metaanaliza povezanosti razvoda i blagostanja dece
TRANSCRIPT
-
8/9/2019 metaanaliza povezanosti razvoda i blagostanja dece
1/21
Psychological
Bulletin
1991, Vol. 110, No.1,26-46
Copyright 1991 by the
American Psychological
Association Inc.
0033-2909/91/J3.00
Parental
Divorce
and theWeil-BeingofChildren:A Meta-Analysis
Paul R.Amato
Department of Sociology
University
of
NebraskaLincoln
Bruce Keith
Department of Sociology
WestVirginia University
This
meta-analysis
involved
92 studies
that
comparedchildrenliving in
divorced
single-parent
families
withchildren
living
in continuouslyintact
families
on measures of well-being. Children of
divorcescoredlowerthanchildren inintactfamiliesacrossa varietyofoutcomes,withthemedian
effectsizebeing.
14
ofastandarddeviation.For someoutcomes,
methodologically
sophisticated
studies yieldedweaker
effect
sizes thandid
otherstudies.
In
addition,
forsome outcomes,
more
recentstudiesyieldedweakereffectsizes
than did
studies
carried
out
during
earlier
decades.
Some
support
wasfoundfor
theoretical
perspectives
emphasizing
parental absenceandeconomic disad-
vantage, but themost
consistent
supportwas found for afamilyconflictperspective.
Thenumberofchildren affected bydivorcehas increased
dramatically since the 1950s. Currently,everyyear more than 1
million children in the United States experience the divorce of
their parents
(US.
Bureau of theCensus,
1989,
p.
87).
Projec-
tions indicate that
38% of
White children
and 75% of
Black
children born
to
marriedparents
will
experience parental
di-
vorce
beforethey reach
the
age of
16
(Bumpass, 1984).
The
large
number ofchildrenaffected bymarital disruption hasgener-
ated both public andscientificconcern over the consequences
ofdivorceforchildren's developmentandwell-being.
Numerous studies
have
examined the implications of paren-
tal
divorce
for
children'sscholasticachievement, conduct, psy-
chological adjustment,
serf-esteem,
social competence,andre-
lationships
with
parents. Many studies
have
foundthat children
from
divorced
families
experience lower levels
of
well-being
across these domains than do children
from
intact families.
However,a good deal of inconsistency exists inthisliterature,
and
many
studies
have failed
to find significant
differences.
In
addition, studies have varied substantially both in their method-
ologies and in thecharacteristics,such as age level and social
class, of the childrenstudied.
This literature has been reviewed in a qualitative
fashion
by
Blechman
(1982),Emery(1982),Goetting
(1981),
Hetherington
(1979), Kurdek
(1981,1983),
and Longfellow (1979), andmore
recentlybyDemoandAcock
(1988),
Edwards
(1987),
Emery
(1988),Hetheringtonand Camara(1984,1988), and Krantz
(1988).Giventheoften contradictory natureofmuchofthis
research,
it is not
surprising that
reviewers
have
sometimes
reached discrepant conclusions. For example, Edwards
(1987)
suggested
that most children recoverfromdivorce with
few
enduringnegative consequences. In contrast, Krantz (1988)
warned
that the
psychosocial adjustment
of
children
of
divorce
Thisresearch was
supported
by a grant from the
University
of Ne-
braskaLincoln
Research Council.
Correspondence concerning this
article
should
be addressed to
Paul
R.
Amato,
Department of Sociology,
University
ofNebraska,Lincoln,
Nebraska 68508-0324.
is at risk. The middle ground was taken by Emery (1988), who
acknowledged that divorce isassociatedwith a numberofnega-
tivechildoutcomes but concludedthatchildren withserious
problems are not markedly overrepresented among divorced
families(p.
70).
Finally,Demo and Acock (1988) found an in-
creasedlikelihood of certain problems for children of divorce
but argued that methodological limitations make itdifficultto
draw
firm conclusions. Overall, reviewers
have
reached quite
differentconclusions
after
examining similarsetsofstudies.
Althoughmany qualitative reviews ofthisliteratureexist,a
meta-analysis has not yet been attempted.\et,asufficientnum-
ber of studies exists atthistime to
justify
a quantitative
review.
Webelieve that a meta-analysis has thepotentialto bring some
order to an otherwise
confused
and contradictory body of find-
ings.
Accordingly,
our first purpose is to estimate the impact of
parental divorce on child well-beingacrossall availablestudies.
This allows us to answer three questions: (a) How large are the
differencesbetween children in divorced andintact
families
on
measures
of
well-being?
(b) Are
thesepooled
differences
statis-
ticallysignificant?
and (c) Are
thesedifferenceslarger
for
some
outcomes thanforothers?
Our meta-analysis has two
further
purposes.The magnitude
of
estimated
effectsmay
varysystematically with studyattrib-
utes, such as the type of analysis used and the nature of the
sample. Consequently, we use meta-analytic techniques to
search
forstudy characteristics
that
account forvariation in
effect
sizes. Our final purpose is to consider the cumulative
evidence
across all available studies for three theoretical
per-
spectives on theeffectsofdivorceonchildren. Each perspective
suggests a number of hypotheses. We
assess
the degree of
sup-
port
for
these hypotheses and,
when
possible,
use
meta-analytic
techniques to combinedataacross studies for this purpose.
Theoretical Considerations
Anumber
of
explanations
and
intervening processes have
been proposedtoaccountfor whydivorce mighthavenegative
effectson
children's
lives
(see Kalter, Kloner, Schreier,
&
Okla,
1989, andMcLanahan,1985,1989, fordiscussions). However,
26
-
8/9/2019 metaanaliza povezanosti razvoda i blagostanja dece
2/21
PARENTAL
DIVORCE AND CHILD WELL-BEING
27
most explanationsfocuson three central notions: parental
ab-
sence, economic disadvantage,
andfamily
conflict.
Parental
Absence
Perspective
Thefamilyis the key
social institution
for
providing
the nur-
turant
socialization
of
youngchildren.Traditionally,ithas been
assumed thatthetwo-parent
family (with
both parents livingin
the same household as the child) is a
better
environment for
children's developmentthanthe single-parentfamily.Accord-
ingly,
the
presumed negativeeffects
of
divorce
areoften
attrib-
utedtosocialization
deficits
thatresultfromgrowingupwith
one parent rather than two.Thisnotion has been criticized for
emphasizingfamilystructureat theexpenseoffamilyprocess
and for
being politically conservative (Marotz-Baden, Adams,
Buech, Munro, & Munro, 1979;
Scanzoni,
Polonko,Teachman,
&Thompson,
1989).
Nevertheless, this perspective
isbasedon
a straightforward principle: If parents are important resources
for childrenlsdevelopment,then,allthings being equal,two
parents should
be
betterthanone.
Itis
well
known that divorce is associated with a
decrease
in
the
quantity
and
quality
ofcontact
between
childrenand
their
noncustodial parent(Amato,1987; Furstenburg&Nord,1985;
White,
Brinkerhoff,
&
Booth,1985).Also, because
most
custo-
dial parents are in the laborforce,they are constrained in the
amountof time andenergy theycan devote to their children
(Brandwein,
Brown,&Fox,1974). Forthesereasons,children
of
divorce
often
experience
a
decrease
in
parental attention,
help, and supervision. This decline in parental support may
increase the likelihood of problems, such as academic failure,
lowself-esteem,
and
misbehavior (Rollins
&
Thomas,
1979).In
addition, the lack of parental models in the household may
result
in
the inadequate learning
of
social skills (such as cooper-
ating,negotiating,andcompromising) necessaryforsuccessin
the larger world.
A
focuson
parental
absence leadsto thefollowinghypothe-
ses. First, children
who
experience
the
death
of a
parent
(or the
lossof aparent
for
anyotherreason)exhibit problems similarto
those of children who experience parental divorce. Second, be-
cause a stepparent can provide an alternative role model and
source of
support, children
of
divorce
havefewer
problems
if
the custodial parent remarries than if the custodial parent re-
mains
single. Third, this perspective suggests that the
disrup-
tive
effectsoflivingin asingle-parent
family
arepartly miti-
gated if noncustodial parents maintain close relationships with
their children. This leads to the hypothesis that thefrequency
andqualityofcontactwiththenoncustodial parentispositively
associated
with
children's well-being.
EconomicDisadvantagePerspective
A
second perspective
is
based
on the
notion
of
economic
disadvantage. Divorce typically leadsto adeclinein thestan-
dard
of
living
of
mother-headedfamilies,oftenpushing them
below poverty level(Duncan &
Hoffman,
1985;Weitzman,
1985).This perspective assumes that it is economic hardship,
ratherthanfamilytypeassuch,thatisresponsiblefor thelow-
ered well-being of children of divorce.
Alack
of
economic resources increases
the
risk
of a
number
of developmental problems in children. Economic hardship
may
negativelyaffectchildren's nutrition
and
health (Williams,
1990).In addition, poor single mothers are unable to afford
private
lessons, educational toys, books, home computers, and
other goods that facilitate children's academic success. Limited
means
may
alsoforcefamilies
to
live
in
neighborhoods
in
which
school programs
are
poorlyfinanced
and
services
are
inade-
quate (McLanahan, 1989). Furthermore,
living
inpovertymay
be
stigmatizing
for
children
and may
facilitate
the
entry
of
ado-
lescents into deviant subcultures(\foydanoff
&Majka,
1988).
Because the economic disadvantage perspective holds that
the problems observed in children of divorce are due primarily
to the loss of income experienced by custodial mothers, it fol-
lows
thatfewdifferencesshouldbeobserved between children
from divorced and intact families if income is controlled statis-
tically,
or if
families
are
matched
on
income level. This
perspec-
tive also leads to the hypothesis that the well-being of children
ofdivorceisenhancedifcustodial mothersremarry,because
this usuallyresults inimprovementsinfinancialstatus. (Note
that this hypothesis is also predicted by the parental absence
perspective)Finally,this perspective suggests the hypothesis
thatchildrenexperience
fewer
problems
if
fathersratherthan
mothershavecustody, becausefathersgenerally earn morein-
come than do mothers.
FamilyConflict Perspective
The third perspective assumes that
conflict
between parents
beforeandduringtheseparation periodis asevere
stressor
for
children.
Interparental
hostility creates an
aversive
homeenvi-
ronment inwhichchildren experiencestress,unhappiness, and
insecurity
(Maccoby&
Martin,
1983).
Obviously, such
an
envi-
ronmentisless than optimalfor thedevelopmentofchildren.
Infact,numerous studies have indicated that
interparental
con-
flict
in
intact marriages has anegative impact on children's
psychologicaladjustment(seeEmery, 1982,
for a
review).Con-
flict is
also
likelyto
stress parents
and
make them lesseffective
indealingwiththeir children
(Hetherington,
Cox,
&Cox,
1982;
Wallerstein
&Kelly,1980). Otherstudiesshow that children
tend to be drawn into
conflict
between theirparents,resulting
infurther
deteriorations in
parent-child
relationships (Amato,
1986;Johnston, Kline,&
Tschann,
1989). Accordingto this
perspective, children
of
divorce exhibit problems,
not
because
of thechangeinfamilystructure,butbecauseof theaccom-
panying
conflict.
Theconflictperspective leads to the hypothesis that children
inintactfamilieswith high levelsofinterparental conflictex-
hibitproblems similar to those of children of divorce. Indeed,
thisperspective suggests that children in harmonious single-
parent
families
may be better adjusted than children in high-
conflictintactfamilies.Asecond hypothesis thatfollowsfrom
this perspective
is
that
the
adjustment
of
children
of
divorce
improveswiththe passage of time since marital dissolution.
The assumption here is that if poor childadjustmentis a reac-
tion to marital discord, then children'sfunctioningshould re-
cover aslevelsofconflict
subside.
Of
course,
in
some cases,
conflictbetween parentsmaycontinueafterthedivorce,partic-
ularlyin relation to
visitation
and the
payment
of
child support.
This leads to a third hypothesis: Children's well-being is in-
-
8/9/2019 metaanaliza povezanosti razvoda i blagostanja dece
3/21
28
PAULR.
AMATO
AND
BRUCE
KEITH
veisely correlated with the level ofpostdivorceconflict that
persistsbetween
parents.
Method
Selection
of
Studies
Studies
were
located through manual searchesat
Psychological
Ab-
stracts, Sociological Abstracts,andtheSoda/SciCTiws/mfec.-computer-
ized
data
bases and the reference sections of reviewarticleswere also
used.
Eachstudy
had to
meet
four
criteriabefore
it was
included
in the
meta-analysis.First, studies
had to
contain
a
sample
of
children living
insingle-parentfamilies
formed
through divorce
or
separation
as
well
as a
sample
of
children fromcontinuously intact
families. A
second
criterionrequiredthe presence of atleastonequantitative measureof
well-being, including scores on tests or scales, ratings, or percentages
inoutcome categories. Third, datafromeachstudyhadtobe presented
ina
formthat allowed
for the
calculation
of at
least
one
effect size.
Fourth, thestudyhad toinvolvechildren; studies that dealt with adult
children of divorcewereset aside for a separate review.(Studies of
college studentswereincluded in this analysis, however). Ninety-two
studies
wereidentifiedthat met these criteria (see the Appendix for a
complete listing); collectively, these studies
involved over
13,000
chil-
dren.
Calculationof Effect
Sizes
We
calculatedeffectsizes
in
severalways.
The
most direct method
of
calculationinvolved
subtracting
themeanscore on the dependent vari-
able
for
theintact samplefromthatof the divorcedsampleand dividing
this
difference
by thepooled whhin-group standard deviation.Often,
meansand
standard deviationswereunavailable
to the reader. In
such
instances,tvalues,Fratios, and correlation coefficientswere
trans-
formedintoeffectsizes, usingthe formulasprovidedbyHedgesand
Olkin (1985).Likewise, percentagedifferences weretranslated into
effect
sizes
by
means
of the
probit
transformation
described
byGlass,
McGaw,
and
Smith (1981). Occasionally,
only
probability values were
reported,so weestimated (valuesfromastandard tablebymatching
the
degrees
of
freedomwith
the
givensignificancelevel. Studies com-
monly
dispatched
with
nonsignificant
findings
early
in a
discussion
of
the results. Becausenonsignificantfindings are as important in a
meta-analysisasthosefoundto be significant, weestimatedthe
effect
sizes in these cases by assuming apvalue of .5. Because all of these
methods produce slight overestimates
of theeffect
size
when the
sam-
ple sizeis small, weusedthe correctionproposedby Hedges and Olkin
(1985) tocalculatetheunbiased estimator. Signs wereaffixedto effect
sizes to reflect the comparativewell-beingof the groups. A negative
signindicatedalowerlevel ofwell-beingforchildrenin thedivorced
group than forthose in the intactgroup;apositive signindicated the
r v rs
We
calculated separateeffectsizes
for
each independent sample in
a
study.
Independent samples existed
when
data were reported sepa-
ratelyforsubgroups
of
children, such
as
boys
and
girls. Altogether,
the
92studies reporteddataon113independent samples. In
most studies,
more than one dependent variable was
used.
Because significance
testsrequirethateffectsizesbeindependent,if two ormoreeffectsizes
were
generated withinthesame outcome category(asdefinedin the
followingparagraph),
the
mean
of
these
was
taken.
In
afewcases,
two
ormorearticles
were
based on the same
data
set; these were treatedasa
single
study in our analysis. In other cases, multiple reports described a
longitudinal
study of the same sample. In these
cases,
we relied on the
publicationthatpresented the most detailedinformation
for
the calcu-
lationof
effect
sizes.Forexample,for thelongitudinal studybyHether-
ington,Cox,andCox,we relied mainly ontheir 1985 publication.In
this way, each independent sample
was
represented onlyonce
in the
meta-analysisfor anyparticular outcome.
Distinctions
can be drawn between two
types
ofeffectsizes, de-
pendingon the nature of the analysisreported. First,it ispossiblefor
effectsizes to be based on the unadjusted,zero-order
differences
be-
tween
intact
and
divorced groups; theseeffect sizes
reflect thetotal
covariationbetween parental divorce and the outcome. Some of
this
covariance may be
causal
and some may be
spurious.
Second, many
studies
adjustthe covariance
forcharacteristics
such as parental educa-
tion,familyincome, andfamilysize.
Similarly;
it is possible to match
children in divorced and
intactfamily
groups on these
variables.Effect
sizes
calculated
on
thisbasis
reflect the
estimated
direct
effect
of
paren-
tal divorce on the child's well-being,afteradjustments are made for
control variables.
If the
control variables
are
ones that precede both
divorce
and
children's outcomes (such
asparental
education),
the re-
sulting
effect
size reflects the
estimated
totaleffectof divorce on
chil-
dren. On the other hand, if the control variables
follow
divorce(suchas
household income), theresultingeffectsize is likely to be an underesti-
mate
of
the
total
effectof
divorce.
Because
of the
clear
differencein
interpretation betweeneffectsizes
with
and without controls, we calculated each separatelywhenboth
adjustedand unadjusted resultswereprovided in astudy.However,
because only a minority of studies used control variables (or matching
of subjects),and becausemanyofthese studiesmixed
pre-
andpostdi-
vorce controls in analyses, we lumped alleffect
sizes
basedonany
controls inthesame category. Consequently, thereadershould beaware
that thereissome ambiguityin theinterpretationof theresultingeffect
sizes.
Variables
Wecoded outcome measures into the following eight
categories:
(a)
academic achievement (standardized achievement tests, grades,
teachers'
ratings, orintelligence) ;(b)conduct (misbehavior, aggression,
ordelinquency);(c)psychological adjustment (depression,anxiety, or
happiness);
(d)
self-concept
(self-esteem,
perceivedcompetence,
or in-
ternal
locus of control); (e) social adjustment (popularity,
loneliness,
or
cooperativeness);(f)
mother-childrelations(affection,
help,
or
quality
ofinteraction); (g)
father-child
relations; and
(h)
other. The categories
reflectthe outcomes most
frequently
studied in relation to divorce and
are
the ones usually discussed in reviews of this literature.
Because one purpose of this study was to searchforsources of varia-
tion ineffectsizes between studies, we coded a variety of studycharac-
teristics.Severalreviewershave pointed
out the
methodological limita-
tions of much of the research onthistopic, including reliance on small
convenience or clinical samples, the use ofpsychometricallyweak
measuresofchild outcomes, and thefailureto control
for
confounding
family-of-origincharacteristics, such associalclass(Blechman,1982;
Demo & Acock, 1988; Emery, 1988; Kurdek,1983).Given thediffer-
ences
in
methodological sophistication between studies,
we recorded
information
on the
following
variables: the type of sample
used
(clini-
cal, convenience, or random sample); the sample size; the use
of single-
versus multiple-item measures of outcomes; and whetherstatistical
controls
(or
matching
of
subjects)wereused.
As an
exploratory mea-
sure,
we also considered the source ofinformationon childoutcomes:
child, parent, teacher, researcher,
or
other
(in
general,
the researcher
categoryrefers tobehavioral observations, whereasthe first 3catego-
ries arebasedon questionnaire on test responses).
Wealso included variables that reflect substantive characteristics of
the sample. The sex of the sample (all boys, all girls, or mixed) was
recorded
because
of the
current interest
in sex
differences
inchildren's
adjustment
to
divorce(Zaslow,1988, 1989). Because differences
in
children's reactions to divorce depending on the age of the childhave
been
reported(e*,
Wallerstein
&Kelly,1980), we included the mean
-
8/9/2019 metaanaliza povezanosti razvoda i blagostanja dece
4/21
PARENTALDIVORCE AND CHILD WELL-BEING
29
age of children in thesample.Longitudinal studieshave shown
that
the
effects
of divorce tend todiminish
with
time
(e.g.,
Hetherington etal.,
1982).For this reason, we recorded the mean number of years since
separationforchildren in the sample.(Unfortunately,thisinformation
was
provided
in
only
40% of
studies.)
We
included
the
year
in
which
datawerecollected to see ifeffectsizes vary over time. If the authors
did not statewhenthedatawere collected, we assumed it was 2 years
before
the
year
of
publication. Finally,
the
country
in
which
the
study
was conducted was recorded for exploratorypurposes.
Wealsoattempted to
code
the race of the sample (Black, White, or
mixed).
But as it
turned out,onlyfivestudies included data
on all
Black
samples. As a result, comparisons between White and Blacksamples,
although
of
considerable interest, were impossible to carry out.Wecan
only
conclude that too little research has been conducted on the impli-
cationsofdivorceforBlack children.Giventhatthe rate ofseparation
and
divorce
is
considerably higher
for
Blacks thanforWhites(Walker,
1988),thisomission
is
surprising.
1
Reliability
Although
the
coding
of
most variables
was straightforward, the
clas-
sification
of outcomes into categories leftsome room for disagree-
ment. Consequently, we performed a reliability check in
which
each of
us independently classified 80 outcomes. TheresultingCohen's kappa
value
was .84
(p