trends in regard to family environment
and psychological distress. Older children reported significantly
better family relations and personal growth, and seemed sig-
nificantly better adjusted psychologically reflecting the correla-
tion between family environment and psychological distress
generally as shown in Table 3. Boys were significantly more
involved in physical aggression, stealing and overall delin-
quency than girls. In looking at family structure it proved useful
to distinguish between single parent families that had always
been so, and those that were the consequence of marital break-
down and separation or divorce. The only significant delin-
quency variables here were stealing and alcohol/drug use and it
was those who came from divorced rather than single parent
families who scored significantly higher on both. In terms of
stealing, those from single parent families scored about the
same as those from intact families, while both scored signifi-
cantly lower than those from divorced families. This would
suggest that it is the disruption of the family structure that is
important. It is also important to note that family structure does
not seem to be implicated, for this sample, in physical aggres-
sion, vandalism, or in terms of overall delinquency scores. If
overall delinquency scores had been used as the sole outcome
Table 2.
Pearson correlations b e t w e en d e l i n q uency, family environment and psyc h o logical distress.
1 2 3 4 5 6 7 8 9 10
1 Physical aggression
2 Alcohol/dru g use .57**
3 Vandalism .63** .48**
4 Stealing .56** .59** .60**
5 Total delinquency .84** .79** .80** .86**
6 Family relations –.21** .02 –.41** –.22** –.26**
7 Systems m aintenance –.25** –.48** –.15* –.32** –.36** .21**
8 Personal growth –.37** –.39** –.36** –.40** –.46** .34** .41**
9 Positive Symptom Index –.26** –.22** –.40** –.41** –.40** –.67** –.20** –.43**
10 Total dist ress –.33** –.27** –.38** –.35** –.41** –.56** –.09 –.28** .68**
11 General Se verity Index –.34** –.28** –.39** –.36** –.41** –.57** –.09 –.29** .69** .99**
*p < .05, **p < .01.
Maintenanc e
Perso nal
grow th
relativ e
Fa mil y
Syste ms
grow th
Crimina l
-.07 .34
Figure 1.
(a) Path model of predictors of delinquency from amos structural equation modelling. (chi-square (df = 2) =
3.73, p = .15; CFI = 0.99; CFI = 0.99; NFI = 0.99; RMSEA = 0.05); (b) Path model of predictors of psychologi-
cal distress and de linquency from AMOS structur al equation modellin g. (chi-square (df = 3) = 3.73, p = .29; CFI
= 0.99; GFI = 0.99; NFI = 0.99; RMSEA = 0.03).
Table 3.
The significant predictors o f d e l i n q uency and psychological distress from hie rarchical multiple regression analysis.
Variable Beta value R2 R2 Change F value Probability <Dependent varia ble
Personal growth –.22 .21 58.7 .001
Criminal relative .20 .31 .10 48.9 .001
General Severity Index –.24 .36 .05 40.5 .001
Sex –.20 .40 .04 36.1 .001
Age .21 .44 .04 33.7 .001
Systems maintenance –.18 .47 .03 31.2 .001
Total delinquency
Personal growth –.22 .21 33.3 .001
Sex –.34 .31 .10 36.6 .001
Criminal relative .22 .36 .05 34.2 .001
General Severity Index –.20 .40 .04 29.3 .001
Age .15 .44 .04 25.9 .001
Family size –.12 .47 .03 22.6 .001
Physical aggression
Family relations – .17 .17 44.0 .001
Personal growth –.20 .23 .06 31.8 .001
Birth order .36 .27 .04 26.8 .001
Age .18 .31 .04 24.3 .001
Family struc t ure .19 .35 .04 23.2 .001
Family size –.22 .38 .03 21.6 .001
Criminal relative .15 .40 .02 20.0 .01
Systems maintenance –.12 .41 .01 18.3 .05
Criminal relative .30 .20 53.3 .001
Family rela tions –.24 .28 .08 41.5 .001
Personal growth –.18 .30 .02 31.2 .01
Age .47 .24 69.5 .001
Systems maintenance –.33 .42 .18 78.9 .001
Birth order .25 .48 .06 65.6 .001
Personal growth –.16 .50 .02 52.9 .01
Alcohol/dr ug use
Family relations –.53 .32 101.6 .001
Alcohol/dr ug use .21 .39 .07 69.9 .001
Birth order –.18 .42 .03 52.6 .001
Criminal relative .11 .44 .02 41.2 .05
General Severity Index
Vandalism –.21 .20 53.2 .001
Criminal relative –.22 .26 .06 37.6 .001
Stealing –.20 .28 .02 28.3 .01
Family size .13 .30 .02 22.7 .05
Family relatio ns
measure, these effects would not have been seen. It was only
because the dimensions of delinquency were looked at sepa-
rately that it can be suggested that while they do tend to co-
exist their causes may be substantially different. This is consis-
tent with the complex models suggested by previous research
(e.g. Stouthamer-Loeber et al., 2002a, 2002b; Tremblay et al.,
1992; Farrigton, 2005).
Family structure does seem to have an effect on family rela-
tions and psychological distress in this data. Again single par-
ent families exhibited similar levels of family relations to intact
families, both significantly more positive than divorced families.
However those from single parent families did exhibit higher
levels of psychological distress than either the intact or di-
vorced families. This apparent anomaly may possibly be ex-
plained by reference to the negative correlation between psy-
chological distress and delinquency. It would seem that psy-
chological distress and delinquency have different causal roots
in familie s.
One factor that has been shown consistently in the crime lit-
erature is the experience of crime through the existence of other
criminal members within the family. In fact this is often drawn
on as evidence for a ‘genetic’ theory of criminality. In this
study having another family member involved in crime does
seem to have a strong effect. Those who reported having a
criminal relative were significantly more involved in all aspects
of delinquency. However the evidence arguably points rather
more to a social tha n a genetic e xplanat ion of c rime, si nce the se
individuals also reported significantly poorer family relations,
systems maintenance, and personal growth. However they did
not differ significantly on any of the indices of psychological
distress. The simplest explanation of this data is that other fam-
ily member’s involve ment in crime is disruptive of family re la-
tionships, reduces cohesion and increases conflict for example,
and provides a model of criminal behavio u r for c h i ld ren. This is
supported by the fact that the best predictors of family relations
are vandalism, having a relative involved i n crime, and stealing.
Hierarchical multiple regression analysis clearly demon-
strates the differences in relationships between family variables,
psychological distress and delinquency. Physical aggression
and stealing share a number of important common predictors
including personal growth and having a family member in-
volved in crime, but also have a number of other correlates
including sex, age and family size. Vandalism however seems
most strongly associated with having a family member in-
volved in crime and family relations. On the other hand having
a family member involved in crime does not seem to impact on
alcohol/drug use. Here age is the major factor with systems
maintenance which is indicative of the organisation and disci-
pline in the family. While this study shows some support for
previous findings it does suggest that a more complex process
related to family relations needs to be invoked in understanding
delinquent behaviour.
In conclusion there are a number of important points which
are indicated by the current study, despite the limitations of
cross-sectional data and self-report measures which must be
acknowledged. Firstly there is a point about methodology in
that it is important to highlight the need for research in the field
to try and separate out various aspects of the complex family
and crime variables that are the focus of study. In this research
it was beneficial 1) to distinguish between single parent and
divorced families, and 2) to consider the dimensions of delin-
quency separately. The results indicate that while delinquency
and psychological distress are both related to family environ-
ment they are not likely to be causally related to each other in
any meaningful way. Secondly it seems that it is not so much a
question of whether a family is intact or not when it comes to
delinquency, but whether the family structure has been dis-
rupted. Finally the role of pre-existing criminal involvement in
the family does seem to be important in at least two possible
ways. One might expect that it provides a model for the child in
the social learning sense. However in this study it also appears
that it is related to a dysfunctional family environment which
may have an impact on delinquent behaviour.
Blackburn, R. (1993). The psychology of criminal conduct: Theory,
research, and practice. Chichester: Wiley.
Carey, G. (1992). Twin imitation for antisocial behaviour: Implications
for genetic and family environment research. Journal of Abnormal
Psychology, 101, 18-25. doi:10.1037/0021-843X.101.1.18
Derogatis, L. R. (1977). SCL-90-R: Administration, scoring and pro-
cedures manual. Baltimore, MD: Clinical Psychometric Research
Derogatis, L. R (1993). Brief symptom inventory: Administration, scor-
ing and procedures manual. New York: National Computer Systems.
Fergusson, D. M., Horwood, L. J., & Lynkskey, M. T. (1994). Parental
separation, adolescent psychopathology, and problem behaviours.
Journal of the American Academy of Child and Adolescent Psychia-
try, 33, 1122-1131. doi:10.1097/00004583-199410000-00008
Fitzgerald, R. (2003). An examination of sex differences in delinquency.
Statistics Canada Catalogue Number 85-561-MIE—001. Ottawa.
Glueck, S., & Glueck, E. (1950). Unravelling Juvenile delinquency.
Cambridge, MA: Harvard University Press.
Haapasalo, J., & Tremblay, R. E. (1994). Physically aggressive boys
from ages 6 to 12: Family background, parenting behaviour, and pre-
diction of delinquency. Journal of Consulting and Clinical Psychol-
ogy, 62, 1044-1052. doi:10.1037/0022-006X.62.5.1044
Hetherington, E. M., Cox, M., & Cox, R. (1985). Long-term effects of
divorce and remarriage on the adjustment of children. Journal of the
American Academy of Ps yc hi at r y, 24, 518-830.
Hirsch, N. D. M. (1937). Dynamic causes of juvenile crime. Cambridge,
MA: Sci-Art Publishers.
Kakar, S. (2005). Gang membership, delinquent friends and criminal
family memebers: Determining the connections. Journal of Gang
Research, 13, 41-52.
LeBlanc, M., & Tremblay, R. E. (1988). A study of factors associated
with the stability of hidden delinquency. International Journal of
Adolescence and Youth, 1, 269-291.
Lindner, M. S., Stanley-Hagen, M., & Cavanaugh B. J. (1992). The
adjustment of children in non-divorced, divorced, single-mother, and
remarried families. Monographs of the Society for Research in Child
Development, 57, 35-72. doi:10.1111/j.1540-5834.1992.tb00302.x
Mack, K. Y., Leiber, M. J., Featherstone, R. A., & Monserud, M. A.
(2007). Reassessing the family-delinquency association: Do family
type, family processes, and economic factors make a difference?
Journal of Criminal Justic e , 35, 51-67.
Moffit, T. (2005). The new look of behavioural genetics in develop-
mental psychopathology: Gene-environment interplay in antisocial
behaviour. Psychologic a l Bulletin, 131, 533-554.
Moos, R. H., & Moos, B. S. (1986) Family environment scale manual
(2nd ed.). Palo Alto, C A : C o ns u lt i n g P s ychologists Press
Nader, D., Singleton, N., & Meltzer, H. (2003). Psychiatric morbidity
among young offenders in England and Wales. International Review
of Psychiatry, 15, 144-147. doi:10.1080/0954026021000046074
Pagani, L., Tremblay, R. E., Vitaro, F., Ke rr, M., & McDuff, P. (1998).
The impact of family transition on development of delinquency in
adolescent boys: A 9-year longitudinal study. Journal of Child Psy-
chology and Psychiatry, 39, 489-499.
Smith, C. A., & Farrington, D. P. (2004). Continuities in antisocial
behaviour and parenting across three generations. Journal of Child
Psychology and Psychiatry, 45, 230-247.
Stouthamer-Loeber, M., & Loeber, R. (1988). The use of prediction
data in understanding delinquency. Behavioural Sciences & the Law,
6, 333-354. doi:10.1002/bsl.2370060305
Stouthamer-Loeber, M., Loeber, R., Wei, E., Farrington, D., & Wik-
strom, P.-O. H. (2002a). Risk and promotive effects in the explana-
tion of persistent serious delinquency in boys. Journal of Consulting
and Clinical Psychology, 70, 111-123.
Stouthamer-Loeber, M., Wei, E. H., Homish, D. L., & Loeber, R.
(2002b). Which family and demographic factors are related to both
maltreatment and persistent juvenile delinquency? Children’s Ser-
vices: Social Policy, Research and Practice, 5, 261-272.
Taylor, K. A. (2001). On the explanatory limits of behavioural genetics.
In D. Wasserman, & R. Wachbroit (Eds.), Genetics and criminal
behaviour. New York: Cambridge University Press.
Tremblay, R. E., Maasse, B. , Perro n, D., Leblan c, M., Schwartz man, A.
E., & Ledingham, J. E. (1992). Early disruptive behaviour, poor
school achievement, delinquent behaviour, and delinquent personal-
ity: Longitudinal analyses. Journal of Consulting and Clinical Psy-
chology, 60, 64-72. doi:10.1037/0022-006X.60.1.64
Tuvblad, C., Grann, M., & Lichtenstein, P. (2006). Heritability for
adolescent antisocial behaviour differs with socioeconomic status:
Gene-environment interaction. Journal of Child Psychology and
Psychiatry, 47, 734-743. doi:10.1111/j.1469-7610.2005.01552.x