MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 1

Unauthorized reproduction of this article is prohibited.

PARENTINGSELF-EFFICACY
Terelationshipbetweenparentingandchildrens
healthy development has been well-documented
(Dishion&McMahon,1998;Gardner,Ward,Bur-
ton,&Wilson,2003),andthisassociationhasbeen
foundtobeparticularlyevidentduringearlychild-
hoodwhenparentsplayamoreprominentrolein
childrenssocializationprocessrelativetoolderages
(Campbell,Shaw,&Gilliom,2000;Gardner,1987;
Shaw & Bell, 1993). Also during early childhood,
parentingself-efcacy(PSE),orperceivedcompe-
tenceintheparentingrole,hasbeenpositivelylinked
withobservedparentingcompetence(e.g.,warmth,
sensitivity,engagement).Infact,PSEhasbeenfound
tomediateassociationsbetweenparentingandchil-
drensdevelopmentaloutcomes(Coleman&Karraker,
2003).InlightofthispromisingresearchonPSE,it
issomewhatsurprisingtofndadearthofliterature
onthedevelopmentalcourseorstabilityofPSEor
longitudinaldatalinkingitsdevelopmentwithchil-
drenssocioemotionaloutcomes.Inaddition,most
Parenting Self-Efcacy and Problem Behavior
in Children at High Risk for
Early Conduct Problems: Te Mediating Role
of Maternal Depression
ChElSEA M. WEAVER, DANIEl S. ShAW, ThoMAS J. DIShIoN, MElVIN N. WIlSoN
AbstrAct
Parenting self-efcacy (PSE) has been positively linked to childrens adjustment and negatively as-
sociated with maternal depression. however, most PSE research has been cross-sectional, limited to
predominantly white, middle-class samples, and has not examined potential mechanisms underlying
associations of PSE with childrens behavior. The present study investigates (1) how PSE changes over
time, (2) the relationship between age 2 PSE and childrens behavior problems two years later, (3) and
the potential mediating role of maternal depression in relation to the association between PSE and
child problem behavior. Participants are 652 ethnically and geographically diverse mothers and their
children, at high risk for conduct problems. PSE increased between ages 2 and 4 and higher initial lev-
els predicted lower maternal- and alternate caregiver-reported age 4 conduct problems after control-
ling for initial problem behavior. Maternal depression at age 3 mediated these relationships. Implica-
tions for parenting interventions, focusing on PSE and maternal psychopathology, are discussed.
Key words: Parenting Self-Efcacy, Child Problem Behavior, Early Childhood, Maternal Depression
researchonPSEhasbeenconductedwithpredomi-
nantlyEuropean-American,middle-classfamilies
(seeRaver&Leadbeater,1999,andZayas,Jankowski,
&McKee,2005fornotableexceptions).Evenfewer
studieshaveexaminedpotentialmediatingfactors
thatmightaccountforassociationsbetweenPSEand
childproblembehavior.WhilePSEhasbeenfound
tomediateassociationsbetweenparentingandchild
outcomes,itispossiblethatunderlyingintrapersonal
characteristicssuchasparentalwellbeingmightcon-
tributetoPSEanditsassociationwithchildoutcomes.
Tecurrentstudysoughttoadvanceourknowledge
onPSEbyexaminingitsdevelopmentalcoursefrom
thetoddlertopreschoolperiod,byexploringitslon-
gitudinalrelationwithemergingchildrensconduct
problems,andtestingwhetherassociationsbetween
PSEandchildrensconductproblemswerepotentially
mediatedbymaternaldepressivesymptoms.Testudy
wascarriedoutwithalargesampleof652children
athighriskfordevelopingearlyconductproblems
2 MANUSCRIPT UNDER REVIEW: Infant Behavior and Development
Unauthorized reproduction of this article is prohibited.
WEAVER,SHAW,DISHION,WILSON
basedonthepresenceofsocioeconomic,family,and
childriskfactors.
Parenting self-Efficacy and General self-Efficacy
theory
PSEhasbeendefnedasthedegreetowhichparents
expecttocompetentlyandefectivelyperformtheir
rolesasparents(Teti&Gelfand,1991),anditisrooted
ingeneralself-efcacytheory.Guidedbysociallearn-
ingtheory,generalself-efcacyreferstothebeliefin
onesabilitytoperformbehaviorssuccessfully(Ban-
dura,1977).Overall,self-efcacyincludesthemoti-
vation,cognitiveresources,andcoursesofaction
necessarytoimplementcontroloveraspecifctask
orevent(Ozer&Bandura,1990).Self-efcacyrefects
perceivedself-competenceasopposedtoexpectations
oftasksuccessorfailure(Bandura,1977).Inother
words,anindividualwithhighself-efcacymayantic-
ipatetaskfailureinasituationthatwouldrealistically
requireadvancedexpertiseinaspecifcdomain.
Ingeneral,peoplewhohavehighself-efcacyhave
beenfoundtotrusttheircompetenceinthefaceof
environmental demands, conceptualize difficult
situationsaschallenges,havelessnegativeemotional
arousalinthefaceofstress,andexhibitpersever-
ancewhenchallenged(Jerusalem&Mittag,1995).
Incontrast,peoplewithlowerself-efcacytendto
doubtthemselvesandexperiencehighlevelsanxiety
whenfacedwithadversity,assumemoreresponsi-
bilityfortaskfailurethansuccess,interpretchal-
lengesasthreats,avoiddifculttasks,andcopedys-
functionally.Inlightofthesecognitive,emotional,
andbehavioralmanifestations,itiscriticaltostudy
self-efcacywithinthedomainofparenting.Specif-
cally,itisimportanttounderstandhowPSEunfolds
overtime,particularlyduringearlychildhoodwhen
childrenarequicklygainingincreasedmobilityand
independenceinthecontextoflimitedcognitiveabili-
ties,therebypresentinguniquechallengestoparents
toprotecttoddlerssafetyandthesafetyofvaluables,
pets,andotherfamilymembers,particularlyyounger
siblings(Shaw&Bell,1993).
the Developmental course of Parenting self-Efcacy
DespitethetheoreticalsignifcanceofPSE,weknow
relativelylittleaboutitsdevelopmentalcourseduring
earlychildhood,aperiodoftimethathasbeenfound
tobecriticalforparentingandchildrenssubsequent
adjustment(Shaw,Bell,&Gilliom,2000).During
toddlerhood,childrenrapidlyacquirearepertoireof
cognitive,social,andmotorskills,challengingpar-
entstotailortheirparentingtechniquestochildrens
changingneeds.Inturn,childrenduringthiskey
developmentalperiodareespeciallydependentonthe
infuenceoftheircaretakersintermsoftheirsocial
andemotionaladjustment.Becauseoftheuniquecir-
cumstancespresentedduringthisperiodofdevelop-
mentnotonlyforchildren,butalsoforthegrowthof
parents,itiscriticaltounderstandhowPSEunfolds
specifcallyduringearlychildhood.Onestudyfol-
lowedchangesinPSEamonglow-incomeminority
mothersfromthethirdtrimesterofpregnancyto
threemonthspost-partum(Zayasetal.,2005).Te
authorsfoundthatPSEsignifcantlyincreasedduring
thetransitiontomotherhood.Inoneofthefewother
longitudinalstudiesusinganethnicallydiverse,mid-
dle-classsample,Gross,Conrad,Fogg,andWothke
(1994)examinedchangesinPSEamongtwocohorts
ofchildrenfromages1to2and2to3,respectively,
withmeasurementsofPSEthreetimeseachyear.PSE
wasfoundtoincreasebetweenages1and2inCohort
1,butremainstablefromages2to3inCohort2.
Inlightofthesestudies,thereissomeevidencethat
PSEinitiallyincreasesinthefrstcoupleofyearsof
childrenslivesandthenshowsmoderatestability;
however,longitudinalmodelingofgrowthparameters
acrossatleastthreemeasurementwavesisneededto
elucidatethedevelopmentaltrajectoryofPSE.
Parenting self-Efcacy and childrens socioemotional
Adjustment
Asearly-onsetbehaviorproblemsbeginningasearly
asthetoddlerperiodhavebeenlinkedtothedevel-
opmentofmoresevereconductproblemsinmiddle
childhoodandadolescence,suchasdelinquencyand
substanceuse(Campbelletal.,2000;Shaw&Gross,
inpress),itiscriticaltoidentifyfactorsinearlychild-
hoodthatcontributetoorprotectchildrenfromthe
maintenanceofproblembehavior.BasedonPSEs
theoreticalandempiricallinkstoseveraldimensions
ofparentingbehavior(Bor&Sanders,2004;Coleman,
Trent,Bryan,King,Rogers,&Nazir,2002),itislogi-
caltoconsideritsassociationswithproblembehavior
duringearlychildhood.Infact,PSEhasbeenlinkedto
childrensdevelopmentintermsofbehavioraladjust-
ment(Bor&Sanders,2004).Forexample,asearly
as5months,PSEwasfoundtobepositivelyrelated
toconcurrentratingsofinfantsoothability(Leerkes
& Crockenberg, 2002). At age 2, Raver and Lead-
beater(1999)foundthatPSEwasinverselyrelatedto
childrensconcurrentdifculttemperamentamong
asampleofurbanimpoverishedfamilies.Further-
more,childrensobservedcompliance,negativity,and
avoidanceofmotheratage2wasfoundtobeassoci-
atedwithconcurrentratingsofPSEamongpredomi-
nantlymiddle-class,mother-toddlerdyads(Coleman
&Karraker,2003).Amongademographicallysimilar
sampleofmotherswithschool-agedchildren,higher
PSEwasconcurrentlyassociatedwithlessemotionally
reactiveandmoresociablebehavior(Coleman&Kar-
MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 3
Unauthorized reproduction of this article is prohibited.
PARENTINGSELF-EFFICACY
raker,2000).Intermsofproblembehavior,lowerlev-
elsofPSEamongmothersofpreschool-agedchildren
athighriskfordevelopingconductproblemswere
foundtobeassociatedwithhigherlevelsofconcur-
rentchildrensdisruptivebehaviors(Bor&Sanders,
2004).Furthermore,mothersofclinically-referred,2-
to8-year-oldchildrenwithconductproblemsreport-
edlowerlevelsofPSEthanacomparativecommunity
sample(Sanders&Woolley,2005).
Despitetheexistenceofseveralstudiesexamining
relationsbetweenPSEanddiferentdimensionsof
childfunctioning,mostworkinthisareahasbeen
cross-sectional.Furthermore,withnotableexceptions
(Raver&Leadbeater,1999;Zayasetal.,2005),most
work on PSE has been conducted using predomi-
nantlywhite,middle-classsamples(e.g.,Colemanet
al.,2002;Leerkes&Crockenberg,2002;Teti&Gel-
fand,1991).Tus,moreresearchthatincorporates
culturallyandsocioeconomicallydiversesamplesis
neededtoextendpreviousfndingstounder-repre-
sentedgroups.
Parenting self-Efcacy and childrens conduct Problems
and the Mediating role of Maternal Depression
Inadditiontoneedingmorestudiesthattracethe
developmentcourseofPSEovertimeandlinkPSE
tospecifcchildoutcomesusingsamplesofchildren
atriskforhighratesofclinically-meaningfulpsycho-
pathology,fewresearchershaveattemptedtoexam-
inepotentialmediatorsofassociationsbetweenPSE
andchildadjustment.Tedevelopmentofgeneral
self-efcacyisthoughttoresultfromfourdomains
ofexperiences,includingindividualhistoriesoftask
successesversusfailures,vicariouslylearningthrough
thesuccessesandfailuresofothers,verbalpersua-
sionfromothers,andaversivephysiologicalarousal,
suchasstressresponses(Bandura,1989).WhilePSE
hasbeentheoreticallylinkedtoanumberofearlylife
experiences(e.g.,historyofchildhoodmaltreatment,
maternalstress),maternaldepressionrepresentsone
importantintrapersonalflterthroughwhichthedet-
rimentalefectsofmaternalPSEonchildrensfunc-
tioningmaybetransmitted.Specifcally,hallmarksof
depressionsuchasfeelingsofhelplessnessandworth-
lessnessmayunderlieanddrivefeelingsoflowmater-
nalPSEandleadtothedevelopmentofchildrens
problembehavior.Tus,associationsbetweenlowPSE
andchildrensconductproblemsmaybeaccounted
forbythenegativecognitionsandafectassociated
with depression. As both clinical depression and
subclinicalelevateddepressivesymptomatologyhave
bothbeenassociatedwithchildrensmaladjustment
(Cummings,Keller,&Davies,2005;Farmer,McGuf-
fn,&Williams,2002),thetermsmaternaldepressive
symptomsanddepressionareusedthroughoutthis
papertodescribeelevatedsymptomsthatweremea-
suredonacontinuousscale.
Tobuildalogicalcaseforconsideringmaternal
depressionasamediatorbetweenPSEandchildcon-
ductproblems,itisimportanttoestablishitsrelation-
shipwithbothPSEandchildfunctioning.Infact,
maternaldepressionhasbeenfoundtobeassociated
withbothlowPSE(Bor&Sanders,2004;Haslam,Pak-
enham,&Smith,2006;Teti&Gelfand,1991;Zayaset
al.,2005)andchildrensproblembehavior,including
conductproblems(Owens&Shaw,2003;Shawetal.,
1994).Forexample,BorandSanders(2004)foundthat
maternaldepressivesymptomswerenegativelyrelated
toconcurrentPSE.However,asmostresearchonthe
relationshipbetweenmaternaldepressionandPSE
hasbeencross-sectional,itisdifculttodetermine
theprecisedirectionofefects.Overall,studiesexam-
iningassociationsbetweenmaternaldepressionand
diferentfacetsofchildadjustmenthavebeenmore
prevalent,includinglongitudinalstudiesexamining
associationswithchildconductproblemscarriedout
inearlychildhood.Whilenotuniformlyconsistent
andsometimessuferingfrominformantresponse
bias(Fergusson,Lynskey,&Horwood,1993)byrely-
ingononeinformantforreportsofbothmaternal
depressionandchildproblembehavior,thepattern
ofoverallresultssuggestalongitudinalassociation
betweenmaternaldepressioninearlychildhoodand
bothpreschoolandlaterschool-ageconductproblems
(Shawetal.,2000).
IntermsofestablishingthelinkbetweenPSEand
maternaldepression,researchongeneralself-efcacy
andnegativelifeeventsconsideredtobedependent
onmaternalbehavior(e.g.,divorce,changeinresi-
dence,lossofjob)suggestthatlowerself-efcacyis
more prevalent among mothers with histories of
depression(Maciejewski,Prigerson,&Mazure,2000).
Tesefndingssuggestthatmaternaldepressioncould
playanimportantroleinafectingthecourseofPSE,
andultimatelyaffectingtheimpactoflowPSEon
childrensproblembehavior.Giventheestablished
linkbetweenmaternaldepressionandchildproblem
behavior,particularlyconductproblems(Owens&
Shaw,2003;Shawetal.,1994),wesoughttotestthe
possibilitythatmaternaldepressionmayserveasan
underlyingmechanismlinkingPSEtochildrenscon-
ductproblems.
Present study
Thecurrentstudyhadthreeprimarygoals.First,
wesoughttocharacterizethetrajectoryofPSEover
threetimepointsduringearlychildhoodamonga
largesampleoffamiliesfacinghighlevelsofsocio-
economic,family,andchildrisk.Second,associations
betweenPSEandlaterchildconductproblemswere
4 MANUSCRIPT UNDER REVIEW: Infant Behavior and Development
Unauthorized reproduction of this article is prohibited.
WEAVER,SHAW,DISHION,WILSON
explored,controllingforage2childproblembehavior
andusingmultiplereportersofchildproblembehav-
iortwoyearsaferinitialassessmentofPSE.Athird
goalwastoexploretheroleofmaternaldepressive
symptomsasapotentialmediatoroftherelationship
betweenPSEandsubsequentchildconductproblems.
Incomparisontothevastmajorityofpriorresearch
onPSE,thecurrentstudy,theEarlyStepsMultisite
Project,includedalongitudinaldesign,theuseofa
large,ethnicallydiversesampleofchildrenathigh
riskforclinically-meaningproblembehavior,andthe
explorationofapotentialmechanismunderlyingthe
relationshipbetweenPSEandchildconductproblems:
maternaldepression.
Intermsofhypotheses,basedonthelimitedlongi-
tudinalresearchonPSE(Grossetal.,1994;Zayaset
al.,2005)andmothersincreasingexperienceinthe
parentingrole,weexpectedtoseeincreasesinPSE
overtime.Second,basedonpriorresearch(Bor&
Sanders,2004,Sanders&Woolley,2005),weexpected
PSEtobenegativelyassociatedwithchildrenslater
conductproblemsevenaferaccountingforinitial
conductproblemsusingbothmaternalandalternate
caregiver(AC)reports(testedasseparatestructural
models).Finally,basedonpriorcross-sectionalstud-
ieslinkingPSEtomaternaldepression(Haslametal.,
2006;Zayasetal.,2005)andmaternaldepressionto
laterchildconductproblems(Owens&Shaw,2003),
maternaldepressivesymptomsatage3wereexpected
tomediateassociationsbetweenPSEatage2andboth
maternalandACreportsofchildconductproblems
atage4.
Method
Participants
Participantswereasampleof652mother-childdyads
drawnfromalargerpreventiveinterventionstudy(N
=731)recruitedbetween2002and2003fromWIC
ProgramsinthemetropolitanareasofPittsburgh,
PA,andEugene,Oregon,andwithinandoutsidethe
townofCharlottesville,VA(Shaw,Wilson,Gardner,
&Dishion,2007).FamilieswereapproachedatWIC
sitesandinvitedtoparticipateiftheyhada2-year-
oldchild,followingascreentoensurethattheymet
thestudycriteriabyhavingsocioeconomic,family,
and/orchildriskfactorsforfuturebehaviorproblems.
Riskcriteriaforrecruitmentweredefnedatorabove
one standard deviation above normative averages
withinthefollowingthreedomains:(a)childbehavior
(conductproblems,high-confictrelationshipswith
adults),(b)familyproblems(maternaldepression,
dailyparentingchallenges,substanceuseproblems,
teenparentstatus),and(c)socio-demographicrisk
(loweducationachievementandlowfamilyincome
usingWICcriterion).Twoormoreofthethreerisk
factorswererequiredforinclusioninthesample.
Ofthe731families(49%female),272(37%)were
recruited in Pittsburgh, 271 (37%) in Eugene site,
and188(26%)inCharlottesville.Moreparticipants
were recruited in Pittsburgh and Eugene because
ofthelargerpopulationofeligiblefamiliesinthese
regionsrelativetoCharlottesville.Acrosssites,the
childrenwerereportedtobelongtothefollowing
racialgroups:27.9%AfricanAmerican(AA),50.1%
EuropeanAmerican(EA),13.0%biracial,and8.9%
otherraces(e.g.AmericanIndian,NativeHawaiian).
Intermsofethnicity,13.4%ofthesamplereported
beingHispanicAmerican(HA).Duringtheperiodof
screeningfrom2002to2003,morethantwo-thirdsof
thosefamiliesenrolledintheprojecthadanannual
incomeoflessthan$20,000,andtheaveragenumber
offamilymembersperhouseholdwas4.5(SD=1.63).
Forty-onepercentofthesamplehadahighschool
diplomaorGEDequivalency,andanadditional32%
hadonetotwoyearsofposthighschooltraining.Te
childreninthesamplehadameanageof29.9months
(SD=3.2)atthetimeoftheage2assessment.
Retention.Ofthe731familieswhoinitiallypar-
ticipated,659(89.9%)wereavailableattheone-year
follow-upand619(84.7%)participatedatthetwo-year
follow-upwhenchildrenwere4yearsold.Atages3
and4,selectiveattritionanalysesrevealednosignif-
cantdiferencesinprojectsite,childrensrace,eth-
nicity,orgender,initiallevelsofmaternaldepression
orchildrensexternalizingbehaviors(parentandAC
reports).Furthermore,nodiferenceswerefoundin
thenumberofparticipantswhowerenotretainedin
thecontrolversustheinterventiongroupsatbothages
3(n=40andn=32,respectively)and4(n=58and
n=53,respectively).
Forthepresentstudy,participantswhocomplet-
edonlyoneassessmentwave(n=52)andprimary
caregiverswhowerenotthebiologicalmotheratall
timepoints(n=27)wereomittedfromallanalyses,
yieldingatotalsamplesizeof652forthecurrent
study.Omittedparticipants(n=79)didnotdifer
fromtheremainderofthesamplebasedonrace,site,
treatmentgroupstatus,maternallevelofeducation,
orage2levelsofmaternaldepressionorchildrens
problembehaviorasindicatedbytheEybergIntensity
andProblemScales(bothmaternalandACreports)
andtheACreportoftheChildBehaviorChecklist
Externalizingsubscale.However,theomittedpartici-
pantsdidhavesignifcantlyhighermaternal-report-
edExternalizingsubscalescoresascomparedtothe
retainedchildren,t(728)=2.16,p=0.03.
ForanalysesusingACreportsofchildrensproblem
behaviors,participantswereincludediftheyhada
participatingACforboththeages2and4assessment
MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 5
Unauthorized reproduction of this article is prohibited.
PARENTINGSELF-EFFICACY
waves,yieldingasamplesizeof267.Forthecurrent
studyatage2,51%ofACswerebiologicalfathers,19%
weregrandmothers,7%werematernalparamours,
andtheremainderrangedinrelationshipfromstep-
fathertocousin.Atage4,44%ofACswerebiological
fathers,19.5%weregrandmothers,7.5%werematernal
paramours,and7.5%wereaunts,withtheremainder
ranginginrelationshipstatus.Sixty-threepercentof
ACswerethesamereporterfromage2toage4.
Design and Procedure
Mothersand,ifavailable,ACs,whoagreedtopartici-
pateinthestudywerescheduledfora2.5-hourhome
visitwhenchildrenwerebetween2yearsand2years
11monthsold.Eachassessmentinvolvedaseriesof
interactivetasksandcaregiverscompletedseveral
questionnaires,includingthoseaboutPSE,maternal
well-being,andchildproblembehavior.Tehome
visitprotocolwasrepeatedatages3and4.
Familiesreceived$100forparticipatingintheage
2homevisit,$120attheage3assessment,and$140
attheage4assessment.Randomizationtotreatment
wasbalancedongendertoassureanequalnumber
ofmalesandfemalesinthecontrolandinterven-
tionsub-sample.Toensureblindness,theexaminer
openedasealedenvelope,revealingthefamilysgroup
assignmentonlyafertheassessmentwascompleted,
andsharedthisinformationwiththefamily.Exam-
inerscarryingoutfollow-upassessmentswerenot
informedofthefamilysassignedcondition.Fora
detaileddescriptionoftheintervention,seeShaw
etal.(2007).Forthepurposesofthecurrentstudy,
treatmentgroupstatuswasusedasacovariateinall
analyses.
MaternalreportsofPSEfromages2,3,and4,were
usedforthecurrentstudy.MaternalandACreports
ofchildproblembehaviorwereusedfromtheages2
and4assessments,withmaternalreportsofdepres-
sionbeingusedfromtheage3assessment.
Maternal Measures
Demographics questionnaire. Demographic data
werecollectedfrommothersduringtheage2visit.
This measure included questions about family
structure,parentaleducationandincome,parental
criminalhistory,andareasoffamilialstressand
strengths.
Parenting self-efcacy.Maternalself-efcacyspe-
cifcallyrelatedtoparentingwasmeasuredusingthe
10-itemEfcacysubscaleoftheParentingSenseof
CompetenceScale(Johnston&Mash,1989).Items
were rated on a 6-point likert scale ranging from
stronglyagreetostronglydisagree.Becauseofthe
high-risknatureofoursample,itemswerereworded
torefectan8thgradereadinglevel.Forexample,the
originalitem,Teproblemsoftakingcareofachild
areeasytosolveonceyouknowhowyouractions
afectyourchild,anunderstandingIhaveacquired
wasrewordedasInowrealizetheproblemsoftaking
careofachildareeasytosolveonceyouknowhow
youractionsafectyouchild.Alphareliabilityofthe
itemswithineachtimepointinthecurrentsample
rangedfrom0.69to0.72.
Center for Epidemiological Studies on Depression
scale (CES-D).TeCES-D(Radlof,1977)isawell-
established and widely used 20-item measure of
depressivesymptomatologythatwasadministeredto
mothersattheage3homeassessment.Participants
reportedhowfrequentlytheyhadexperiencedalist
ofdepressivesymptomsduringthepastweekona
scalerangingfrom0(lessthanaday)to3(5-7days).
Itemsweresummedtocreateanoveralldepressive
symptomsscore.Forthecurrentsample,theinternal
consistencywas0.77.
childrens Measures
Child Behavior Checklist (CBCL; Achenbach &
Rescorla, 2000).TeCBCLforAges1.5-5isa99-
itemquestionnairethatassessesbehavioralproblems
inyoungchildren.MotherscompletedtheCBCLat
theages2and4visits.Tebroad-bandExternal-
izingfactorwasusedtoevaluatethefrequencyof
problembehaviorduringthestudyperiod.Internal
consistenciesforExternalizingwere0.86and0.91
formaternalreportsatages2and4,respectively,
andwere0.90and0.91forACreportsatages2and
4,respectively.
Eyberg Child Behavior Inventory. This 36-item
behaviorchecklistalsowasadministeredattheages
2and4assessments(Robinson,Eyberg,&Ross,1980).
TeEybergincludestwofactorsthatfocusontheper-
ceivedintensityandwhetherornotthebehaviorisa
problemforcaregivers.TeIntensityfactormeasures
thecaregiver-reportofthestrengthoftheproblem
behaviorusingaseven-pointscale.TeProblemfac-
torconsistsofdichotomousratingsofwhetherornot
eachbehaviorservesasaproblemforthecaregiver.
TeInventoryhasbeenhighlycorrelatedwithinde-
pendentobservationsofchildrensbehavior,difer-
entiatedclinic-referredandnon-clinicpopulations
(Robinsonetal.,1980),andshowedhightest-retest
reliability(0.86)andinternalconsistency(0.98;Web-
ster-Stratton,1985).Inthecurrentstudy,alphareli-
abilitiesfortheProblemfactorwere.85and.92atages
2and4formaternalreports,respectively,and.91and
.93atages2and4forACreports,respectively.Alpha
reliabilitiesfortheIntensityfactorwere.86and.94at
ages2and4formaternalreports,respectively,and.92
and.94atages2and4forACreports,respectively.
6 MANUSCRIPT UNDER REVIEW: Infant Behavior and Development
Unauthorized reproduction of this article is prohibited.
WEAVER,SHAW,DISHION,WILSON
results
Descriptive statistics and Intercorrelations
Meansandstandarddeviationsforstudyvariablesare
presentedinTable1.ItshouldbenotedthatT-scores
(M=50,SD=10)arepresentedforallmeasuresof
childrensconductproblems.Onaverage,children
werenearlyonestandarddeviationabovethenorma-
tivemeanformaternalreportsofage2externalizing
behaviorsasmeasuredbytheChildBehaviorCheck-
list(M=59.23,SD=7.95),age2maternal-reported
Eybergintensityofbehavior(M=58.94,SD=7.95),
andmaternal-reportedEybergproblembehaviorscale
atbothages2and4(M=59.07,SD=8.40andM=
59.72,SD=10.94,respectively).Intermsofmaternal
depression,scoresof16andaboveontheCES-Dhave
beenfoundtorefectclinically-meaningfuldepressive
symptoms(Eaton&Kesslerm,1981;Myers&Weiss-
man,1980).Specifcally,thiscut-ofscoreshoweda
modestrelationshipwithclinicaldepressionincom-
munitysamples(Wetzler&vanPraag,1989).Onaver-
ageatage3,mothersinthesamplereportedelevated
levelsofdepressivesymptomatology(M=15.59,SD=
11.02),withover60%scoringatorabove16.
Intercorrelationsofstudyvariablesarepresentedin
Table2.Withtheexceptionofthenon-signifcantcor-
relationbetweenage2PSEandAC-reportedEyberg
problemscoreatage4andthetrendbetweenage2
PSEandAC-reportedexternalizingproblems(r=-
.10,p=0.10),allvariablesweresignifcantlycorrelated
intheexpecteddirections.Specifcally,(1)PSEscores
werepositivelyassociatedacrosstime,(2)PSEscores
werenegativelycorrelatedwithchildrensproblem
behaviorsatages2and4andmaternaldepression
atage3,and(3)age3maternaldepressionwasposi-
tivelyassociatedwithallfactorsofchildrensproblem
behaviorsatages2and4.
Data Analysis Approach
Latentgrowthcurvemodeling(LGCM)wasusedto
modelinitiallevelsof,andchangesin,PSEovertime.
LGCMcapturesindividualdiferencesindevelopment
longitudinallybyusingstructuralequationmodeling
toestimatethemeanintercept(startingpoint)and
slope(rateofchange)ofindividualsbasedontheir
observedscoresonmultipleindicatorsofaspecifc
constructofinterest.Giventhescoresontheobserved
variables,maximumlikelihoodestimatesareusedto
fndthemostlikelyvaluesoftheunobservedlatent
growthparameters.Temeanlatentinterceptand
slopearethenusedtodescribetheshapeoftheaver-
agegrowthcurve(McArdle&Bell,2000).
LGCMscanbeusedtomakepredictionsaboutthe
meansandcovariancesofthedata,yieldingftindi-
cesincludingChi-squaregoodnessofftsignifcance
test,theComparativeFitIndex(CFI;Bentler,1990),
Tucker-Lewis Index (TLI; Tucker & Lewis, 1973),
RootMeanSquareErrorofApproximation(RMSEA;
Browne&Cudeck,1993),andtheStandardizedRoot
MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 7
Unauthorized reproduction of this article is prohibited.
PARENTINGSELF-EFFICACY
MeanSquareResidual(SRMR).CFIandTLIvalues
greaterthan0.95andRMSEAandSRMRvaluesless
than0.05indicategoodmodelft.Anonsignifcant
Chi-squarevalueindicatesthattheproposedcovari-
ancematrixdoesnotdiferfromtheobservedcovari-
ancematrix,suggestingadequatemodelft.Because
theChi-squaretestissensitivetosamplesize(Wida-
man&Tompson,2003),theratioofChi-square/df
providesabetterftindexwithlargersamplesizes
(Bollen,1989),withvalueslessthan2.5indicating
adequatemodelft.Forallanalyses,missingdatawere
handled using full information
maximumlikelihoodestimation
inMplusVersion4.0(Muthn,&
Muthn,1998).
Tests of mediation were con-
ductedusingthebootstrapsam-
plingmethodtotestthesignifi-
canceoftheindirectpath(Shrout
&Bolger,2002).Dataarerandomly
drawnwithreplacementfromthe
observeddatasettocreateanew
datasetofthesamesize.Tispro-
cedureisrepeatedaspecifednum-
beroftimes(1000forthecurrent
study),andtheindirecteffectis
estimatedfromeachofthedatas-
ets,yieldingaconfdenceinterval.
An advantage of the bootstrap
methodisthatitdoesnotassume
a normally distributed param-
eterestimate,whichisimportant
given that distributions of indi-
rect effect estimates repeatedly
havebeenfoundtobepositively
skewed(Mackinnon,Lockwood,
Hoffman, West, & Sheets, 2002;
Shrout&Bolger,2002).Assuch,
when assuming normality, the
resulting confidence intervals
produceasymmetricerrorrates,
therebydecreasingthepowernec-
essarytodetecttheindirectefect
(MacKinnonetal.,2002).Because
bootstrapmethodologydoesnot
assumeanormaldistribution,it
isamorepowerfultestthantra-
ditional multi-step approaches,
andisrecommendedforsmallto
moderatesamplesizes(Shrout&
Bolger,2002).
Unconditional Latent Growth curve
Model of PsE
Goal 1: Change in PSE over chil-
drens ages 2, 3, and 4.First,inorder
tomodelchangeinPSEovertime,anunconditional
(i.e.,nocovariateswereincludedinthemodel)LGCM
wasftusingMplusVersion4.0(Muthn,&Muthn,
1998).Allftindicesindicatedthatthemodelwasa
goodfttothedata,2(2)=4.93,p>.05;CFI=0.99;
TLI=0.99,RMSEA=0.04;SRMR=0.03.Parameter
estimatesoftheunconditionalLGCMsuggestedthat
averageinitiallevelsofPSEweresignifcantlydiferent
fromzero(b=30.40,SE=0.17,p<.01);however,this
isnotparticularlymeaningfulgiventherewasnozero
Figure 1.Latentgrowthcurvemodelofparentingself-efcacy(PSE)fromages2
to4,maternaldepressionatage3,andmaternal-reportedchildrensproblembe-
haviorsatage4(showingunstandardizedpathcoefcients;mediationalmodelin
boldforemphasis).
Figure 2.Latentgrowthcurvemodelofparentingself-efcacy(PSE)fromages2to
4,maternaldepressionatage3,andalternatecaregiver-reportedchildrensprob-
lem behaviors at age 4 (showing unstandardized path coefcients; mediational
modelinboldforemphasis).
8 MANUSCRIPT UNDER REVIEW: Infant Behavior and Development
Unauthorized reproduction of this article is prohibited.
WEAVER,SHAW,DISHION,WILSON
pointonthePSEscalethatwasadministered.More
importantly,therewassignifcantindividualvariabil-
ityaroundtheintercept(b=13.08,SE=0.95,p<.01),
meaningthatindividualssignifcantlydeviatedfrom
averagelevelsofPSEatage2.Intermsofgrowth,
PSEsignifcantlyincreasedbetweenages2and4(b
=0.40,SE=0.08,p<.01).Terewasnotsignifcant
individualvariationaroundtheaverageslope.Asa
result,predictorsandoutcomeswereonlyextendedto
theinterceptparameterforallconditionalLGCMs.
conditional Latent Growth curve Models of PsE
Goals 2 and 3: Relationships between PSE, childrens
problem behaviors, and maternal depression.Separate
conditionallatentgrowthcurvemodelswereftfor
maternalversusACreportsofchildrensproblem
behavior.Parameterestimatesofbothmodelsarepre-
sentedinFigures1(usingmaternalreportsofproblem
behavior)and2(usingACreportsofproblembehav-
ior).Inbothmodels,errortermswereallowedtocor-
relatebetweeneachmanifestindicatorofchildrens
behaviorproblemsanditscounterpartacrossages
2and4toaccountforsharedmethodvariance.For
example,theage2externalizingerrortermwascorre-
latedwiththeage4externalizingerrorterm.Further-
more,errorvariancesbetweentheEybergintensity
scoreandtheEybergproblemscorewereallowedto
correlatewithintimepointatage2andatage4to
accountforsharedmethodvariance.Forparsimony,
errorvarianceswerenotrepresentedinthemodels.
IntermsoftheLGCMusingmaternalreportsof
childrensbehaviorproblems,allftindicessuggested
thatthemodelwasagoodft,2(54)=109.19,p=.00;
2/df=2.02,CFI=0.98;TLI=0.97,RMSEA=0.04;
SRMR=0.04.AsshowninFigure2,lowerlevelsof
PSEatage2(i.e.intercept)predictedhigherlevelsof
maternal-reportedchildrensproblembehavioratage
4(b=-0.35,SE=0.10,p<.01)afercontrollingforthe
signifcantefectsofage2levelsofproblembehavior
onbothinitiallevelsofPSE(b=-0.25,SE=0.04,p<
.01)andage4problembehavior(b=0.70,SE=0.09,
p<.01),aswellasmaternaleducationalattainment
onPSE(b=-0.47,SE=0.14,p<.01)andsignifcant
treatmentefectsonbothmaternaldepression(b=
-1.71,SE=0.81,p<.05)andchildrensbehaviorprob-
lems(b=-1.77,SE=0.55,p<.01).Sitediferences
werealsocovariedinthemodel;however,efectswere
non-signifcant.
Signifcantrelationshipsbetweeninitiallevelsof
PSEandmaternaldepressionatage3andbetween
age-3depressionandchildrensbehaviorproblemsat
age4wereintheexpecteddirection(b=-.89,SE=
0.14,p<.01andb=0.09,SE=0.03,p<.01,respec-
tively).Intermsofthemediatingefectofmaternal
depressionatage3,theconfidenceintervalofthe
indirectefectbasedon1000bootstrapsamplesdid
notincludezero(95%CI:-0.13to-0.02),suggest-
ingsignifcantmediationofmaternaldepressionon
therelationshipbetweenPSEatage2andmaternal-
reportedchildrensproblembehavioratage4.Next,
becausethebootstrappingtechniqueisrecommended
whenfttingmodelswithmoderatesamplesizesand
theSobeltestrequiresmorepowertodetectefects
(Mackinnonetal.,2002),mediationalsowastested
usingthemorerigorous,largesample,Sobelmethod
(Sobel,1982).ResultsusingtheSobeltestwerealso
signifcant(z=-2.69,p<.01),furthersupportingthe
mediatingroleofmaternaldepression.
BecauseofthepotentialreporterbiasintheLGCM
usingonlymaternalreports,theidenticalmodelwas
fit using AC reports of childrens problem behav-
ioratages2and4.Allftindicessuggestedthatthe
modelwasaverygoodfttotheobserveddata,2
(54)=87.53,p=.00;2/df=1.62,CFI=0.97;TLI=
0.96,RMSEA=0.04;SRMR=0.04.Resultsshowed
similarpatternsofefectswithasignifcantnegative
relationshipbetweenPSEatage2andchildrensprob-
lembehaviorsatage4(b=-0.34,SE=0.15,p<.01).
Again,signifcantrelationshipsbetweeninitiallevels
ofPSEandmaternaldepressionatage3andbetween
age-3maternaldepressionandchildrensbehavior
problemsatage4wereintheexpecteddirection(b
=-.96,SE=0.25,p<.01andb=0.09,SE=0.04,p
<.05,respectively).Ascomparedtothemodelwith
onlymaternalreports,themodelwithAC-reported
childrensproblembehaviorshowednonsignifcant
pathwaysbetweenmaternallevelofeducationandini-
tialPSEandbetweenage2problembehaviorandage
3maternaldepression.Terewerealsononsignifcant
treatmentefectsonmaternaldepressionandage4
problembehavior.
Withrespecttomediation,theconfdenceinter-
valoftheindirectefectofdepressionbasedon1000
bootstrapsamplesdidnotincludezero(95%CI:-0.17
to-0.01),providingfurtherconfrmationofthemedi-
atingroleofmaternaldepression.Resultsusingthe
Sobeltestoftheindirectefectwerealsosignifcant
fortheAC-reportmodel(z=-2.13,p<.01).
Discussion
Three primary goals were pursued in the current
study:(1)toexaminethestabilityanddevelopmental
courseofPSEovertime,(2)toexaminelongitudinal
associationsbetweenPSEandchildrensbehavior
problems,(3)andtotestthemediatingroleofmater-
naldepressioninreferencetoassociationsbetween
PSEandchildrenslaterbehaviorproblems.Interms
ofthefrstgoal,consistentwithourexpectationsand
thelimitedavailableextantliterature(Grossetal.,
1994;Zayasetal.,2005)aswellaslogicalexpectations
MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 9
Unauthorized reproduction of this article is prohibited.
PARENTINGSELF-EFFICACY
frommothersgainingmoreexperienceintheparental
role,wefoundthatPSEsignifcantlyincreasedfrom
thetoddlertopreschoolperiod.Furthermore,individ-
ualmotherslevelsofPSEdidnotsignifcantlydevi-
atefromtheaveragetrajectoryoftheentiresample,
suggestingthatmothersvariedlittleintheirrateof
changefromages2to4.Intermsofthesecondgoal,
aspredicted,lowerlevelsofPSEatage2(i.e.,intercept
parameter)predictedhigherratesofchildrensprob-
lembehavioratage4evenafercontrollingforini-
tiallevelsofconductproblems.Tisrelationshipwas
foundusingbothmaternalandACreportsofproblem
behavior.Importantly,ACreportsofchildconduct
problemsprovidedcorroboratingevidenceofmater-
nalreportsofassociationsbetweenPSEandlaterchild
adjustment,acommonlimitationofpriorresearch
inthisarea.Last,usingbootstrapmethodology,we
foundthatmaternaldepressionatage3mediatedthe
relationshipbetweenage2PSEandage4childrens
problembehaviorasreportedbybothmothersand
ACs.Tisresult,alsocorroboratedbytwoinformants,
isconsistentwiththenotionthatmaternaldepression
representsonefactorthatmayunderlielowPSEand
accountforassociationsbetweenPSEandchildcon-
ductproblems.Itisalsoworthnotingthatmediation
efectswerecorroboratedusingtheSobeltest,which
requiresmorepowertodetectasignifcantefect.
These findings extend the current literature by
assessing the course and stability of PSE at three
timepointsoverthecourseoftwoyears.Asprevi-
ouslydiscussed,moststudiesexaminingPSEhavenot
beenlongitudinalindesignorwerelimitedbytheuse
oftwotimepoints(e.g.,Zayasetal.,2005).Byusing
threeassessmenttimepoints,wewereabletomodel
latent growth parameters to describe the average
growthcurveofthesample,whichbothcorroborated
andextendedpreviousfndingsregardingincreasesin
PSEduringearlychildhood(Grossetal.,1994;Zayas
etal.,2002).Moreover,thecurrentsamplewasethni-
callydiverseandlow-income,furtherextendingthe
extantresearchonPSE,whichhastypicallyfocused
onmajority,middleclasssamples(e.g.,Colemanetal.,
2002;Leerkes&Crockenberg,2002;Teti&Gelfand,
1991).
InadditiontomodelingPSEovertime,thecur-
rentstudyaddstotheliteraturebydemonstrating
thatage2levelsofPSEpredictedalatentconstruct
ofage-4childrensproblembehavior,usingfactors
fromtwopreviouslyvalidatedinstruments(i.e.,CBCL
andEybergInventory).Importantly,theseassocia-
tionsremainedevidentevenafercontrollingforage
2levelsofchildconductproblems.Veryfewstudies
havelinkedPSEtochildrensbehavioraladjustment
(seeBor&Sanders,2004;Sanders&Wooley,2004for
notableexceptions),muchlessoveratwo-yearperiod
andcontrollingforchildrensbehavioraladjustment.
Inadditiontocontrollingforinitiallevelsofproblem
behavior,theefectofage2levelsofproblembehavior
oninitiallevelsofPSEwerealsocovaried.Tis,in
essence,controlledforchildefectsintheestimation
ofage2PSE,whichisimportantgiventhesample
consistedofchildrenalreadyshowingelevatedrates
ofbehavioralmaladjustmentuponrecruitment.
Tecurrentstudyalsoextendedpreviousresearch
examiningassociationsbetweenPSEandchildadjust-
mentbyinvestigatingthepotentialmediatingfunc-
tionofmaternaldepression.Teresultsareconsis-
tentwiththehypothesisthatnegativecognitionsthat
characterizedepressionlikelyprovidetheframework
throughwhichlowPSE,orfeelingsofincompetence
intheparentingrole,directlyafectsthesocioemo-
tionaladjustmentofchildren.Inotherwords,depres-
sivecharacteristicssuchaswithdrawal,sadafect,and
worthlessnesslikelyprovideavehiclethroughwhich
PSEelicitschildrensdisruptivebehavior.
Intermsofatheoreticalexplanationaccounting
forthemediatingroleofmaternaldepression,learned
helplessness(Abramson,Seligman,&Teasdale,1978)
representsonepotentialintrapersonalmechanism.
Tetheoryoflearnedhelplessnesspositsthatwhen
confrontedwithanegativeevent,individualswho
attributepooroutcomestointernal,stable,and/or
global factors are more likely to have depressive
responsesthanthosewhoattributenegativeoutcomes
toexternal,unstable,orspecifcfactors.Intermsof
PSEinthecontextofdepression,foramotherhighon
learnedhelplessness,lowPSEmaybeduetoattribu-
tionsthatherchildsbehaviorisuncontrollable.As
aresult,childrensnegativebehaviorswouldlikely
worsen,withmaternalfeelingsofincompetencerein-
forced.Moreover,lowerPSEmaypersistasdepressed
mothersbecomefocusedonthechallengesofsynchro-
nizingparentingtechniqueswithchildrenschang-
ingdevelopmentalneeds,especiallyamongchildren
showingearlymarkersofbehavioralmaladjustment.
Asaresult,childrensconductproblemsmaygener-
ateaviciousintrapersonalcycle,learnedhelplessness
fuelinglowPSE,whichinturnreinforcesfeelingsof
learnedhelplessness.
Tefndingsofthecurrentstudyhaveimplications
forpreventionprogramstargetingchildrensconduct
problems.Specifcally,giventheassociationsfound
betweenPSEandbothmaternalandACreportsof
laterchildconductproblems,thefindingssuggest
thatpreventionprogramsbeaimedatincreasingPSE
amongmothersduringearlytoddlerhood.Although
thecurrentresultsarelimitedbythestudyscorrela-
tionaldesign,limitinginferencesaboutcausality,they
dosuggestthatassociationsbetweenPSEandchild
conductproblemsareevidentfortoddlersathigh
10 MANUSCRIPT UNDER REVIEW: Infant Behavior and Development
Unauthorized reproduction of this article is prohibited.
WEAVER,SHAW,DISHION,WILSON
riskforshowingearly-startingconductproblems.
TargetingPSEinfuturepreventiontrialswouldthus
bealogicalnextsteptotestwhetherincreasingPSE
wouldresultinimprovementsinchildbehavior.Tat
maternaldepressionmediatedassociationsbetween
PSEandchildconductproblemssuggestsmaternal
depressionisanothertargetforimprovingbothPSE
andchildbehavior.Itislikelythatimprovementsin
motherscognitionsandattributionsmightleadto
improvementsinPSE.Assuccessfulpsychological
(e.g.,cognitive-behavioraltherapy)andpharmaco-
therapeutic(SSRIs)methodsfortreatingdepression
alreadyexist,intheshorttermitmightbeeasierto
testtheseideasbytargetingmaternaldepression.
Animportantlimitationofthecurrentstudythat
warrantscarefulconsiderationismethodbias.Allof
themainstudyvariableswerequestionnairereports,
andinthemodelusingallmaternalreportsofchil-
drens problem behavior, all of the variables were
fromthesamereporter.ByusingtheACreportsof
childrens behavioral maladjustment, we direct-
ly addressed reporter bias; however, method bias
remainsproblematic.Itshouldbenotedthattwoof
thethreeconstructsofinterestinthecurrentstudy
(i.e.,PSEanddepressivesymptomatology)areopti-
mallymeasuredusingselfreportsastheyrepresent
internalfeelingsandcognitions.Intermsofassessing
childrensproblembehavior,futureresearchshould
usemultiplemethodsofdatacollectionincluding
directobservationsofchildrenengagedwithseveral
salientcaretakers,andideally,acrossvariousenvi-
ronmentalcontexts.
FuturedirectionsinthestudyofPSEshouldalso
examinelongitudinaltransactionalprocessesbetween
bothPSEandchildrensproblembehavioraswellas
PSEandmaternaldepression.First,aswehaveshown
thatmotherscognitionsandbehaviorsimpacttheir
children,itislikelythatchildrensbehaviorsimpact
mothersdevelopmentaltrajectoriesofPSE.Forexam-
ple,achildmayrespondwithrule-breakingbehavior
toelicitaresponsefromamotherwhoiswithdrawn.
Techildsbehaviormayreinforcematernalfeelings
ofincompetenceandtheinabilitytoexerciseefective
parentalmanagementofherchild.Inturn,lowlevels
ofPSElikelyperpetuateanegativecyclebetweenthe
mother-childdyad.Similartransactional,intraper-
sonalprocesseslikelytakeplacebetweenPSEand
maternaldepression.Advancedlongitudinalstatisti-
caltechniquessuchasparallelprocessesandlatent
difference score models can be used to shed light
onthesecomplexdynamics,providinginformation
regardingtherelativeimportanceofeachconstruct
acrossdevelopment.Asaresult,morepreciseand
efectivepreventionprograms,intermsofdevelop-
mentaltimingandapotentialchangingfocusoftar-
getedconstructs,maybeplannedandimplemented
withmothersofchildrenathighriskfordeveloping
conductproblems.
references
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D.
(1978). Learned helplessness in humans: Critique and
reformulation. Journal of Abnormal Psychology, 87, 49-
74.
Achenbach, T. M., & Rescorla, L. A. (2000). Manual for the
ASEBA preschool forms & profles. Burlington VT: Uni-
versity of Vermont Department of Psychiatry.
Bandura, A. (1977). Self-efcacy: Toward a unifying theory
of behavioral change. Psychological Review, 84, 191-215.
Bandura, A. (1989). Regulation of cognitive processes
through perceived self-efcacy. Developmental Psychol-
ogy, 25, 729-735.
Bentler, P. M. (1990). Fit indexes, Lagrange multipliers,
constraint changes and incomplete data in structural
models. Multivariate Behavioral Research, 25, 163-172.
Bollen, K. A. (1989). Structural equations with latent vari-
ables. New York, NY: John Wiley & Sons.
Bor, W., & Sanders, M. R. (2004). Correlates of self-report-
ed coercive parenting of preschool-aged children at high
risk for the development of conduct problems. Australian
and New Zealand Journal of Psychiatry, 38, 738-745.
Brown, M. W. & Cudeck, R. (1993). Testing structural
equation models: Alternative ways of assessing model ft.
Newbury Park, CA, Sage: 136-262.
Campbell, S. B., Shaw, D. S., & Gilliom, M. (2000). Early
externalizing behavior problems: Toddlers and pre-
schoolers at risk for later maladjustment. Development
and Psychopathology, 12, 467-488.
Coleman, P. K., & Karraker, K. H. (2000). Parenting self-
efcacy among mothers of school-age children: Concep-
tualization, measurement, and correlates. Family Rela-
tions, 49, 13-24.
Coleman, P. K., & Karraker, K. H. (2003). Maternal self-ef-
fcacy beliefs, competence in parenting, and toddlers be-
havior and developmental status. Infant Mental Health
Journal, 24, 126-148.
Coleman, P. K., Trent, A., Bryan, S., King, B., Rogers, N.,
& Nazir, M. (2002). Parenting behavior, mothers self-
efcacy beliefs, and toddler performance on the Bayley
Scales of Infant Development. Early Child Development
and Care, 172, 123-140.
Cummings, E. M., Keller, P. S., & Davies, P. T. (2005). To-
wards a family process model of maternal and paternal
depressive symptoms: Exploring multiple relations with
child and family functioning. Journal of Child Psychol-
ogy and Psychiatry, 46, 479-489.
Dishion, T. J., & McMahon, R. J. (1998). Parental monitor-
ing and the prevention of child and adolescent problem
behavior: A conceptual model and empirical formula-
tion. Clinical Child and Family Psychology Review, 1,
61-75.
Eaton, W.W., & Kesslerm, L.G. (1981). Rates of symptoms
of depression in a national sample. American Journal of
Epidemiology, 114, 528-538.
MANUSCRIPT UNDER REVIEW: Infant Behavior and Development 11
Unauthorized reproduction of this article is prohibited.
PARENTINGSELF-EFFICACY
Farmer, A., McGufn, P., & Williams, J. (2002). Measuring
psychopathology. Oxford: Oxford University Press.
Fergusson, D. M., Lynskey, M. T., & Horwood, J. L. (1993).
Te efect of maternal depression on maternal ratings of
child behavior. Journal of Abnormal Psychology, 21, 245-
269.
Gardner, F. E. (1987) Positive interaction between mothers
and conduct-problem children: Is there training for har-
mony as well as fghting? Journal of Abnormal Psychol-
ogy, 15, 283-293.
Gardner, F., Ward, S., Burton, J., & Wilson, C. (2003). Te
role of mother-child joint play in the early development
of childrens conduct problems: A longitudinal observa-
tional study. Social Development, 12, 361-378.
Gross, D., Conrad, B., Fogg, L., & Wothke, W. (1994). A
longitudinal model of maternal self-efcacy, depression,
and difcult temperament during toddlerhood. Research
in Nursing and Health, 17, 207-215.
Haslam, D. M., Pakenham, K. I., & Smith, A. (2006). Social
support and postpartum depressive symptomatology:
Te mediating role of maternal self-efcacy. Infant Men-
tal Health Journal, 27, 276-291.
Jerusalem, M., & Mittag, W. (1995). Self-efcacy in stress-
ful life transitions. In A. Bandura (Ed.) Self-efcacy in
changing societies (pp.177-201). New York: Cambridge
University Press.
Johnston, C., & Mash, E. J. (1989). A measure of parenting
satisfaction and efcacy. Journal of Clinical Child Psy-
chology, 18, 167-175.
Leerkes, E. M., & Crockenberg, S. C. (2002). Te develop-
ment of maternal self-efcacy and its impact on mater-
nal behavior. Infancy, 3, 227-247.
Maciejewski, P. K., Prigerson, H. G., & Mazure, C. M.
(2000). Self-efcacy as a mediator between stressful life
events and depressive symptoms: Diferences based on
history of prior depression. British Journal of Psychiatry,
176, 373-378.
Mackinnon, D. P., Lockwood, C. M., Hofman, J. M., West,
S. G., & Sheets, V. (2002). A comparison of methods to
test mediation and other intervening variable efects.
Psychological Methods, 7, 83-104.
McArdle, J., & Bell, R. (2000). An introduction to latent
growth curve modeling for developmental data analysis.
In T. Little, K. Schnabel, and J. Baumert (Eds.). Mod-
eling Longitudinal and Multilevel Data. Mahwah, NJ:
Lawrence Erlbaum Associates.
Myers, J., & Weissman, M. (1980). Use of a self-report symp-
tom scales to detect depression in a community sample.
American Journal of Psychiatry, 137, 1081-1084.
Muthn, L. K. & Muthn, B. (1998). M-Plus users guide.
Los Angeles, CA: Authors.
Owens, E. B., & Shaw, D. S. (2003). Predicting growth curves
of externalizing behavior across the preschool years. Jour-
nal of Abnormal Child Psychology, 31, 575-590.
Ozer, E. M., & Bandura, A. (1990). Mechanisms governing
empowerment efects: A self-efcacy analysis. Journal of
Personality and Social Psychology, 58, 472-486.
Radlof, L.S. (1977). Te CES-D scale: A self-report depres-
sion scale for research in the general population. Applied
Psychological Measurement, 1, 385-401.
Raver, C. C., & Leadbeater, B. J. (1999). Mothering under
pressure: Environmental, child, and dyadic correlates of
maternal self-efcacy among low-income women. Jour-
nal of Family Psychology, 4, 523-534.
Robinson, E. A., Eyberg, S. M., & Ross, W. A. (1980). Te
standardization of an inventory of child conduct prob-
lem behaviors. Journal of Clinical Child Psychology, 9,
22-29.
Sanders, M. R., & Woolley, M. L. (2005). T erelationship
between maternal self-efcacy and parenting practices:
Implications for parent training. Child: Care, Health &
Development, 31, 65-73.
Shaw, D. S., & Bell, R. Q. (1993). Developmental theories of
parental contributors to
antisocial behavior. Journal of Abnormal Child Psychology,
21, 493-518.
Shaw, D.S., Bell, R.Q., & Gilliom, M. (2000). A truly early
starter model of antisocial behavior revisited. Clinical
Child and Family Psychology Review, 3, 155-172.
Shaw, D. S., Keenan, K., & Vondra, J. I. (1994). Te devel-
opmental precursors of antisocial behavior: Ages 1-3.
Developmental Psychology, 30, 355-364.
Shaw, D. S., Wilson, M. N., Gardner, F, & Dishion, T. J.
(2007, March). Te use of motivational feedback in a
family-based intervention to prevent early conduct prob-
lems. In F. Gardner (Chair), Using systematic feedback
to guide contextually-based interventions to prevent
problem behavior across settings and development. Pa-
per presented at the 2007 Society for Research in Child
Development Biennial Meeting, Boston, MA.
Shrout, P. E. & Bolger, N. (2002). Mediation in experimen-
tal and nonexperimental studies: New
procedures and recommendations. Psychological Meth-
ods, 7, 422-445.
Sobel, M. E. (1982). Asymptotic confdence intervals for in-
direct efects in structural equation models. In S. Lein-
hardt (Ed.), Sociological methodology (pp. 290-312). San
Francisco, CA: Jossey-Bass.
Teti, D. M., & Gelfand, D. M. (1991). Behavioral compe-
tence among mothers of infants in the frst years: Te
mediational role of maternal self-efcacy. Child Devel-
opment, 62, 918-929.
Tucker, L. R., & Lewis, C. (1973). A reliability coefcient
for maximum likelihood factor analysis. Psychometrika,
38, 1-10.
Webster-Stratton, C. (1985). Comparisons of behavior
transactions between conduct-disordered children and
their mothers in the clinic and at home. Journal of Ab-
normal Child Psychology, 13, 169183.
Wetzler, S., & van Praag, H.M. (1989). Assessment of de-
pression. In S. Wetzler (Ed.), Measuring mental illness:
Psychometric assessment for clinicians (pp. 71-88).
Washington, DC: American Psychiatric Association.
Widaman, K. F., & Tompson, J. S. (2003). On specifying
the null model for incremental ft indices in structural
equation modeling. Psychological Methods, 8, 16-37.
Zayas, L. H., Jankowski, K. R. B., & Mckee, M. D. (2005).
Parenting competence across pregnancy and postpartum
among urban minority women. Journal of Adult Devel-
opment, 12, 53-62.