European Journal of Developmental Psychology Parenting and

Transkript

European Journal of Developmental Psychology Parenting and
This article was downloaded by: [Melhuish, Edward]
On: 17 May 2011
Access details: Access Details: [subscription number 921782607]
Publisher Psychology Press
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK
European Journal of Developmental Psychology
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713684600
Parenting and preschool child behaviour among Turkish immigrant,
migrant and non-migrant families
Melek Daglara; Edward Melhuisha; Jacqueline Barnesa
a
Birkbeck, University of London, London, UK
First published on: 29 April 2010
To cite this Article Daglar, Melek , Melhuish, Edward and Barnes, Jacqueline(2011) 'Parenting and preschool child
behaviour among Turkish immigrant, migrant and non-migrant families', European Journal of Developmental
Psychology, 8: 3, 261 — 279, First published on: 29 April 2010 (iFirst)
To link to this Article: DOI: 10.1080/17405621003710827
URL: http://dx.doi.org/10.1080/17405621003710827
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
EUROPEAN JOURNAL OF DEVELOPMENTAL PSYCHOLOGY
2011, 8 (3), 261–279
Parenting and preschool child behaviour among Turkish
immigrant, migrant and non-migrant families
Melek Daglar, Edward Melhuish and Jacqueline Barnes
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
Birkbeck, University of London, London, UK
When families migrate the new culture and culture of origin may conflict, with
possible consequences for parenting and children’s development. Turks form
one of the largest immigrant groups in Western Europe, and there is also much
movement within Turkey. This study compares three groups; Turkish
immigrants to the UK (N ¼ 142), migrants within Turkey (N ¼ 229), and
Turkish non-migrants (locals, N ¼ 396). The children were 39–71 months old
(M ¼ 58 months, SD ¼ 6.5), 392 were boys and 375 were girls. Parents supplied
data on family characteristics and parenting, and teachers supplied data on
children’s behaviour. Using Baumrind’s parenting model and allowing for
background effects, compared to non-migrants and migrants, the immigrant
parents were less permissive and more authoritarian. Children in immigrant
families had more externalizing problems, internalizing problems and
emotional dysregulation and less social competence than migrant and nonmigrant children. Multilevel models and structural equation models both
found that these effects upon child behaviour were evident after taking into
account demographic factors and were not eliminated by taking into account
parenting style differences, and thus suggest that immigration and migration
are risk factors for child behaviour. Effects of immigrant and migrant status
were partly direct and partly indirect via their effects on parenting.
Keywords: Child behaviour; Immigration; Migration; Parenting style;
Preschool children.
Children are born into families that are influenced by the surrounding
culture and parents’ cultural roots. Migration is increasingly part of life and
in many countries immigrant families are a substantial part of the
population. For example children in immigrant families account for nearly
one-in-four children in the USA (Hernandez, Denton, & Macartney, 2008).
Correspondence should be addressed to Edward Melhuish, Institute for the Study of
Children, Families & Social Issues, Birkbeck, University of London, 7 Bedford Square, London
WC1B 3RA, UK. E-mail: [email protected]
Ó 2010 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business
http://www.psypress.com/edp
DOI: 10.1080/17405621003710827
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
262
DAGLAR, MELHUISH, BARNES
When families migrate the new culture and parents’ culture of origin may
conflict, leading to stress within the family and parenting problems, which
may cause children’s psychological difficulties (Kia-Keating, 2006). We can
distinguish three groups in relation to migration: immigrant (between
countries); migrant (within country); and local (non-migrant) families. For
these groups there may well be differing consequences for parenting and
child development. This paper focuses on such possible differences.
Some studies have found that children of first-generation migrant families
are at risk of emotional and behavioural problems. Diler, Avci, and
Seydaoglu (2003), for instance, found lower self-esteem and increased
depression and anxiety among 10- to 12-year-old children of families who
had migrated within Turkey. Where migration is between countries the stress
associated with migration may well be greater. A study of London secondaryschool children (Leavey et al., 2004) found more difficulties, particularly
emotional symptoms and peer problems, among immigrant, predominantly
Kurdish, children than their UK-born counterparts. Also Mohammadi,
Taylor, and Fombonne (2006) found more psychological problems for
Iranian immigrant children, aged 6–15 years, in the UK than for similar-aged
children living in Iran, particularly for anxiety/depression and for aggressive
behaviour. In the Netherlands, Janssen et al. (2004) found that self-reported
emotional and behavioural problems were greater for Turkish immigrant
adolescents than their Dutch peers. On the other hand, while Derluyn,
Broekaert, and Schuyten (2008) report higher levels of peer problems and
avoidance for immigrant adolescents in Belgium compared to Belgian peers,
externalizing behaviour, hyperactivity and anxiety were lower for the
immigrant adolescents. Thus some (but not all) emotional and behavioural
problems for children may be greater in migrant and immigrant families.
However, different immigrant groups may have different experiences.
Stevens et al. (2003) found that Turkish and Moroccan immigrant groups in
the Netherlands with similar migration histories and from somewhat similar
cultures differed in that teachers reported higher levels of externalizing
problems for the Moroccan boys than their Dutch and Turkish counterparts, although Turkish parents reported more problems than Dutch and
Moroccan counterparts. Similarly Murad, Joung, van Lenthe, Bengi-Arslan,
and Crijnen (2003) found more self-reported behaviour problems among
Turkish immigrant adolescents than Dutch counterparts. Another Dutch
study found that immigrant parents were less likely to recognize externalizing problems in their children than Dutch parents (Zwirs, Burger, Schulpen,
& Buitelaar, 2006).
Baumrind (1971) focused on which style of discipline is optimal for child
development and described three models of parental control that
differentiate parents on the basis of their authority over their children and
their level of child-centeredness. The authoritative parenting style (demanding
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
263
but child-centred) is a constellation of parent attributes that includes
emotional support and clear, bi-directional communication, firm limit setting,
reasoning, and responsiveness. The authoritarian parenting style (demanding
and adult-centred) consists of punitive and directive discipline strategies. The
permissive parenting style (undemanding and child-centred) places few
demands or restrictions on the child.
From another perspective, Kagitcibasi (1990, 2007) has proposed three
distinct family interaction patterns: (1) traditional—with intergenerational
interdependence; (2) individual—generational independence; and (3) dialectical synthesis of (1) and (2) with material independence but emotional
interdependence between generations. Rural Turkish families typically show
the traditional pattern. While in modern urban societies, patterns (2) and (3)
with greater child independence are most common. As Kagitcibasi (2003)
pointed out, immigrant families from rural origins maintain traditional
collectivistic values of interdependence.
It is likely that migration will have an impact on both discipline and the
overall relationship between generations. Hence for a family migrating from
traditional rural environments to an urban setting, child independence may
be seen as a threat to the family, because the child will focus on individual
rather than family welfare. Yet in modern urban societies family interactions
emphasizing intergenerational independence are common where parents ‘‘let
go’’ of the child to allow individuation—separation and the development of
autonomy, which is regarded as healthy development. In such contexts
family interdependence might be regarded as dysfunctional. Similarly Kwak
(2003) has argued that immigrant parents show little change in values, and
their children adapt better to the host society than parents who are likely to
maintain the values of the culture of origin. Hence intergenerational conflict
may arise.
With regard to discipline there is probably an interaction between
immigrant family values and the demands of the new context. Possibly
immigrant parents behave more strictly to try to maintain traditional family
values, which may be maladaptive in the new context. This is congruent with
the finding from Kotchick and Forehand (2002) that African-American,
Asian-American and Latino parents are more authoritarian than EuropeanAmerican, parents. In addition, Turkish parents are traditionally more
likely to have an authoritarian style, with the father making most decisions
(Fisek & Sunar, 2005). A study of childrearing practices in three generations
of urban middle-class Turkish families showed continuities and changes
over generations (Sunar, 2002). All three generations emphasized the
importance of the family over the individual, with newer generations moving
towards more authoritative and less authoritarian parenting, with more use
of rewards and reasoning and encouragement of independence, though girls
still received more control than boys.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
264
DAGLAR, MELHUISH, BARNES
Parents may behave in a more restrictive manner to protect their children
from perceived risk (Querido, Warner, & Eyberg, 2002), when living in a
new, unfamiliar country or one where the language is different to their own.
This could have unfortunate consequences in that harsh or restrictive
parenting is likely to lead to subsequent child problems (Aunola & Nurmi,
2005; Harkness & Super, 1995; Pearson & Rao, 2003).
Also relevant is the issue of acculturation. Sociocultural difficulties in
migrant populations have been linked to cultural incongruity and less
interaction and identification with host nationals, which indicates lower
acceptance of the host culture (Bhugra & Arya, 2005; Hazuda, Stern, &
Haffner, 1988). However, adaptation to the host culture is in turn associated
with fewer child emotional and behaviour problems (Bhugra & Bhui, 1998;
Sowa, Crijnen, Bengi-Arslan, & Verhulst, 2000).
Migration effects and the relevance of acculturation are often estimated
by comparison with the local population but migration between countries
(immigration) compounds cultural, linguistic and migration status. However, it is possible that immigrant (between countries) and migrant (within a
country) groups differ on unmeasured individual differences that may be
relevant. Also, other factors may be important (e.g., language barriers,
discrimination against immigrant but not migrant families) in comparing
migration groups and locals, which means that the usual comparisons do
not fully explore the relevance of migration status versus being from another
culture.
An alternative strategy is to compare migrant and non-migrant groups in
the same culture, which identifies the impact of a new location with no
language change (and probably less discrimination); and then to compare
both these groups with similar families who have moved to a new country.
Turks form one of the largest immigrant groups in Western Europe, and
there is migration within Turkey in addition to immigration to other
countries (Crul & Vermeulen, 2003). When they move, Turkish families
often maintain strong links to Turkish culture, with traditional family
values, limited contact with host community, preference for Turkish
marriage partners, and maintenance of the Turkish language between
generations (Crul & Doomernick, 2003), all factors that limit the host
culture (Berry, 1992). In order to investigate the different risks associated
with migration and immigration, for parenting style and for child
development, this study investigated three groups; Turkish immigrants to
the UK, migrants within Turkey, and Turkish non-migrants (locals). It was
predicted that parenting would be more authoritarian, and child behaviour
problems more evident, for immigrant (between countries) compared to
migrant (within country) or local (non-migrant) families, and to a lesser
extent for migrant families compared to local families. Also relevant and
likely to support the hypothesized differences are additional contextual
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
265
factors such as discrimination against immigrant but not migrant groups,
but the relevant stressors would be least for the local group, who might be
anticipated to have the most optimal parenting. The study involved
preschool children and their families, and is the first study (to our
knowledge) of migration with families of children at this young age.
METHOD
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
Participants
Recruitment took place in preschool centres in the United Kingdom (UK)
and Turkey in areas home to recently migrated families. Participants were
767 families with 4- to 6-year-old preschool children (392 boys and 375 girls;
age range ¼ 39–71 months, M ¼ 58 months, SD ¼ 6.5). The participation
rate was 97% of all families approached. Participants in Turkey were from
two neighbouring cities—Istanbul and Kocaeli and were classified as either
‘‘local’’ or ‘‘migrant’’ according to whether they had lived in the city for less
or more than five years, since this time period would include most or all of
the target child’s life. The Turkish migrant families (N ¼ 229) had all lived in
their current city less than 5 years. The Turkish local families (N ¼ 396) had
all lived more than 5 years in their current city, and usually for much longer
(see Table 1). Participants in the UK (‘‘immigrant’’) lived in London
(N ¼ 142), the majority in neighbourhoods with substantial Turkish
populations. Unfortunately selection based on time in the UK was not
feasible and all had been resident for several years or more. All migrant and
immigrant families were first-generation migrants, and appeared to have
migrated for ‘‘a better life’’, i.e., primarily economic reasons. See Table 1 for
more details of participant characteristics.
Procedure
Parents and teachers received a letter describing the goals and methods of
the study, and signed informed consent was obtained. Mothers were
interviewed at home by a Turkish-speaking woman when questionnaires
were also completed. Sometimes fathers were also present, but the mother
was always the main respondent. Teachers at the preschool completed
questionnaires. All interview questions and questionnaires were translated
into Turkish and back translated into English by different professional
translators. A pilot study with Turkish parents ensured there were no
problems resulting from translation that could cause misunderstanding. All
interviews and questionnaires for all parents in Turkey or the UK were in
Turkish as were questionnaires for teachers in Turkey; teachers in England
completed questionnaires in English.
266
DAGLAR, MELHUISH, BARNES
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
TABLE 1
Characteristics of the study groups
Resident 1–3 years
Resident 3–5 years
Resident 5–10 years
Resident 10–15 years
Resident 16þ years
Mother rural origin
Father rural origin
Family socioeconomic status
High level
Low level
Self employed
Unemployed
Family education
University/college
High school
Elementary/secondary
Immigrant
(N ¼ 142)
N (%)
Migrant
(N ¼ 229)
N (%)
Local
(N ¼ 396)
N (%)
8
14
57
33
30
100
92
(5.6)
(9.9)
(40.2)
(23.2)
(16.1)
(70.4)
(64.7)
116 (50.7)
113 (49.3)
0
0
0
96 (41.9)
107 (46.7)
118
144
134
60
96
0
0
(29.8)
(36.4)
(33.8)
(15.2)
(24.2)
3
31
52
56
(2.1)
(21.8)
(36.6)
(39.4)
45
77
104
3
99
112
176
9
(25.0)
(28.3)
(44.4)
(2.3)
(19.7)
(33.6)
(45.4)
(1.3)
Tests and
significance
w2(2) ¼ 154.3***
w2(2) ¼ 107.8***
w2(6) ¼ 218.6***
w2(4) ¼ 161.8***
17 (12.0)
44 (13.2)
81 (57.0)
83 (36.2)
107 (46.7)
39 (17.0)
179 (45.2)
183 (46.2)
34 (8.6)
Note: ***p 5 .001.
Measures. Parent interview questions covered children’s birth details,
activities in the home, caregivers, parents’ education and occupation, family
structure, and migration history including parental place of rearing (rural or
urban). Parents completed a 32-item version of the Parenting Styles and
Dimensions Questionnaire (PSDQ; Robinson, Mandleco, Olsen, & Hart,
2001). The 32 items (answered using a 5-point scale; 1 ¼ never, 5 ¼ always)
produce 7 subscale scores that are combined to identify a predominant style,
one of three classifications (Baumrind, 1971); authoritative (Cronbach’s
alpha ¼ .86), authoritarian (Cronbach’s alpha ¼ .82) and permissive (Cronbach’s alpha ¼ .64).
Teachers completed the Child Behaviour Scale (CBS; Ladd & Profilet,
1996), and an Extended Strengths and Difficulties Questionnaire (ESDQ;
Belsky, Melhuish, Barnes, Leyland, & Romaniuk, 2006). The CBS had 59
items and 6 subscales: aggressive with peers; prosocial with peers; excluded by
peers; asocial with peers; hyperactive–distractible; and anxious–fearful (mean
Cronbach’s alpha ¼ .75). The ESDQ had 40 items describing social/emotional
behaviour to derive the 5 original Strengths and Difficulties Questionnaire
(Goodman, 1997) subscales: emotional problems; conduct problems;
hyperactivity–inattention; peer problems; and prosocial behaviour plus 3
additional subscales: independence; emotional dysregulation; and co-operation, where the extra items had been derived from questionnaires used by
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
267
Sammons et al. (2004; mean Cronbach’s alpha ¼ .73). Both CBS and ESDQ
items were scored on a 3-point scale; not true, sometimes true, certainly true.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
Statistical analysis
The 14 subscales of the ESDQ and CBS were transformed to standardized
scores (M ¼ 0; SD ¼ 1) with a common measurement scale. Principal
components analysis (with oblique rotation) was conducted to reduce the
number of dependent variables and thereby reduce the likelihood of chance
significant results. Subscales were assigned to the component on which they
had the highest positive loading and then averaged to produce a composite
factor. Five composite child behaviour factors (accounting for 67% the
variance) emerged (see Table 2). They were labelled to reflect the
contributing subscales: externalizing behaviour; hyperactivity; internalizing
behaviour; emotional dysregulation; and social competence, and constituted
the child behaviour outcomes.
Multilevel models were used to take account of the clustering of children
within preschool centres. The independent variables considered as predictors
of parenting and child behaviour were: child gender; birth weight; number of
siblings; parents’ age; highest parental educational level; socioeconomic
TABLE 2
Child behaviour constructs derived from principal components analysis
Component
Emotional problems (ESDQ)
Conduct problems (ESDQ)
Hyperactivity–inattention (ESDQ)
Peer problems (ESDQ)
Prosocial behaviour (ESDQ)
Independence (ESDQ)
Co-operation (ESDQ)
Emotional dysregulation (ESDQ)
Aggressive with peers (CBS)
Prosocial with Peers (CBS)
Asocial with peers (CBS)
Excluded by peers (CBS)
Anxious-fearful (CBS)
Hyperactive-distractible (CBS)
11
2
3
4
5
0.25
0.622
0.26
0.81
70.68
70.30
70.65
0.18
0.63
70.63
0.43
0.83
0.38
0.20
0.10
0.25
0.96
0.19
70.17
70.10
70.08
0.15
0.36
70.13
0.09
0.22
0.09
0.96
0.86
70.38
0.04
0.28
70.25
70.08
70.25
0.03
–0.02
70.19
0.20
0.29
0.50
0.01
0.11
0.31
0.15
0.47
–0.41
–0.13
70.04
0.86
0.10
–0.36
0.56
0.47
0.44
0.11
70.11
70.23
70.06
70.39
0.37
0.89
0.34
70.04
0.18
0.37
70.34
70.39
70.28
70.01
Notes: 1Component labels: 1 ¼ externalizing; 2 ¼ hyperactivity; 3 ¼ internalizing; 4 ¼ emotional
dysregulation; 5 ¼ social competence. 2Loading in italics indicates the subscale is included in
that component. EDSQ ¼ Extended Strengths and Difficulties Questionnaire; CBS ¼ Child
Behaviour Scale.
268
DAGLAR, MELHUISH, BARNES
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
status (SES; highest parental occupational status); and migration status. In
addition interaction effects for gender 6 parent education, gender 6 occupational status, and gender 6 migration status were tested. Non-significant
variables were dropped from the final models.
Logistic multilevel models were used to identify predictors for categorical
outcomes (parenting classification) and linear multilevel models for
continuous outcomes (child behaviour). There were two steps; first all
predictors were entered, then gender 6 parent education, gender 6 occupational status and gender 6 migration status interactions were added. For
child behaviour outcomes a third step added parenting classification to the
model. Finally, structural equation models were used to link all variables in
the study and test alternative interpretations of relationships between
predictors and outcomes.
RESULTS
Participant characteristics
The three groups were similar for child’s birth weight and parents’ ages (see
Table 1). The proportion of parents of rural origin differed. The majority of
the immigrant group parents and just under half the migrant group parents
but less than a quarter of the local group had been born in a rural area.
More of the immigrant group had lower educational level and lower SES
while the migrant and local groups did not differ markedly. The migrant
group had slightly lower educational and SES levels than the local group.
Parenting style
The PSDQ produces 7 subscale scores. The subscales connection, regulation
and autonomy produce an authoritative dimension score, the subscale
indulgent produces a permissive dimension score, and the subscales physical
coercion, verbal hostility and punitive produce an authoritarian dimension
score. The largest dimension score defines the predominant parenting style
as one of three categories, authoritative, authoritarian or permissive (see
introduction). Forty percent of parents were predominantly permissive, just
over a third authoritative (35%), and a smaller group predominantly
authoritarian (24%).
Group differences. The immigrant, migrant and local groups differed. In
particular only 19% of immigrant parents showed authoritative parenting,
while 28% were permissive and more than half (54%) were authoritarian. In
contrast 34% of the migrant parents were authoritative, 43% permissive and
only 23% authoritarian. For the local group 42% of parents were
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
269
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
authoritative, 44% permissive and only 14% authoritarian. Comparing the
three groups on parenting classification indicates significant differences,
w2(4, n ¼ 767) ¼ 89.1, p 5 .0001.
Predicting parenting style. The results of logistic multilevel models for
each parenting classification are shown in Table 3. The final model for
authoritative parenting indicated that this was less likely for girls (27%)
than boys (43%). Where parents had a high school or college/university
education authoritative parenting was more likely as compared to parents
with elementary/secondary education. After allowing for these variables,
migration status was not significantly related to the likelihood of
authoritative parenting. There were no significant interactions involving
gender, parent education or migration.
The final model for permissive parenting revealed that it was less likely
for parents with a college/university education (see Table 3), compared to
parents with elementary/secondary education, and those with high-school
education were not different to those with elementary/secondary education.
Allowing for parent education, permissive parenting was less frequent in the
immigrant group, but there was no significant difference between the
migrant (43%) and local (44%) groups. There were no significant
interactions involving gender, parent education or migration group.
Predominantly authoritarian parenting (see Table 3) was more likely for
girls (36%), than for boys (13%). A significant gender 6 parent education
interaction indicated that the effect of parental education differed for boys
and girls. Separate analyses for boys and girls revealed that parental
education had a significant effect for boys only. Boys were less likely to
receive authoritarian parenting when parents had a high-school education,
and particularly when parents had a college/university education. Also,
allowing for the above effects, for both boys and girls, authoritarian
parenting was more likely for the immigrant group than the local group.
In summary, authoritative parenting was more likely for boys, while
authoritarian parenting was more likely for girls. Authoritative parenting
increased as parent education increased. Permissive parenting was less likely
for parents with a college/university education, and in the immigrant group.
Authoritarian parenting increased in the immigrant group and for boys only
decreased with greater parent education.
Child behaviour
Group differences. Comparing the standardized scores for the child
behaviour constructs (see Table 4), one-way analyses of variance (ANOVAs) identified significant group differences for all five. For externalizing
behaviour, F(2, 764) ¼ 61.74, p 5 .001, all groups were significantly different
270
2
ns
ns
41.9
34.1
19.0
1.98 to 6.93
10.07 to 35.59
3.71***
18.93***
9.1
26.3
60.6
43.7
42.8
27.5
40.9
53.0
25.1
43.9
36.8
0.32 to 0.64
0.45***
43.4
26.9
Observed
(%)
Permissive
95%
confidence
interval
Authoritative
Odds
ratio
ns
0.40***
ns
0.27***
ns
Odds
ratio
0.22 to 0.72
0.19 to 0.39
95% confidence
interval
14.4
23.1
53.5
boys2 39.1
boys2 9.7
boys2 4.1
12.8
36.3
Authoritarian2
Observed
(%)
ns
3.60***
0.25***
0.11***
4.17***
Odds
ratio
2.22 to 5.86
0.12 to 0.52
0.04 to 0.21
2.81 to 6.20
95%
confidence
interval
Notes: Comparison group. For authoritarian parenting style the gender and immigration effects reflect the whole sample. The effect of parent education
on authoritarian parenting style only significant for boys, figures in italics represent data for boys only. *p 5 .05; **p 5 .01; ***p 5 .001.
1
Parenting style
Gender
Boys1
Girls
Parental education
Elementary/secondary1
High school
College/university
Migration
Local1
Migrant
Immigrant
Observed
(%)
TABLE 3
Results of binary logistic regressions to identify factors predicting parenting style classification
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
271
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
TABLE 4
Child behaviour standardized scores by group
Group
Externalizing
Mean (SD)
Hyperactivity
Mean (SD)
Internalizing
Mean (SD)
Emotional
dysregulation
Mean (SD)
Social
competence
Mean (SD)
Immigrant
Migrant
Local
Total
0.69
0.11
70.31
0.00
0.51
0.18
70.11
0.00
0.61
70.05
70.19
0.00
0.55
0.00
70.20
0.00
70.29
70.04
0.13
0.00
(0.79)
(0.96)
(0.96)
(1.00)
(1.08)
(0.92)
(0.96)
(1.00)
(0.97)
(0.86)
(1.00)
(1.00)
(1.08)
(0.97)
(0.91)
(1.00)
(0.84)
(1.00)
(1.03)
(1.00)
(post hoc Sheffe, p 5 .005), the immigrant group having the highest score,
followed by the migrant group with the local children having the least
externalizing behaviour. For hyperactivity, F(2, 764) ¼ 23.89, p 5 .001,
internalizing problems, F(2, 764) ¼ 37.144, p 5 .001, and emotional dysregulation, F(2, 764) ¼ 31.84, p 5 .001, the immigrant group had significantly
more problems than the migrant and local groups, which did not differ from
each other. For social competence, F(2, 764) ¼ 9.68, p 5 .001, the local
group had a higher score than both migrant and immigrant groups, who did
not differ from each other. Replicating the comparisons for boys and girls
separately all results were the same except that for boys the group difference
for hyperactivity was less marked, F(2, 389) ¼ 3.70, p 5 .5, and post hoc
tests only differentiated the immigrant group from the local group.
Predicting child behaviour. A summary of the effect size (ES; in standard
deviation units) for statistically significant predictors from the final
multilevel models for behaviour outcomes is shown in Table 5.
The model for externalizing behaviour indicated less externalizing
behaviour for girls (ES ¼ 0.43, p 5 .001) and less externalizing behaviour
for the children of parents with a high school (ES ¼ 0.34, p 5 .001), or
college/university education (ES ¼ 0.42, p 5 .001) compared to children of
parents with less education. There was more externalizing behaviour for the
children in migrant (ES ¼ 0.27, p 5 .001) and immigrant (ES ¼ 0.52,
p 5 .001) groups than for the local group. In terms of parenting style,
more externalizing behaviour was associated with both permissive
(ES ¼ 0.71, p 5 .001) and authoritarian (ES ¼ 0.99, p 5 .001) parenting (as
compared with authoritative).
Hyperactivity was less for girls (ES ¼ 0.44, p 5 .001) and there was a
significant gender interaction with the immigrant group. For SES
comparisons the unemployed group was the reference (comparison)
category (lowest point of scale). Children of all employed parents showed
less hyperactivity than those with unemployed parents (low SES: ES ¼ 0.41,
p 5 .01; high SES: ES ¼ 0.42, p 5 .05; self-employed: ES ¼ 0.28, p ¼ .07).
272
DAGLAR, MELHUISH, BARNES
TABLE 5
Summary of multilevel models for child behaviour showing effect sizes (in standard
deviation units) for statistically significant predictors
Externalizing
Hyperactivity
Internalizing
Emotional
dysregulation
Social
competence
74
48
53
34
70.29**
70.35**
0.14**
Variance
39
explained (%)
Child gender—comparison ‘‘boy’’
Girl
70.43***
70.44***
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
Parent education—comparison ‘‘elementary’’
High School
70.34***
College
70.42***
SES—comparison ‘‘unemployed’’
Low
Self-employed
High
Migration—comparison ‘‘local’’
Immigrant
0.52***
Migrant
0.27***
70.41**
70.280
70.42*
girls1 0.68***
Parenting—comparison ‘‘authoritative’’
Permissive
0.71***
Authoritarian
0.99***
0.35***
0.37***
70.19*
70.190
0.17*
0.72***
0.40***
70.66***
70.27***
0.28***
0.51***
0.29**
0.44***
70.69***
71.03***
Notes: 1For hyperactivity the effect of immigrant compared to local group was significant only
for girls. 0p 5 .10; *p 5 .05; **p 5 .01; ***p 5 .001.
Separate analyses for boys and girls revealed that, for girls only, there was
greater hyperactivity among the immigrant group compared to the local
group (ES ¼ 0.68, p 5 .001). When parenting was added to the model for all
children, authoritarian (but not permissive) parenting was associated with
increased hyperactivity as compared with authoritative parenting
(ES ¼ 0.35, p 5 .001).
There was more internalizing behaviour for girls (ES ¼ 0.37, p 5 .001).
Also children of self-employed (ES ¼ 0.17, p 5 .05) parents were reported by
teachers to have more internalizing behaviour. The immigrant group
showed substantially more internalizing behaviour (ES ¼ 0.72, p 5 .0001) in
contrast to children in the migrant group, who were equivalent to the local
group. Both permissive (ES ¼ 0.28, p 5 .0001) and particularly authoritarian parenting (ES ¼ 0.51, p 5 .0001) were associated with more internalizing
behaviour (compared to authoritative parenting).
Children whose parents had a high-school education (ES ¼ 0.29, p 5 .01)
or better (ES ¼ 0.35, p 5 .01) were reported to show less emotional
dysregulation. Immigrant children were said to have more emotional
dysregulation (ES ¼ 0.40, p 5 .001). Emotional dysregulation was also
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
273
greater for children whose parents were either permissive (ES ¼ 0.29,
p 5 .01) or authoritarian (ES ¼ 0.44, p 5 .001) compared to children of
authoritative parents.
Initial effects for parents’ education on social competence became nonsignificant when parenting style was added. In the final model children in
both immigrant (ES ¼ 0.66, p 5 .001) and migrant (ES ¼ 0.27, p 5 .001)
groups showed substantially less social competence than the local group.
There was significantly less social competence shown by children in the
permissive (ES ¼ 0.69, p 5 .001) and authoritarian (ES ¼ 1.03, p 5 .001)
parenting groups compared to the authoritative group.
For all five child behaviour variables the effects of migration status were
not significantly reduced when parenting classification was added to the
model, thus indicating that the migration had effects upon child behaviour
that were independent of parenting effects.
Structural equation models were used to consider simultaneously the
relationships between all significant predictor variables and all parenting
and child behaviour variables in the study. Three types of model were
compared:
1. No mediation: All predictors have independent effects upon child
behaviour.
2. Full mediation: Parent and child characteristics have effects upon
parenting, which then affects child behaviour, but no independent
effects upon child behaviour.
3. Partial mediation: Parent and child characteristics have independent
effects upon child behaviour as well as effects mediated through
parenting.
In the models, the PSDQ subscale scores determining parenting classification were indicators of a latent variable of parenting, and the five child
behaviour scores were indicators of a latent variable of child social
adjustment. Parental education was treated as an ordinal variable, and
dummy variables were created for immigrant (yes/no), migrant (yes/no)
groups and for gender (boy: yes/no). The significant predictors from the
multilevel models for child behaviours were included in the models and
any paths not adding significantly to a reduction in chi-square were
dropped. The best-fitting model was a partial mediation model, which
showed a significant reduction in chi-square over both the no-mediation
and full-mediation models (Dw2 ¼ 57 to 537, df ¼ 3 to 39, p 5 .0001). Also
these models were consistent with the multilevel model results and all
showed a good fit to the data (CFI 4 .95; RMSEA 5 .05, C1 0.03 to 0.06;
Hoelter index 4 200). The partial mediation model is shown in Figure 1.
Parent education only had an indirect effect upon the latent child social
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
274
DAGLAR, MELHUISH, BARNES
Figure 1. Structural equation model. This shows relationships between PSDQ subscales and
latent parenting variable, child behaviour scores and latent child social adjustment variable, and
between parent education and immigration/migration status and latent variables.
adjustment variable through a direct effect on the latent parenting variable,
which in turn influenced the latent child social adjustment variable.
Gender had an indirect effect on child adjustment through its influence on
the latent parenting variable and also a direct effect on child social
adjustment. The immigrant and migrant groups had indirect effects on
child social adjustment through their effects on the latent parenting
variable and also direct effects on child social adjustment. In addition there
were further indirect effects via parent education. In all direct and indirect
paths the effects were stronger for the immigrant rather than migrant
group.
DISCUSSION
Results reported here indicate that immigrant parents had a more
authoritarian parenting style than migrant or local parents, and that
children in immigrant families had more externalizing problems, internalizing problems and emotional dysregulation and less social competence than
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
275
migrant and non-migrant children. Structural equation modelling indicated
that immigrant and migrant status had both direct effects on child social
adjustment and also indirect effects than involved paths via parent education
and parenting. These effects suggest that immigration and migration are
potential risk factors for child behaviour.
The first hypothesis was confirmed in that immigrant parents were found
to be more authoritarian. In addition, this study provides strong support for
the usefulness, with Turkish parents, of Baumrind’s model of parenting,
which was originally developed in the USA and proposes that authoritative
parenting is associated with optimal development from the preschool years
(Baumrind & Black, 1967) through to adolescence (Steinberg, Elman, &
Mounts, 1989). In this sample of Turkish parents both permissive parenting
and, more particularly, authoritarian parenting were associated with a
greater likelihood of adverse child behaviours and a lesser likelihood of child
behaviour that may be advantageous (i.e., social competence). Beyond the
effects identified using this model of parenting, such as more authoritarian
parenting directed to girls than boys, and more from parents with fewer
educational qualifications, there was also, as predicted, an effect of
migration status on parenting, being most marked when the family was
immigrant. The three groups of parents were almost identical in age
(mothers’ Mage ¼ 32–33 years; fathers’ Mage ¼ 36), so migration differences
did not reflect age effects. Also, these results were found with preschool
children, which have not been studied in previous research on immigration
and migration.
For both migrant and immigrant groups the higher likelihood of their
using an authoritarian style of parenting may reflect anxiety about the local
area, as many of the parents in both groups had been raised in rural areas
that could seem more family friendly than urban locations. Beyond the
possible influence of the stress of moving to a culture differing from their
culture of origin, the immigrant group, for whom the effect was stronger,
had the additional stress of living in a culture where the language was
different and where cultural norms may be potentially challenging and have
characteristics that they wished to protect their children from (e.g., possible
discrimination). This is akin to the tendency of parents living in inner city
environments to use more authoritarian strategies, as a means of protecting
children from community danger or possibly harmful influences (Querido
et al., 2002).
The second hypothesis was also confirmed in that immigrant status was
associated with more child behaviour problems, with a lesser but also
significant effect of migration within Turkey. Taking into account child
gender and family demographic characteristics, and parenting style, both
groups were more likely to exhibit behaviour problems and relate less well to
peers than children living in Turkey in families that were local to the area.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
276
DAGLAR, MELHUISH, BARNES
Specifically, immigrant children were likely to have more externalizing
problems, internalizing problems and emotional dysregulation and less
social competence with peers; and girls were likely to have more
hyperactivity than the migrant or local groups. The findings from the
multilevel models were supported by the structural equation models that
indicated that immigration/migration status affects child social adjustment
directly in addition to its effects upon parenting. This suggests that, beyond
the tendency for immigrant parents to be more authoritarian, there may be
additional stresses within the family and for the children living in a
predominantly non-Turkish culture as they mix with peers that are risk
factors for behavioural and emotional problems. This is congruent with
evidence from Shrake and Rhee (2004) that cultural conflict can be
associated with increased behaviour problems in Korean-American adolescents, and older immigrant children and adolescents in the UK (Leavey
et al., 2004; Mohammadi et al., 2006).
Children in the migrant group in Turkey were likely to have more
externalizing problems and less social competence than children in families
local to the area but it did not appear that they were more susceptible to
internalizing or emotional problems. Possibly their parents’ moves led to
some disruption and stress but the situation of living within a different
culture, with the attendant language differences and cultural expectations
posed a greater risk factor for children’s behavioural development (Berry,
1992). While authoritarian parenting was associated with more behaviour
problems and less social competence, the effect of immigration was evident
in addition to the parenting style effect.
With regard to relationship identified between immigrant status and
internalizing problems or emotional dysregulation, it is noteworthy that
these problems are often less evident in the early years and only come to the
attention of adults, teachers or parents when children display overt distress
such as marked depression. There is potential to offer support for immigrant
parents and their young children that include not only a focus on avoiding
harsh discipline but also supporting the children to express their feelings
about adaptation to the host culture. Sowa et al. (2000) found that problem
behaviours in Turkish immigrant children in Holland were associated with
low integration and that more integration leads to lower levels of problem
behaviour. Generally, there is better functioning in children of immigrant
parents who are successfully adapted in terms of psychological and
sociocultural outcomes, socioeconomical and professional social status,
who speak the language of the host culture well, and have supportive friends
or ethnic community. Moreover, parental emotional well-being, family
functioning, ethnic community vitality with strong social support network
have all been linked to decreased levels of psychological stress (Sowa et al.,
2000; Stansfeld et al., 2004). Hence, becoming assimilated into the host
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
277
culture may be protective for family functioning and children’s development. Adverse social conditions interact in a complex manner but this study
indicates that immigrant or migrant status should be considered in addition
to the more commonly studied demographic and family factors in
understanding parenting and child behaviour.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
Limitations of this study
There may be unmeasured differences between the migration groups in this
study that might enable more explanation of the differences identified in
their parenting or their children (e.g., personality of parents). In addition it
would have been useful to obtain details about reasons for migration or
immigration, perceptions of the areas in which they lived, religious activity
and (for the immigrant group) the extent to which they were acculturated
into the UK, so that these factors could be linked with parenting practices.
These could be the focus of further research.
Manuscript received 8 January 2009
Revised manuscript accepted 29 January 2010
First published online 29 April 2010
REFERENCES
Aunola, K., & Nurmi, J.-E. (2005). The role of parenting styles in children’s problem behaviour.
Child Development, 76, 1144–1159.
Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology
Monographs, 4, 1–103.
Baumrind, D., & Black, A. E. (1967). Socialization practices associated with dimensions of
competence in preschool boys and girls. Child Development, 38, 291–327.
Belsky, J., Melhuish, E., Barnes, J., Leyland, A., Romaniuk, H., & the NESS Research Team.
(2006). Effects of Sure Start local programmes on children and families: Early findings.
British Medical Journal, 332, 1476–1478.
Berry, J. W. (1992). Acculturation and adaptation in a new society. International Migration, 30,
69–86.
Bhugra, D., & Arya, P. (2005). Ethnic density, cultural congruity and mental illness in migrants.
International Review of Psychiatry, 17, 133–137.
Bhugra, D., & Bhui, K. (1998). Transcultural psychiatry: Do problems persist in the second
generation? Hospital Medicine, 59, 126–129.
Crul, M., & Doomernik, J. (2003). The Turkish and Moroccan second generation in the
Netherlands: Divergent trends between and polarization within the two groups. International Migration Review, 37, 1039–1064.
Crul, M., & Vermeulen, H. (2003). The second generation in Europe. International Migration
Review, 37, 965–986.
Derluyn, I., Broekaert, E., & Schuyten, G. (2008). Emotional and behavioural problems in
migrant adolescents in Belgium. European Child and Adolescent Psychiatry, 17, 54–62.
Diler, R. S., Avci, A., & Seydaoglu, G. (2003). Emotional and behavioural problems in migrant
children. Swiss Medical Weekly, 133, 16–21.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
278
DAGLAR, MELHUISH, BARNES
Fisek, G. O., & Sunar, D. (2005). Contemporary Turkish families. In J. L. Roopnarine & U. P.
Gielen (Eds.), Families in global perspective (pp. 169–183). Needham Heights, MA: Allyn &
Bacon.
Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of
Child Psychology and Psychiatry, 38, 581–586.
Harkness, S., & Super, C. (1995). Culture and parenting. In M. H. Bornstein (Ed.), Handbook of
parenting: Biology and ecology of parenting (pp. 211–234). Hillsdale, NJ: Lawrence Erlbaum
Associates, Inc.
Hazuda, H., Stern, M., & Haffner, S. (1988). Acculturation and assimilation among Mexican
Americans: Scales and population-based data. Social Science Quarterly, 69, 687–705.
Hernandez, D. J., Denton, N. A., & Macartney, S. E. (2008). Children in immigrant families:
Looking to America’s future. Social Policy Report, 22, 3–23.
Janssen, M. M., Verhulst, F. C., Bengi-Arslan, L., Erol, N., Salter, C. J., & Crijnen, A. A.
(2004). Comparison of self-reported emotional and behavioural problems in Turkish
Immigrant, Dutch and Turkish adolescents. Social Psychiatry and Psychiatric Epidemiology,
39, 133–140.
Kagitcibasi, C. (1990). Family and socialisation in cross-cultural perspective. In J. Berman
(Ed.), Cross-cultural perspectives: Nebraska symposium on motivation, 1989 (pp. 135–200).
Lincoln, NE: Nebraska University Press.
Kagitcibasi, C. (2003). Autonomy, embeddedness and adaptability in immigration contexts.
Human Development, 46, 145–150.
Kagitcibasi, C. (2007). Family and human development across cultures (2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum Associates, Inc.
Kia-Keating, M. (2006). Refugee youth in resettlement: Exposure to war and protective factors.
PhD thesis, Boston University.
Kotchick, A. B., & Forehand, R. (2002). Putting parenting in perspective: A discussion of the
contextual factors that shape parenting practices. Journal of Child and Family Studies, 11,
255–269.
Kwak, K. (2003). Adolescents and their parents: A review of intergenerational family relations
for immigrant and non-immigrant families. Human Development, 46, 115–136.
Ladd, G. W., & Profilet, S. M. (1996). Child Behavior Scale: A teacher-report measure of young
children’s aggressive, withdrawn and prosocial behaviours. Developmental Psychology, 32,
1008–1024.
Leavey, G., Hollins, K., King, M., Barnes, J., Papadopoulos, C., & Grayson, K. (2004).
Psychological disorder among refugee and migrant schoolchildren in London. Social
Psychiatry and Psychiatric Epidemiology, 39, 191–195.
Mohammadi, R., Taylor, E., & Fombonne, E. (2006). Prevalence of psychological problems
among Iranian immigrant children and adolescents in UK. Iranian Journal of Psychiatry, 1,
12–18.
Murad, S. D., Joung, I. M. A., van Lenthe, F. J., Bengi-Arslan, L., & Crijnen, A. A. M. (2003).
Predictors of self-reported problem behaviours in Turkish immigrant and Dutch adolescents
in the Netherlands. Journal of Child Psychology and Psychiatry, 44, 412–423.
Pearson, E., & Rao, N. (2003). Socialization goals, parenting practices, and peer competence in
Chinese and English preschoolers. Early Child Development and Care, 173, 131–146.
Querido, J. G., Warner, T. D., & Eyberg, S. M. (2002). Parenting styles and child behaviour in
African-American families of pre-school children. Journal of Clinical Child Psychology, 31,
272–277.
Robinson, C. C., Mandleco, B., Olsen, S. F., & Hart, C. H. (2001). The Parenting Styles and
Dimensions Questionnaire (PSQD). In B. F. Perlmutter, J. Touliatos, & G. W. Holden
(Eds.), Handbook of family measurement techniques. Vol. 3: Instruments & index (pp. 319–
321). Thousand Oaks, CA: Sage.
Downloaded By: [Melhuish, Edward] At: 16:30 17 May 2011
PARENTING, CHILD BEHAVIOUR, AND IMMIGRATION
279
Sammons, P., Sylva, K., Melhuish, E., Siraj-Blatchford, I., Taggart, B., Elliott, K., et al. (2004).
The Effective Provision of Pre-school Education (EPPE) project. Technical paper 11: The
continuing effect of pre-school education at age 7 years. London: Institute of Education.
Shrake, E. K., & Rhee, S. (2004). Ethnic identity as a predictor of problem behaviours among
Korean American adolescents. Adolescence, 39, 601–622.
Sowa, H., Crijnen, A. A. M., Bengi-Arslan, L., & Verhulst, F. C. (2000). Factors associated
with problem behaviours in Turkish immigrant children in the Netherlands. Social
Psychiatry and Psychiatric Epidemiology, 35, 177–184.
Stansfeld, S. A., Haines, M. M., Head, J. A., Bhui, K., Viner, R., Taylor, S. J. C., et al. (2004).
Ethnicity, social deprivation and psychological distress in adolescents in east London.
British Journal of Psychiatry, 185, 233–238.
Steinberg, L., Elman, J. D., & Mounts, N. S. (1989). Authoritative parenting, psychosocial
maturity, and academic success among adolescents. Child Development, 60, 1424–1436.
Stevens, G., Pels, T., Bengi-Arslan, L., Verhulst, F. C., Vollebergh, W. A. M., & Crijnen,
A. A. M. (2003). Parent, teacher and self-reported problem behavior in The Netherlands—
Comparing Moroccan immigrant with Dutch and with Turkish immigrant children and
adolescents. Social Psychiatry and Psychiatric Epidemiology, 38, 576–585.
Sunar, D. (2002). Change and continuity in the Turkish middle class family. In R. Liljestrom &
E. Özdalga (Eds.), Autonomy and dependence in the family. Istanbul, Turkey: Swedish
Research Institute in Istanbul.
Zwirs, B. W. C., Burger, H., Schulpen, T. W. J., & Buitelaar, J. K. (2006). Different treatment
thresholds in non-Western children with behavioural problems. Journal of the American
Academy of Child and Adolescent Psychiatry, 45, 476–483.

Benzer belgeler