University of Southern California Dissertations and Theses

Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

(USC Thesis Other)

Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INTERGENERATIONALSUPPORTBETWEENGRANDPARENTSAND
GRANDCHILDRENINRURALCHINAANDITSEFFECTON
THEPSYCHOLOGICALWELL-BEINGOFOLDERADULTS
by
LingXu
ADissertationPresentedtothe
FACULTYOFTHEUSCGRADUATESCHOOL
UNIVERSITYOFSOUTHERNCALIFORNIA
InPartialFulfillmentofthe
RequirementsfortheDegree
DOCTOROFPHILOSOPHY
(SOCIALWORK)
August2013
Copyright2013 LingXu
ii
Acknowledgments
Thisdissertationwouldnotbepossiblewithoutthesupportfrommanypeople.First,I
wouldliketogivespecialthankstoDr.IrisChi,myadvisor,forherremarkablementorship.She
isanamazingscholar,andherenthusiasmandextensiveknowledgeregardingresearchinvolving
olderChineseadultshasinspiredmyinterestingerontologyresearch.Shealwaysprioritizesthe
needsofherstudentsandwasalwayssupportivewhenIneededadviceonmydissertationor
otheracademicprojects.Shesetahighacademicstandardandencouragedandpushedmeto
reachthatstandard.DuringmysixyearsatUSC,Dr.Chiwaslikebothamentorandamother,
nurturingmeandwatchingmegrowintellectuallyandsocially.Ourrelationshiphasgone
beyondatypicalteacher–studentrelationshipandwilllastforalifetime.
Iwanttothankmycommitteemembers,Dr.MerrilSilverstein,Dr.LawrencePalinkas,
andDr.Chih-PingChou,fortheirsteadfastsupport.Inadditiontoguidingmetofinishmy
dissertation,theygavemeeverypossibleformofhelpandsupportthroughoutmyexperiencein
thePhDprogram.TheyweremorethanhappytowritedifferentreferenceletterswheneverI
neededthem,andIwillneverforgetourmanydiscussionsofimportantissues.Ifeelsoluckyto
havethemasmembersofmydissertationcommitteeandamdeeplygratefulfortheirsupport.
Iownagreatdealofthankstomyhusband,AnshiJin,forhisunreservedloveand
support.Icanrecallmanynightsandweekendswhenhehandledthehouseworkandbabysitting
whileIfocusedonmyresearchprojects.HeconsoledandencouragedmewhenIstruggledwith
academicprojectsorpapers.Hehasalwaysstoodbehindmeandofferedalltheemotionaland
financialsupportIneeded.
VeryspecialthankstomyfriendsYawenLi,MayGuo,WeiyuMao,andRongXiaofor
thesupporttheyhavegivenmeoveralltheseyears.Wenotonlystudiedandrelaxedtogether,
iii
butwealsosharedourhappinessandsadness.Theyhavebeenstrongsourcesofsupport
throughoutdifficulttimesinmygraduatecareer.Iwouldalsoliketothankmycohortfriends,
Dahlia,Rohini,Ian,Min-Kyoung,andHsin-Yi,fortheirfriendshipandsupport.Thehappytimes
andactivitieswesharedhelpedmesurvivemanytoughtimesduringthepastfiveyears.
Iamalsoverygratefultomymotherandmyparents-in-lawfortheireverlastingloveand
support.TheycametotheUnitedStatesforsixmonthseachtohelpcareformychildrensoI
wouldhavetimetofocusonmyacademicwork.Withouttheirassistanceandsacrifice,Iwould
nothavebeenabletoachievesomuchinmyacademiccareer.Ialsowanttothankmytwo
lovelylittleboys.Theygavemethemotivationtomoveaheadmanytimes.Theiradorablefaces
andlovemakemeforgetthedifficultiesofacademicworkandhelpmeovercomeexhaustion
fromwork.ThankyouandIloveyou.
Last,Iwanttothankallthepeoplewhosupportedmeandprovidedassistancethroughout
thePhDprogram.Amongthesepeople,specialthanksaregiventoEricLindberg,whogaveme
alotofsupportineditingmymanuscriptsanddissertation,andMalindaSampson,forher
emotionalandfinancialsupport.
iv
TableofContents
Acknowledgments...........................................................................................................................ii
ListofTables..................................................................................................................................vi
ListofFigures................................................................................................................................vii
Abstract.........................................................................................................................................viii
Chapter1:IntroductionandTheoreticalFramework.......................................................................1
Background................................................................................................................................1
TheoreticalFramework..............................................................................................................4
RuralChinaasaResearchSetting.............................................................................................8
PurposesoftheStudy...............................................................................................................10
Chapter2:DescriptionoftheDataset............................................................................................13
SurveySite...............................................................................................................................13
DataSampleandCollection.....................................................................................................15
ContentsofQuestionnaires......................................................................................................19
Qualityofthedataset...............................................................................................................20
AdvantagesofUsingAnhuiData............................................................................................21
Chapter3:DeterminantsofSupportExchangebetweenGrandparentsandGrandchildren:
TheRolesofGrandparentCaregiving,PatrilinealHeritage,andEmotionalBonds.....................23
Introduction..............................................................................................................................23
IntergenerationalSupportExchangebetweenGrandparentsandGrandchildren..............23
DeterminantsofSupportExchangebetweenGrandparentsandGrandchildren...............25
ThePresentStudy..............................................................................................................29
Methods....................................................................................................................................30
Data....................................................................................................................................30
Measurement......................................................................................................................31
DataAnalysis.....................................................................................................................34
Results......................................................................................................................................35
CharacteristicsofGrandparentsandGrandchildren..........................................................35
Grandparent-GrandchildSupportExchange......................................................................37
DeterminantsofGrandparent-GrandchildSupportExchange...........................................38
Discussion................................................................................................................................39
Chapter4:ChangePatternsofInstrumentalSupportfromAdultChildrenand
GrandchildrenamongRuralChineseGrandparents......................................................................44
Introduction..............................................................................................................................44
InstrumentalSupportfromYoungerGenerations.............................................................46
ThePresentStudy..............................................................................................................48
v
Methods....................................................................................................................................48
Data....................................................................................................................................48
Measurement......................................................................................................................49
DataAnalysis.....................................................................................................................51
Results......................................................................................................................................53
ChangePatternsofInstrumentalSupportfromTwoYoungerGenerations......................53
PredictorsofInstrumentalSupportChangePatterns.........................................................55
Discussion................................................................................................................................57
Limitations.........................................................................................................................60
Conclusions........................................................................................................................62
Chapter5:TheDynamicRelationshipbetweenInstrumentalSupportfromGrandchildren
andDepressionamongRuralChineseGrandparents:ALongitudinalStudy................................63
Introduction..............................................................................................................................63
InstrumentalSupportandDepression................................................................................64
ThePresentStudy..............................................................................................................67
Methods....................................................................................................................................68
Data....................................................................................................................................68
Measurement......................................................................................................................69
DataAnalysis.....................................................................................................................71
Results......................................................................................................................................75
SampleCharacteristicsatBaseline....................................................................................75
BivariateLatentGrowthCurveAnalysis...........................................................................77
BivariateLatentDifferenceScoreAnalysis......................................................................79
Discussion................................................................................................................................81
Limitations.........................................................................................................................83
ImplicationsforPolicyandPractice..................................................................................84
Conclusions........................................................................................................................85
Chapter6:ConclusionsandImplications......................................................................................86
SummaryofMajorResearchFindings....................................................................................87
ContributionsofthisDissertation............................................................................................90
ImplicationsforFutureResearch.............................................................................................92
ImplicationsforPoliciesandPractice......................................................................................95
References......................................................................................................................................98
vi
ListofTables
Table2.1.Cronbach’sAlphaReliabilityScoresforSomeVariablesintheAnhuiStudy............21
Table3.1.CharacteristicsofGrandparents(N=926)...................................................................35
Table3.2.CharacteristicsofGrandchildren(N=1,843)...............................................................36
Table3.3.SupportExchangebetweenGrandparentsandGrandchildren....................................37
Table3.4.DeterminantsofSupportExchangebetweenGrandparentandGrandchild.................39
Table4.1.SampleCharacteristicsofOlderAdultsinRuralChinaatBaseline(N=731)..............49
Table4.2.DistributionsofChangofSupportfromAdultChildrenandGrandchildrenover9
YearsinRuralChinabyPercentage(N=2,248)...........................................................................50
Table4.3.ModelFitforLatentClassAnalysisofInstrumentalSupportamongRuralOlder
ChineseAdults...............................................................................................................................53
Table4.4.ConditionalLatentClassProbabilitiesofaThree-ClassModelofIntergenerational
SupportPatternsinRuralChina....................................................................................................54
Table4.5.MultinomialOddsRatiosPredictingClassMembershipinaPartiallyConstrained
Three-ClassModelofIntergenerationalSupportPatternsinRuralChina....................................56
Table5.1.SampleCharacteristicsatBaseline(N=731)...............................................................76
Table5.2.MeansandStandardDeviationsofAge,ProvisionofSupporttoGrandchildren,and
DepressionOverTime..................................................................................................................76
Table5.3.BivariateLatentGrowthCurveModel:EstimatesandCoefficientswithoutControl
Variables........................................................................................................................................78
Table5.4.BivariateLatentDifferenceScoreAnalyses:EstimatesandCoefficientswithout
ControlVariables...........................................................................................................................79
vii
ListofFigures
Figure1.1.FamilySystemsofGrandparenting...............................................................................6
Figure2.1.SampleSizeandAttritionintheAnhuiStudyinChina,2001–2009..........................18
Figure5.1.LatentGrowthCurveModelwithoutControlVariables............................................72
Figure5.2.LatentDifferenceScoreModelwithoutControlVariables........................................74
viii
Abstract
Socialsupportisespeciallyimportantforpeopleinlatelifeduetophysicalandlife-course
changes.ThemajorityofolderadultsinChinareceivesocialsupportprimarilyfromtheiradult
children.However,formanyruralresidents,supportingolderadultshasnotbeeneasydueto
lowerstandardsoflivingandmigrationtourbancities.Ontheotherhand,duetoincreased
longevity,alargeproportionofolderadultshaveadultgrandchildrenwhomayalsoserveasan
importantsourceofsupport.However,littleattentionhasbeenpaidtothedynamicsof
grandchildsupportamongolderadultsinruralChina,despitedemographicshiftsthatindicate
grandparentsandgrandchildrenareleftbehindandlivewitheachotherinruralChina.This
dissertationexaminedhowadultgrandchildrenprovidefamilysupporttooldergrandparentsand
howthissupportaffectsthepsychologicalwell-beingofolderadultsinruralChina.
Thisdissertationconsistsofthreeindependentstudies.Itsmainpurposewastoaddtothe
existingknowledgeregardingintergenerationalsupportbyinvestigatingthedynamicsofsupport
providedtoolderadultsbytheirgrandchildreninruralChina.Usingalongitudinaldatasetfrom
agroupofolderadultsinaruralregionoftheAnhuiprovinceinmainlandChina,this
dissertationhadthefollowingspecificaims:(1)toinvestigatesupportexchangebetween
grandparentsandgrandchildrenandassociatedfactors;(2)toexploretheunderlyingchange
patternsofinstrumentalsupportthatolderparticipantsreceivedfromadultchildrenand
grandchildrenduringthepastnineyears;and(3)toexaminehowsupportfromgrandchildren
affectedlevelsofdepressionamongolderadults.
Theresultsofthisdissertationidentifiedbothsimilaranddifferentpatternsof
intergenerationalsupportbetweengrandparentsandgrandchildreninruralChinacomparedwith
ix
thoseinWesternsocieties.Thefirststudyrevealedthatgenerallymoregrandparentsreceived
supportfromtheirgrandchildrenthanthosewhoprovidedsupporttotheirgrandchildren,
includinginstrumental,financial,andemotionalsupport.Ahigherpercentageofolderadults
exchangedsupportwiththegrandchildrenoftheireldestchildifthatchildwasmaleratherthan
female.Thisstudyalsofoundthatolderadultswhohadstrongemotionalbondswiththemiddle
generation,especiallywithsons,andpastgrandchildcaregivingexperienceweremuchmore
likelytoreceiveandprovidesupporttotheirgrandchildren.Thesecondstudyrevealedthree
underlyingchangepatternsofinstrumentalsupportfromadultchildrenandgrandchildren:a
decreaseinsupport,anincreaseinsupport,andminorlevelsofsupport.Thethirdstudyindicated
thatanincreaseinsupportfromgrandchildrenwasnegativelyassociatedwithanincreasein
depression.However,depressionlevelsamongolderparticipantsplayedaleadingrolein
predictingsupportfromgrandchildren,ratherthanviceversa,thusindicatingthesignificant
influenceofmentalhealthconditionsinlaterlife.
Altogether,thefindingsofthisdissertationrevealedthatgrandchildrenareimportant
sourcesofsupportorsocialcapitalforolderadultsinruralChinaandsuchsupporthasbeneficial
effectsonthepsychologicalwell-beingofolderadults.However,supportprovidedby
grandchildrentoolderadultswasfacilitatedorhinderedbythemiddleparentgeneration.
ComparingresearchfindingsfromWesternsocieties,resultsofthisdissertationcaninformthe
developmentofcomparativetheoreticalmodelsofintergenerationalsupportacrosscultures.
Theseresearchfindingsalsohavepolicyandpracticeimplicationsforintergenerationalfamily
caregiversofolderruralChineseadults.Itisclearthereisaneedformoreresearchon
intergenerationalsupportexchangebetweengrandparentsandgrandchildreninvariouscultural
contexts.
1
Chapter1
IntroductionandTheoreticalFramework
Background
Socialsupportisimportantforhealthandwell-beingacrossthelifecourse,butitis
especiallyimportantinlatelifeduetophysicalandlifecourse-changes(Zimmer&Kwong,
2003).WhereasolderadultsinWesterncountriesreceivesocialsupportfromavarietyof
sources,themajorityofolderadultsinChinareceivesupportprimarilyfromtheiradultchildren
(Cheng&Chan,2006).Therefore,intergenerationalrelationshipsandsupportexchangeinthe
familycontextplayanimportantroleaspeoplegrowolder(M.Guo,2011).Intergenerational
supporthasbeenfoundtoprotectthephysicalandpsychologicalwell-beingofolderadultsin
China(e.g.,Chen&Silverstein,2000;Silverstein,Cong,&Li,2006).However,formanyrural
residents,supportingtheirolderparentshasnotbeeneasybecauseoflowerstandardsofliving
andmigrationofthemiddlegenerationtourbancities.Itisimportanttoexamineotherfamily
sourcesofeldercare,suchasgrandchildren,andtodevelopacomprehensiveunderstandingof
therolesthatgrandchildrenplayinsupportingolderadults.Demographicshiftshaveresultedin
changingrolesforgrandparentsandgrandchildrenincontemporaryruralChinesefamilies,
particularlygiventheincreasingoverlapbetweenthelifecoursesofgrandparentsand
grandchildren(Smith,2011).
Duetoincreasedlongevity,theworld’sagingpopulationisincreasinglyexpanding.
Grandparenthoodhasbecomeinseparablylinkedtooldage,andtheroleisviewedasanormal
stageofthefamilycycle(Hoff&Tesch-Römer,2007).Mostpeoplenotonlybecome
grandmothersorgrandfathers,butalsoseetheirgrandchildrenreachadulthoodandstartfamilies
2
oftheirown.Asaresult,familyrelationshipsandsupportexchangeinthree-generationfamily
networkshavebecomerelativelycommonandevenasocietalnorm(Harper,2005;Hoff,2007).
Bengtson(2001)suggestedthatgrandparentswillplayanincreasinglyimportantrolein
multigenerationalfamilies.
Assupportproviders,grandparentsoftenserveasprimarycaregiversofgrandchildrenin
theabsenceofparentsorassecondarycaregiverstosupplementotherformsoffamilycare.As
supportreceivers,olderadultsprimarilyreceivesupportfromtheiradultchildren.However,in
manycases,adultchildrendonotserveascaregiversforolderadults.Asaresult,grandchildren
oftenassumeimportantsupportrolesfortheirgrandparents.Researchhasshownthatadult
grandchildrenarepotentialsourcesofinstrumental,financial,andemotionalsupportfor
grandparents(Bengtson,2001;Harwood&Lin,2000).Giventheincreasinglysalientrolesthat
grandchildrenplayinthelivesofolderadults,itiscriticaltoexploresupportexchangebetween
grandparentsandgrandchildren;examinetheunderlyingmechanisms,reasons,orconditions
associatedwithsuchsupportexchange;andinvestigatehowsupportfromgrandchildrenaffects
thewell-beingofolderadults.ThisisespeciallytrueforruralChina.
By2050,Chinawillhaveoneoftheworld’soldestpopulations,withmorethanonein
fiveresidentsolderthan65yearsandmorethanoneinthreeolderthan60years(UnitedNations,
2004).Ruralelders,whoconstitutemorethantwo-thirdsofChina’solderpopulation,receive
fewerpublicservicesandlesseconomicsupportthantheirurbancounterparts(Jiang,1995).
Furthermore,olderadultsinruralareasarecommonly“leftbehind”bymigrantchildrenseeking
jobsineconomicallyexpandingurbancities(Goldstein,Zhigang,&Goldstein,1997).Onone
hand,thislabormigrationamongtheruralworking-agepopulationhasresultedingreater
geographicalseparationbetweengenerations,reducingopportunitiesforolderpeopletolivewith
3
andreceivesupportfromtheiradultchildren(Silversteinetal.,2006).Thus,thebackboneofthe
eldersupportsystemisweakenedbythedecreasedavailabilityofadultchildren(Jiang,1995).
Ontheotherhand,olderadultswhoareleftbehindaremorelikelytolivewithorinteractwith
theirgrandchildren(T.Z.Chen,2009;Huang,2009;Li,2004;Silversteinetal.,2006),thereby
makingitpossibleandpracticalforgrandparentsinruralChinatoexchangesupportwiththeir
grandchildren.
GrandchildsupportforolderadultsalsohasculturalsupportinruralChina.Grandparents
inruralChinavoluntarilyinvolvethemselvesinsupportinggrandchildrenforthepurposeof
improvingthefamily’sfinancialsituation(Baker&Silverstein,2012).Inaddition,grandparents
maintainaclosebondwiththemiddlegenerationthroughemotionalandfilialobligationor
remittancesfromadultchildrenwhohavemigratedelsewhere(Silversteinetal.,2006).
Grandparent–grandchildsupportexchangealsohashistoricalculturalrootsinthatgrandparents
oftentakecareofgrandchildren.Inreturn,grandchildrenareexpectedtorespectandcarefor
theirgrandparentsastheyenteradulthood.Thisexpectationmightchangeincomingdecades
becauseofthepossibleerosionoffilialpietyasaresultofmodernization,industrialization,and
globalization.Giventhatsocialservicesandcommunitycareforolderadultsarelargely
underdevelopedinruralcommunities(Yipetal.,2007)andthatsupportfromadultchildrenis
unstablebecauseoftheirmigration,howruralChineseolderadultsseeksupportfromanother
familygroupofgrandchildrenhasbecomeanimportantresearchtopic.Therefore,itiscrucialto
exploresupportexchangebetweengrandchildrenandolderadultsinruralChina,examinethe
underlyingmechanismsthatdrivethissupportexchange,exploretherelationshipbetween
supportfromgrandchildrenandadultchildren,andinvestigatetheeffectofsupportfrom
grandchildrenonthepsychologicalwell-beingofolderadults.
4
TheoreticalFramework
Thethreestudiesinthisdissertationwereguidedbycorporatetheory,intergenerational
solidaritytheory,aswellasfamilysystemstheory,althougheachstudyhadindividualguiding
theories.Thesetheoriesprovideageneralunderstandingofwhyandhowgenerationswithin
familiessupportoneanother,aswellashowthatsupportaffectsthewell-beingofindividuals.
CorporateTheory
Thecorporatemodelisanimportanttoolforunderstandingintergenerationalsupport
withintheChinesefamily.Itcanbeseenasabranchofreciprocalexchangetheory.Thismodel
treatsfamilyasacloselyrelatednetwork.Allfamilymembersinthenetworkcareforthewell-
beingofothermembersandseekoptimaldistributionofresourceswithinthenetwork(Sun,
2002).Indevelopingcountries,formalmechanismsofsupportarerareornonexistent;exchanges
maymanifestasacomplexdiffusionofresourcesbetweenandwithingenerations(Agree,
Biddlecom,Valente,&Chang,2001;Attias-Dunfut,2003;Silverstein,2006).Chinesefamilies
havelongbeenportrayedascorporatekingroupsconsistingofmembersrelatedbybloodand
marriage(Sun,2002;Tu,Freedman,&Wolf,1993).Alargerfamilyconsistingofchildrenand
grandchildrenrepresentsabiggerpoolofpotentialresourcesforolderadults.Whenseniorsare
inneedofhelp,youngergenerationsareexpectedtoprovidesupport.Therefore,thecorporate
modelprovidesanotherimportantrationalofwhyIincludedgrandchildrenasimportantfamily
supportresourcesinthedissertationstudy.
IntergenerationalSolidarityTheory
Theintergenerationalsolidaritytheoryoutlinesthebehaviorsandsentimentsthatlink
generations—typicallytherelationshipsbetweenparentsandchildren,butmorerecently
betweengrandparentsandgrandchildren(Bengtson,Olander,&Haddad,1976).Itisamongthe
5
mostpopularorganizingframeworksforunderstandingfamilyrelationshipsanddynamicsin
laterlife(Lüescher&Pillemer,1998).Ingeneral,theintergenerationalsolidaritytheoryoffersan
explanationforwhyandinwhatwaysfamilymembersdeveloplinkswithoneanotheracross
generations.
Thistheorypositsthatties,support,andinteractionsbindfamiliestogether(Bengtsonet
al.,1976).Accordingtothistheory,familymembersdifferintheircommitmentto
intergenerationaltransfersandsolidarity.Familymemberswithastrongsenseofsolidaritytend
toassistoldergenerations.Chinesefamiliesareoftencharacterizedbysolidarityandcohesion.
Therefore,adultchildrenandgrandchildreninafamilywithstrongsolidaritywouldbewilling
andhappytoprovidesupporttoolderadults,especiallywhenothergenerationscannotprovide
thissupportduetovaryingreasons.
Accordingtotheintergenerationalsolidaritytheory,familymembersneedtodo
somethingthatspecificallybenefitsothermembersofthefamilytomaintainsolidfamilyunity.
Althoughthetheoryplacesaspecificfocusonthedegreetowhichmembersoftheyounger
generationprovidesupporttooldergenerations,itstillemphasizesreciprocityasaprecondition
ofstrongintergenerationalsolidarity.Hence,supportisrequiredinbothdirections.
Theintergenerationalsolidaritytheoryalsodescribesvariousformsoffamilysupport
acrossgenerations.Althoughthepresentsolidaritytheoryoutlinessixdimensionsofsupport,
threeclassicaltypesofsupportareparticularlysalient:(1)practicalsupportorinstrumental
support,suchasassistancewithhouseholdtasks,provisionofpersonalcaretoolderadults,or
helpwithbabysitting;(2)financialsupport,suchasmoneygiventochildrenforeducationor
livingexpensesandfinancialaidgiventoolderadultsinneed;and(3)emotionalsupport,such
6
Grandparent
Parent
Chinese
Culture
Grandchild
Ruralareaof
China/migration
OtherEcologicalContexts
Individuallifecourse:Patterns
andtiming
asattention,advice,orconsolationsharedacrossgenerations.Thisdissertationadaptedthese
threedimensionsofsupporttoexamineintergenerationalexchangesofsupportinruralChina.
FamilySystemsTheory
Thefamilysystemstheoryalsoprovidedatheoreticalrationaleorframeworkforthe
threestudiesinthedissertation.Thefamilysystemstheorypositsthatfamiliesaresystemsof
interconnectedandinterdependentindividuals,noneofwhomcanbeunderstoodinisolation
fromthesystem(Bowen,1966;Kerr&Bowen,1988).King,Russell,andElder(1998)
illustratedtheinterdependenceofgenerationswithinafamily(Figure1.1).
Figure1.1.FamilySystems ofGrandparenting(Kingetal.,1998)
Theinterlockingovalsrepresenteachmemberofathree-generationalfamilystructure;
theshadedarearepresentstheoverlappinglivesofgrandparentsandgrandchildren,whichare
alsoembeddedintheparentgeneration.Inaddition,membersofthefamilysystemexperience
individualanduniquepatternsandoccurrencesofvariouslife-courseevents(Kingetal.,1998).
7
Minuchin(1988)proposedseveralbasicconceptsofsystemstheory,threeofwhichcan
beappliedtothisdissertation.First,thefamilygroupiscomprisedofsmallergroupings,or
subsystems.Eachindividualinthehouseholdispartofmultiplesubsystems.Patternsof
interactionbetweensubsystemsarehierarchicallyorganized(Minuchin,1988).Grandparents,
parents,andgrandchildrenmayformsubgroupsorsubsystemsdependingonfamily
characteristics.Additionally,thesesubsystemshaveaninternalhierarchicalorderwith
grandparentsoccupyingthehighestlevel,whichisconsistentwithConfucianideology.Despite
theirintrinsicrolefunctions,grandparentsandgrandchildrenarealsoengagedinmultiple
systemsassupportreceiversandproviders.Thisconstitutesthebasicfoundationof
intergenerationalsupportexchangewithinthefamilysystem.
Second,familymembersareintenselyconnectedonanemotionallevel.Familymembers
soprofoundlyaffectoneanother’sthoughts,feelings,andactionsthatitoftenseemsasifthey
arelivinginthesame“emotionalskin”(Minuchin,1988).Parentalactionsorfeelingstoward
olderadultswilllikelyaffecttheactionsandbehaviorsoftheyoungergeneration.Therefore,the
emotionalbondbetweengrandparentsandtheiradultchildrenmaymediategrandparent–
grandchildsupportexchange,atopicexaminedinthefirststudyofthisdissertation.Inaddition,
familymembersinfluenceoneanother.Themiddlegenerationmayserveasamodelorexample
fortheyoungergenerationtointeractwiththeirgrandparents,whichwasanissueinvestigatedin
thesecondstudyofthisdissertation.
Third,familysystemsexperienceperiodsofhomeostasisandchange.Familymembers
solicitoneanother’sattention,approval,andsupportandreacttooneanother’sneeds,
expectations,anddistress.Asonememberorsubsystemchanges,othersmustadaptinresponse.
Asaresult,familyprocessesbecomeunstableduringperiodsofchangeandtransition.Limited
8
orabsentsupportfromthemiddlegenerationcanbeseenasaperiodofchangeandtransition
(Minuchin,1988).Howgrandchildrenrespondtothischangeandwhethersupportfrom
grandchildrenaffectsthepsychologicalwell-beingofolderadultsformedthebasesofthesecond
andthirdstudiesofthisdissertation.
RuralChinaasaResearchSetting
Ashometothelargestpopulationandthelargestagingpopulationintheworld,Chinais
facingdramaticandconcurrentdemographicchangesintermsofurbanizationandpopulation
aging.MassivemigrationfromruraltourbanareasofChinasince1979constitutesthelargest
flowoflaborawayfromagricultureinworldhistory(Taylor,2001;Zhang,1998).Meanwhile,
theChinesepopulationisagingatamuchmorerapidpacethanallotherindustrializedsocieties.
Thecombinationofthesetwodemographicchangeshasinstigatedmajorsocialshifts,foremost
amongthembeingtheurgentneedforeldercare.Caringforolderadultshasbecomeacommon
challengeforfamilies,communities,andthestate(Luo,2009),particularlyinruralareasof
China.
Ofthe118.9millionChinesepeopleaged65yearsorolderin2010,approximatelytwo-
thirdslivedinruralregions(NationalBureauofStatisticsofChina,2010).Theirlivelihoodand
well-beingisoftenendangeredbylittleornonexistentincomeandalackofpensionsandhealth
insurance(Feng&Xiao,2007).Comparedwiththeirurbancounterparts,ruralolderadults
experiencemorefunctionallimitationsandworsehealth(He,Sengupta,Zhang,&Guo,2007).
However,theChinesegovernmenthasnotdevelopedcomprehensivesocialwelfaresystemssuch
associalsecurityandMedicareforolderadultsinruralareas.Althoughmorethanhalfofolder
adultsinurbanChinaareeligibleforvariouspensionandhealthcarebenefits,only4.67%of
ruralolderadultsreceivepensions(Feng&Xiao,2007;Giles&Mu,2007).Manyolderadultsin
9
ruralareascontinuetoworkintotheir70sandhavetorelyonadultchildrenforallformsof
supportinlaterlife.Therefore,issuesoffamilyrelationsandintergenerationalsupportmaybe
moreimportantforruralolderChineseadultsthantheirurbancounterparts(M.Guo,2011).First,
olderadultsinruralChinamaybemoresensitivetointergenerationalsupportbecauseofa
strongeremphasisonConfucianidealscomparedtourbanregions.Second,becauseruralolder
Chineseadultsaremorelikelytorelyontheirchildrenforsupportthantheirurbancounterparts
becauseofthelackofsocioeconomicresources,intergenerationalrelationshipsmighthavea
strongereffectonthesenseofidentityandpsychologicalstatusofruralolderadults.
However,significantrural-to-urbanmigrationandtheincreasingmobilityofadult
childrenareimportantchangesinfluencingtheexperiencesofallfamilymembersinruralChina.
Thenumberofpeoplemigratingfromruraltourbanareashasmorethanquadrupledinrecent
decades,from34millionin1989to147millionin2005(NationalBureauofStatisticsofChina,
2005),constitutingapproximatelyonequarterofthetotalChinesepopulationinruralareas
(Goodkind&West,2002).Thisresultedinanestimated18millionruralolderadultswith
migrantchildreninChinain2000.Duetothemassiverural-to-urbanshift,migrantadultchildren
areincreasinglyunabletoprovidefamilialcare,particularlyphysicalcarefortheirelderly
parents.Becauseadultchildrenplayacrucialroleinprovidingvariousformsofsupporttoolder
adultsinChina(Chi&Chou,2001;Whyte,2004),removingyoungandmale-dominated
populationsfromruralareashasledtoagrowingconcernregardingthesupportandwell-being
ofolderpeople(M.Guo,2011).Olderparentswhoareleftbehindinruralareaswerefoundto
experienceadeclineinpsychologicalwell-beingcausedbytheabsenceoftheirchildren(Du,
Ding,Li,&Gui,2004).Althoughstudieshaveshownthatgeographicalmobilitydoesnotresult
infamiliesabandoningtheirelders(Antonucci,Jackson,&Biggs,2007;M.Guo,Aranda,&
10
Silverstein,2009),separationbetweenadultchildrenandagingparentscanbeastructuralbarrier
impedingintergenerationalsupportexchange.Howtheimmigrationofadultchildrenaffects
intergenerationalrelationshipsandsupportforruralolderadultsandhowruralChineseolder
adultssecuresupportfromfamilymembershasbecomeanincreasingfocusamongresearchers.
Giventhegrowingnumberofruralfemalemigrantworkers(NationalBureauofStatistics
ofChina,2010),ruralolderadultsaremorelikelytoassumeprimarycaregiverrolestotheir
grandchildrenwhenbothparentsareabsent(skipped-generationhouseholdstructure).Recent
studiesonchildrenofmigrantworkersfoundthataroundonethirdofthesechildrenlivedwith
theirgrandparents(Duan&Wu,2009;D.G.Wu,2006).Thisprovidesanopportunityforolder
adultstofunctionasgrandparentsandinteractandexchangesupportwiththeirgrandchildren.
Thus,grandchildrenbecomeanotherimportantsourceofsupportforruralolderadultstorelyon.
Insum,thechangingsociodemographiccontextinruralChinahasprovidedboth
opportunitiesandrestraintsforintergenerationalsupport.Byexaminingpatternsof
intergenerationaldynamicsbetweengrandparentsandgrandchildrenandcomparingthemto
respectivefindingsfromWesternstudiesandstudiesconductedinurbanChina,thisresearchcan
informthedevelopmentofcomparativetheoreticalmodelsofintergenerationalrelationsacross
differentculturesandsocialcontexts.
PurposesoftheStudy
Thisdissertationincludesthreeindependentbutinterrelatedstudiesthatwereguidedby
thefamilysystemstheoryandaddressintergenerationalsupportexchangebetweengrandparents
andgrandchildrenwithconsiderationofthemiddlegeneration.Thefirststudyinvestigated
currentgrandparent–grandchildsupportexchange(receiptandprovision),howpastcaregiving
bygrandparentsaffectscurrentsupportexchange,andwhethertheemotionalbondbetween
11
grandparentsandtheiradultchildrenwereassociatedwithcurrentsupportexchange.Thesecond
studyusedchangescoresbetweenadjacentwavesfrom2001to2009toexaminethe
interrelationsbetweeninstrumentalsupportfromgrandchildren,adultchildren,orboth.This
studyexploredtheunderlyingpatternsofchangeininstrumentalsupportfromgrandchildrenand
adultchildrenandexaminedwhetherinstrumentalsupportfromgrandchildrensupplantedor
complementedsupportfromadultchildren.Usingfour-wavelongitudinaldata,thethirdstudy
investigatedtheeffectofreceivingsupportfromgrandchildrenonthepsychologicalwell-being
ofgrandparents.Specifically,itexaminedhowinstrumentalsupportfromgrandchildrenand
psychologicalwell-beingchangedovertime,howchangesininstrumentalsupportinfluenced
changesindepressivesymptoms,andwhethersupportordepressionplayedapredictingroleifa
reciprocallinkagewasfound.
Thisdissertationaddressedthefollowinggapsintheknowledgebase.First,most
previousresearchonintergenerationalrelationshipshasbeenconstrainedtoexaminationof
supportexchangepatternsordynamicrelationshipsbetweenolderadultsandtheiradultchildren
andtherelatedimpactonpsychologicalwell-being(e.g.,Cong&Silverstein,2008a,2008b;M.
Guoetal.,2009;Silversteinetal.,2006;Song,Li,Zhang,&Feldman,2007),grandparent–
grandchildrelationships(e.g.,Monserud,2008a,2008b,2010),ortheeffectsofgrandparent
caregivingonthewell-beingofolderadults(Goodman&Silverstein,2002;B.Guo,Pickard,&
Huang,2007;Harris,2002;McGowen,Ladd,&Strom,2006;Pruchno&McKenney,2002).
However,supportexchangebetweengrandparentsandgrandchildreninthecontextofan
emigrantmiddlegeneration,thedynamicrelationshipbetweentwoyoungergenerationsinterms
ofprovidingsupporttothesameolderadult,andtherelatedimpactsonthepsychologicalwell-
beingofolderadultshaveremainedlargelyunexplored.Second,therearefewlongitudinal
12
studiesofintergenerationalsupportinagingfamilies.Byusingfour-wavelongitudinalsurvey
datafromthe“Well-BeingoftheElderlyinAnhuiProvince”studyinChina,thisdissertation
allowedforestimateswithhigherreliabilitycomparedwithcross-sectionalstudiesorstudies
usingtwo-wavedesignsand,thus,moreaccuratelycapturedthecausalitybetween
intergenerationalsupportandpsychologicalwell-being.Third,anotherareararelytouchedupon
inpreviousresearchisthatofreciprocallinkagesandwhethergrandparent–grandchildsupportor
psychologicalwell-beingofolderadultsplaysapredictingroleshouldsuchalinkexist.Previous
studiesmainlyfocusedontheeffectsofsupportexchangewithadultchildrenonthe
psychologicalwell-beingofolderadults(Cong&Silverstein,2008b;M.Guoetal.,2009).Less
isknownabouthowpsychologicalwell-beingaffectssupportexchange,suchasreceiving
instrumentalsupportfromgrandchildren,asexaminedinthethirdstudyofthisdissertation.The
lackofknowledgeonthistopiccloudsourunderstandingofthedynamicrelationshipbetween
grandparent–grandchildsupportandpsychologicalwell-beingamongelders.
13
Chapter2
DescriptionoftheDataset
Dataforthisstudyweredrawnfromthe“Well-BeingofElderlyinAnhuiProvince”
study(hereafterreferredtoastheAnhuistudy)inChina,afour-wavelongitudinalsurveystudy
jointlyconductedfrom2001to2009bytheSocialofSocialWorkandSchoolofGerontologyat
theUniversityofSouthernCaliforniaandthePopulationResearchInstituteofXi’anJiaotong
University.Datawerecollectedfromasampleofadultsaged60yearsorolderlivinginrural
townshipsinChaohu,AnhuiProvince.
SurveySite
ThedatawerecollectedintheChaohuregionofAnhuiProvince.AnhuiProvinceis
locatedinthegeographiccenterofChinaandhasalargemigrantlaborerpopulation.In2000,
morethan7.8millionrurallaborersmigratedtonearbyurbancities,comprising15.4%ofthe
totalruralpopulationinAnhuiprovince(ChinaCensusBureau,2004).Chaohuislocatedinthe
geographiccenterofAnhuiprovinceandischaracterizedbymoderateeconomicdevelopment
andpopulationdensity.Thispoor,ruralregionofAnhuiprovinceisknownforitsrelativelyhigh
ratesofworking-agelabormigrationtothecitiesofHefei,Nanjing,andShanghai(Silversteinet
al.,2006).In2000,therewereapproximately3.76millionruralresidentsintheChaohuregion,
comprising83.7%ofthetotalpopulation(ChinaCensusBureau,2004).Amongtheserural
residents,12%ofworkinglaborershadmigratedtourbancities(ChinaCensusBureau,2004).
Thus,ChaohuisrepresentativeofAnhuiprovinceandmostruralareasincentraloreasternparts
ofChina.
14
Chaohuisalsoknownforitsrelativelyhighdensityofolderadults.Accordingtocensus
figuresin2000,therewere387,046olderadultsaged60yearsoldorolderlivingintheregion
(ChinaCensusBureau,2004).Excludingtheworking-agemigrantpopulation,theproportionof
olderadultsinChaohuareawasabovethenationalaverageof10%.Surveyresultsindicatedthat
insomevillageswithhighratesofmigration,youngchildrenandolderadultswerethemain
residents.Basedonthecurrentmigrationrate,ithasbeenestimatedthattheelderlyratewill
increaseinthenearfuture(Zuo,2011).
InChaohu,theannualpercapitanetincomewasmorethan2000RMB($307.69)in2000;
moneybroughtbackbymigrantlaborerscomprised8%ofnetincome.Anhuistudydata
indicatedthatmostlandhadbeenleasedtoothersbecauseeldersandyoungchildrenwereleft
behindbythemigrantmiddlegeneration.Thus,itwasestimatedthatmoneyfrommigrant
laborerswasthemainsourceofincomeformanyfamilies.
Inrecentyears,mostofthemigrantshaveservedasseasonaltemporaryworkers(Zuo,
2011);returningtotheirhometownsforfestivalsorharvestseasonsandmigratingtocitiesfor
workinotherperiods.Thosemigrantsretainasenseofbelongingandsecurityregardingtheir
ruralhometown(Zuo,2011).LikemostotherruralChineseareas,99%ofresidentsinChaohu
identifyasmembersoftheHanethnicityandhaveadeepideologyofpatrilinealheritage(Cong
&Silverstein,2012).Inaddition,manyolderadultsstillretainthetraditionalthinkingthat
havingmorechildrencanbringthemhappinessinoldage.Becausefamilysupport,especially
fromadultchildren,isthemainformofsupportinoldage,therearefewelderlyinstitutionsor
seniorcentersinthearea(Zuo,2011).Thistypicalruralarearepresentedagoodopportunityto
gainadeepunderstandingoffamilysupportanddynamicsacrossgenerationsinruralChina.
15
DataSampleandCollection
Datasampleatbaseline.Thesamplewasselectedusingastratifiedmultistagemethod.
First,12ruraltownshipswererandomlyselectedfromall126townshipsintheChaohuregion.
Second,sixadministrativevillageswererandomlyselectedineachtownship.Third,withineach
selectedvillage,allpeopleaged60andolderwerestratifiedtoformtwosamplingframesbased
onage:(1)thoseaged60–74and(2)those75andolder.Toguaranteethatonlyoneolderperson
perhouseholdwasinterviewed,twoapproacheswereusedduringthesamplingprocess.If
householdpartnersweredividedbetweenthetwoagegroups,theyoungerpartnerwasdropped
andasubstituterespondentwasrandomlyselectedasareplacement.Ifhouseholdpartnerswere
inthesameagegroup,thenthepartnerchosensecondwasdroppedandasubstituterespondent
wasrandomlyselected.Duringthesurvey,iftheoriginallyselectedparticipantswereyounger
than60yearsold,werefoundtobedeceased,orwerenotintownatthetimeofsurvey,another
participantwasdrawnfromthesamplingframetoreplacethatperson.Surveyswereconducted
inrespondents’homes.ThebaselinesurveywasconductedinApril2001.Ofthe1,802eligible
participantsrandomlyselectedatbaseline,1,715completedthesurvey,yieldingaresponserate
of95.3%.Therewasnorandomsamplingperformedduringfollow-upsurveys,exceptforthe
4thwavewhichincludednewcohortofdespondences.Forthenewcohortof416participantsin
2009,asimilarsamplingprocedurewasadoptedasthebaselinesurvey.
Follow-upsurveys.Follow-upinterviewswereconductedinNovember2003,December
2006,andApril2009.Whenconductingthefollow-upsurveys,researchershadtoensurethat
interviewsubjectshadparticipatedinthebaselinesurveyin2001.Adetailedprocesswas
followedtoidentifyparticipantsduringfollow-upsurveys:
16
(1)Detailedinformationwasrecordedregardingthevillagenumber,groupnumber,age,
andIDnumberduringthebaselinesurvey.Morethanoneresearchercheckedthisbackground
informationverycarefullyandentereditintotherecord.
(2)Beforeeachfollow-upsurvey,researcherscomparedthebasicinformationofeach
participantinthe2001datasettothoseintherecord,thusobtainingkeyinformationaboutthe
individualbackgroundandfamilystructureofeachparticipant.
(3)Basedonthekeyinformation,theinterviewerswentintoeachrespondent’shousehold
toconductthesurvey.Beforebeginningthesurvey,interviewersmadesurethatthebackground
informationoftherespondentmatchedtheinformationintherecord.
(4)Aftercheckingthebackgroundinformationandconfirmingtherewasnoerrorin
identifyingtheparticipant,reviewersgaveeachparticipantthesameIDsurveynumberasin
2001andstartedthesurvey.
(5)Olderadultswhocouldnotbefoundduringfollow-upsurveysweretreatedasa
missingcaseaftercarefullycheckingthebasicinformation.Reasonsfornotfollowingupwere
enteredintotherecord.
In2003,1,391participantswereinterviewed,representing79.8%oftheoriginalsample.
Ofrespondentswhowerenotlocated,76hadmovedoutofthevillage(6%)and240died(14%).
Twenty-threeformerrespondentswerelocatedbutdeclinedtoparticipate,terminatedtheir
interviews,and/orweretooilltobeinterviewed.In2006,ofthe1,391Wave2respondents,236
peoplehaddied,57peoplehadpermanentlyortemporarilyleft,and31peopledeclinedto
participateorcouldnotbelocated,resultinginasampleof1,067respondents.In2009,1,224
participantswereinterviewed,including808originalparticipantsandanewcohortof416
participants.Figure2.1illustratesthesamplesizeandattritionoftheAnhuistudyovertime.For
17
thisdissertation,Iusedsamplesfromdifferentwavesofthesurveyfordifferentstudies,as
describedlater.
Surveyimplementation.Thebaselinesurveywasconductedduring6daysinApril2001.
Theresearchgroupincludedsurveyorganizers,surveysupervisors,interviewers,and
interviewees.Thesurveysupervisorsweredoctoralandmaster’sstudentsfromXi’anJiaotong
Universityandtheinterviewersweremainlylocalinstitutionstafffamiliarwiththedialect,the
participantsandtheChaohuregion.Eachtownshiphadasurveysupervisorwhowasinchargeof
threetofourinterviewers.Trainingwasconductedduringthefirstdayforbothsurvey
supervisorsandinterviewers.Thecontentoftrainingincludedsurveyskills,rulesregarding
fillingoutthequestionnaire,andgettingfamiliarwiththequestionnaires.Mockinterviewswere
conductedattheendofthefirstdaytoidentifyandsolveandproblemsorerrors.On-sitesurveys
wereconductedduringtheremainingfivedays.Inthebeginning,eachinterviewerconducted
pilotsurveysundertheguidanceofsupervisorstomakesuretheyunderstoodandwerefamiliar
withthesurveyprocessandwerenotmakingerrors.Afterthatinitialreview,interviewers
conductedsurveyindependently.Surveysupervisorscheckedeachfilledquestionnaireand
returnedittotheintervieweriftherewereanymistakesorunnecessarymissinginformation.
Sometimesinterviewersneededtoreturntoparticipantstocollectmissinginformation.In
addition,thesupervisorrandomlyselectedsomequestionnairesandaskedtheparticipantssome
keyquestionsagaintomakesurethesurveywasconductedaccurately.Questionnaireswith
severeproblemsthatcouldnotbecorrectedweretreatedasinvalid.Proceduresduringfollow-up
surveysweresimilartothebaselinesurveyin2001andalsolastedfor6dayseach.
18
Figure2.1.SampleSizeandAttrition intheAnhuiStudyinChina,2001–2009(M.Guo, 2011)
2006
1,368(79.8%)
Declined/missing: 87(4.7%)
1,715(95.3%)
TotalNon-respondent:347(100%)
Deceased: 240(69.2%)
Left: 84(24.2%)
Declined/missing: 23(6.6%)
2001
2003
1,802
TotalNon-respondent:301(100%)
Deceased: 236(78.4%)
Left: 57(18.9%)
Declined/missing: 8(2.7%)
TotalNon-respondent:259(100%)
Deceased: 173(66.8%)
Left: 33(13.7%)
Declined/missing: 53(20.5%)
Newcohort: 416
2009
1,067(78.0%)
808(75.7%)
1,224
19
ContentoftheQuestionnaires
ThepurposeoftheAnhuistudywastounderstandintergenerationalrelationships,
supportexchange,andthephysicalandmentalhealthstatusofolderadultsinruralChina.The
threefollow-upsurveysmeasuredchangesinfamilysupport,especiallythepatternsoffamily
supportcausedbythemigrationofthemiddlegenerationandhowthosechangesaffectedthe
well-beingofolderadults.Questionnairesusedforthefoursurveyswerenotexactlythesame.
Researchersrevisedthequestionnairesbasedonfindingsoftheprevioussurvey(s)andnew
researchpurposes.Nonetheless,themaincomponentsofthequestionnaireremainedthesame,
includingthefollowinginformation.
(1)Backgroundinformation,includingage,gender,maritalstatusandfamilystructure
information.
(2)Sociodemographicinformation,includinghousing,education,occupation,etc.
(3)Healthstatusinformationintermsofself-ratedhealth,activitiesofdailyliving
(ADLs),instrumentalactivitiesofdailyliving(IADLs),physicalmobility,andspecific
multimorbidityinformation.
(4)Informationaboutalllivingchildrenofeachparticipant,includingage,gender,
education,occupation,maritalstatus,livingarrangement,geographicalproximity,timeofand
reasonformigration,frequencyofinteraction,supportexchange(instrumental,financial,and
emotional),whetherparticipantscaredfortheiryoungchildren,intensityofcaregiving,etc.
(5)Informationaboutparticipants’attitudesregardingtraditionalold-agesupport.
(6)Informationaboutthepsychologicalwell-beingofolderadults,suchaslife
satisfaction,depression,andworry.
20
(7)Informationaboutprovidingandreceivingdailypersonalcareandassistancewith
householdchoirswiththefollowingpersons:(1)sons;(2)daughters,(3)daughters-in-law;(4)
sons-in-law;(5)spouse;(6)grandchildren;(7)grandchildren-in-laws;(8)otherrelatives.
(8)Informationaboutolderadults’economicstatus,suchasincome,pension,financial
supportfromothers,andfinancialsatisfaction.
(9)Informationaboutolderadults’cognitiveability.
Tocollectin-depthinformationaboutintergenerationalinteractionandsupportexchange
witheachoftherespondent’sadultchildren(firstthreewaves)andeachofthegrandchildrenin
theireldestchild’sfamily(fourthwave),theinterviewwasstructuredsuchthatrespondents
discussedeachchildatlength,beginningwiththeeldestandmovingontootherchildren,if
applicable.Thisapproachwasalsoadoptedwhencollectinginformationaboutgrandchildren.
Togatherinformationaboutolderadultswhodidnotparticipateinfollow-upsurveysdue
tomigrationordeath,separatequestionnaireswereaddedin2003,2006,and2009.The
additionalmigrationquestionnaireincludedquestionsabouttimeof,locationof,andreasonsfor
migration,aswellaslivingarrangementsaftermigrationandwhethertheyreturned.The
additionalmortalityquestionnaireincludedquestionsabouttimeof,locationof,andreasonsfor
death,livingarrangementsbeforedeath,timespentinbedbeforedeath,medicalcostsand
paymentbeforedeath,formsofsupportreceivedbeforedeath,andfinancialarrangement(will
andtestament,etc)afterdeath.
QualityoftheDataset
ThedatasetoftheAnhuistudyhasbeenshowntobeofhighqualityinpreviousstudies
(e.g.,Cong&Silverstein,2008a,2008b;M.Guoetal.,2009;Silversteinetal.,2006;Songetal.,
2007).Duringthedatacollectionprocess,researchersuseddifferentformsofquality-control
21
methodstoensurehighquality.Forexample,attheendofeachsurvey,theinterviewerfilledout
surveyprocessformstoindicatetheobjectivityandaccuracyoftheinformationgainedfromthe
survey.Althoughthereweresomeerrors,theywerewithinanacceptablerange.Responserates
werequitehighatbaseline(95.3%),andtheattritionratewasapproximately15%to20%
betweenanytwoadjacentsurveypoints,whichisreasonableforlongitudinalstudies.
Thehighqualityofthisdatasetisalsoreflectedinthemeasurementofvariables.For
example,weusedADL,IADLs,andmobilitytomeasurethehealthstatusofolderadultsanda
scaleofintergenerationalsolidaritytomeasuretheemotionalclosenessbetweenolderadultsand
theiradultchildren.Thosemeasurementswereconsistentwiththoseusedinmostprevious
studies,includingthoseinWesternsocieties.Thealphascoresofthesemeasurements,which
alsoindicatethehighqualityoftheAnhuidataset,areshowninTable2.1.
Table2.1
Cronbach’sAlphaReliabilityScoresforSomeVariablesintheAnhui
Study
Variable 2001 2003 2006 2009
ADLs .92 .96 .95 .89
IADLs .89 .93 .89 .90
Mobility .94 .91 .93 .87
Emotionalcloseness .86 .96 .83 .81
Note.ADL=activityofdailyliving;IADL=instrumentalactivityofdailyliving.
ReliabilityscoresobtainedfromZuo,2011.
AdvantagesofUsingAnhuiData
TheAnhuistudydatasetprovidesadvantagesnotfoundinmostotherstudieson
intergenerationalsupportinlaterlife.AsoneofthelargestlongitudinalstudiesinChina,it
featuresinformationondemographics,healthstatus,andthepsychologicalwell-beingofolder
adults.Asecondadvantageisitsextensivedataonintergenerationalrelationshipsandsupport
exchangeofolderadultswitheachoftheiradultchildrenandgrandchildrenoftheireldest
22
children.Thus,thissurveydataallowsfortheexplorationofthedynamictrajectoryof
grandparent–grandchildsupportwithinthecontextofsupportexchangewithadultchildren.A
thirdadvantageisthatthesurveyisuniquetoruralagingfamilies,whichoftenexperiencethe
migrationofadultchildrentourbancitiestoseekjobs.Thus,thestudiesinthisdissertation
providedageneralpictureofintergenerationalsupportexchangewithinruralagingChinese
familiesasawhole.
23
Chapter3
DeterminantsofSupportExchangebetweenGrandparentsandGrandchildren:TheRoles
ofGrandparentCaregiving,PatrilinealHeritage,andEmotionalBonds
Introduction
MostpreviousstudiesofsupportexchangeinChinesefamilieshavefocusedonsupport
andcareprovidedbyadultchildrentotheirolderparents(e.g.,Cong&Silverstein,2008a,2008b;
M.Guoetal.,2009).Therefore,olderadultsaregenerallyprofiledasrecipientsofcarefrom
adultchildren.Theexchangeofcarebetweengrandparentsandgrandchildrenhaslargelybeen
neglectedbyresearchers(Whyte,2003).InruralChina,whereolderadultsaremorelikelytolive
andinteractwiththeirgrandchildrenduetomigrationofthemiddlegeneration(T.Z.Chen,2009;
Li,2004;Huang,2009;Silversteinetal.,2006),developingabetterunderstandingofsupport
exchangewithgrandchildrenisparticularlyimportant.Mutualityofsupportprovisionisa
necessarypreconditionoffamilysolidarity,whichhasbeenemphasizedinChineseculture
throughouthistory.Ifintergenerationalsolidarityisappliedtograndparent–grandchild
relationships,theremustbereciprocityofsupportbetweenthetwogenerations(Bengtonetal.,
1976;Hoff,2007).Thisstudyexploredsupportexchangebetweengrandparentsand
grandchildreninruralChinaandthedeterminantsofsuchexchangewithaspecificfocusonthe
influenceofpastgrandparentcaregiving,emotionalbondswiththemiddlegeneration,and
patrilinealheritage.
IntergenerationalSupportExchangebetweenGrandparentsandGrandchildren
Socialexchangetheoryprovidesausefulparadigmforunderstandingthese
intergenerationaltransactions(Silverstein,2006).Thistheoryincludesthepremisethatsocial
24
relationshipsaregovernedbyanormofreciprocity,theexpectationthatadebtshouldberepaid
(Molm&Cook,1995;Silverstein,Conroy,Wang,Giarrusso,&Bengtson,2002).Accordingto
thedirectexchangemodel,grandparentsandgrandchildrenmaysimultaneouslyexchange
differentgoodsandservices.Withreciprocalreceivingandgivingofsupport,intergenerational
solidaritycanremainwithinthefamilycontext(Bengtonetal.,1976).
Empiricalstudiesalsohaveprovidedevidenceofintergenerationalsupportexchange
betweengrandparentsandgrandchildren.SomeWesternresearchersfoundthatgrandparentsand
adultgrandchildrenprovideeachotherwithemotionalandpracticalsupport(Ashton,1996;
Harwood&Lin,2000;Langer,1990).Grandparentsprovidealltypesofcaregiving(e.g.,
custodial,part-time,co-parenting,etc.)totheirgrandchildrenwhentheparentalgenerationis
unabletoassumeacaregivingfunction.Studieshavefounddirectdownwardfinancialsupport
fromgrandparentstograndchildrenisnotuncommoninWesternsocieties(Attias-Donfut,2003;
Silverstein,2006).Ontheotherhand,adultgrandchildrencanbecomecaregiversfortheir
grandparents(Dellmann-Jenkins,Blankemeyer,&Pinkard,2000;Fruhauf,Jarrott,&Allen,2006;
Hoff,2007;Piercy,1998).Astudyofthree-generationfamiliesinFrancefoundthatmorethan
halfofthegrandchildrenprovidedsupportiveservicestotheirgrandparents,suggestinga
reciprocalflowofassistance(Attias-Donfut,2003).Otherresearchersfoundmutualityofsupport
betweengrandparentsandgrandchildren,especiallyintermsofinstrumentalsupportexchange
(Silverstein,Giarrusso,&Bengtson,1998).Intermsofthebalanceofmutualsupport,Hoff
(2007)foundthatinstrumentalandfinancialsupportexchangebetweengrandparentsand
grandchildrenwasimbalancedorasymmetricalattheoldergeneration’sexpenseinGermany.
Accordingtothisstudy,theproportionofgrandparentsprovidingsupporttotheirgrandchildren
washigherthanthatofgrandchildrenprovidingsupporttotheirgrandparents.Substantial
25
downwardfinancialtransferswerenotmatchedbyreciprocalupwardinstrumentalsupport—not
eventowardtheoldestold,whomightbemorefragilewithadvancedageandmayneedmore
instrumentalsupport(Hoff,2007).However,anearlierstudybyLanger(1990)foundthatmost
oftherelationshipsbetweengrandparentsandadultgrandchildrenwereimbalancedatthe
grandchildren’sexpense,withgrandparentsreceivingmorethantheygave.
However,mostpreviousstudiesfocusedoneithersupportprovisionorreception,rather
thanexaminingthematthesametime.Thereisaneedtoconductmorestudiestoanalyzethe
supportexchangebetweenthesetwogenerations.Inaddition,muchoftheliteratureonthistopic
isprimarilylimitedtoWesternsocietiesandmaynotbeapplicabletoruralChinesegrandparents.
SupportexchangepatternscouldbeverydifferentforgrandparentsinChinathanthosein
Westerncountries.Ononehand,olderChineseadultsoftenpaygreaterattentiontothe solidarity,
continuity,orgenerativityoftheirfamilythanolderadultsinWesterncultures.Grandparentsin
ruralChinaoftenvoluntarilyinvolvethemselvesingrandchildcaregivingandsupporttoimprove
thefamily’sfinancialsituation(Baker&Silverstein,2012).Ontheotherhand,filialpietyis
emphasizedamongyoungergenerations,andgrandchildrenaresocializedtorespectandsupport
oldergenerationsasneeded.Therefore,mutualsupportinruralChinamightbedifferentfrom
thatofWesterncountries.Clearly,thereisaneedforabetterunderstandingofhowsupport
exchangepatternsbetweengrandparentsandgrandchildrendifferinruralChina.
DeterminantsofSupportExchangebetweenGrandparentsandGrandchildren
Pastgrandparentcaregiving.Previousinteractionbetweengrandparentsand
grandchildreninfluencescurrentsupportexchangebetweenthetwogenerations.Basedonsocial
exchangetheory,reciprocationdoesnotneedtobeimmediateandmaylagbymanyyears.
Grandparentsmayinvestintheiryounggrandchildren,andwhenthegrandchildrenareolder,
26
theypaybacktheirgrandparentswithsupportandattention.Monserud(2008a,2010)foundthat
earlierpatternsofinteractionsbetweengrandparentsandgrandchildrenaccountedforlater
grandparent–grandchildsupportexchange.Otherstudiesalsosuggestedthatcaregivingprovided
bygrandparentsorinteractionswithgrandparentswhengrandchildrenwereyoungresultedin
strongerrelationshipsandsupportiveinteractionlaterinlife(Brown,2003;Hodgson,1992).
ThisdelayedsocialexchangemodelorpreviousresearchfindingsinWesternsocieties
maybeapplicabletoruralChina.However,thealtruismmodel,anothersub-modelproposedby
Silversteinandcolleagues(2002)basedonthesocialsupportexchangemodel,canalsobe
appliedinruralChina.Accordingtothismodel,inaffectiverelations,peoplebehave
altruisticallybydoingsomethingforothersbecausetheyareconcernedwiththewelfareof
others.Insocietieswithfewsourcesofpublicsupport,suchasruralChina,familymembersare
oftenintertwinedinawebofmutualsupporttransferstoaddressthecollectiveneedsofthe
family(Silverstein,2006).Therefore,ruralChineseelderswhoengageintheculturally
importantactivityofgrandparentcaregivingtotheirgrandchildrenwhentheyareyoungmaynot
expectthemtoreciprocate.Ontheotherhand,familyrelationsarealsogovernedbysocialnorms,
suchasthekinshipobligationoftheyoungergeneration.Thisnormmandatesthatyounger
generationsshouldcarefortheirgrandparentsintimesofneed,andthisshouldnotdependon
whatsupporttheyreceivedinthepast.
Togainabetterunderstandingofthecurrentsupportexchangebetweengrandparentsand
grandchildreninruralChina,thisstudytestedtheapplicabilityofthedelayedexchangemodel,
oraltruismmodel.
Theroleofthemiddlegeneration.Theparentgenerationalsoplaysanimportantrolein
intergenerationalsupportbetweengrandparentsandgrandchildren.Familysystemstheory
27
emphasizestheinterdependencebetweenfamilymembersandprovidesageneralframeworkfor
examiningrelationshipsanddynamicexchangebetweengrandparentsandgrandchildrenwithin
thefamilysystem.Accordingtothistheory,grandparent–grandchildrelationshipsareembedded
inasystemoffamilytiesandshouldbeexaminedwithinthecompletekinshipnetwork(Cox&
Paley,1997).
Empiricalstudiesalsohaveshownthatexchangesofsupportbetweengrandparentsand
grandchildrencanbecontingentonthedynamicnatureofthegrandparent–parentrelationship.
Monserud(2008a,2008b,2010)foundthatconcurrentrelationsbetweenparentsand
grandparentscaninfluencegrandparent–grandchildrelationshipsandsupportexchange.
Goodman(2003)pointedoutthatcloseordistantrelationshipsamonggrandparents,parents,and
grandchildrencanaffectoneanother.Studiesofthreegenerationshaveshownthemiddle
generationoccupiesastrong,mediatingpositioninrelationshiptoadjacentgenerations.The
middlegenerationmayserveasagatekeeperornormsetterandfacilitateorunderminethe
interactionbetweengrandparentsandgrandchildren(Attar-Schwartz,Tan,&Buchanan,2009;
Barnett,Scaramella,Neppl,Ontai,&Conger,2010;Goodman,2003).However,thoseprevious
studiesinWesternsocietiestestedtheroleofthemiddlegenerationinthegrandparent–
grandchildrelationship,ratherthansupportexchangebetweenthetwogenerations.Givenits
generalinfluenceonintergenerationalinteractions,itwashypothesizedthatthemiddle
generationalsoplaysaroleingrandparent–grandchildsupportexchange.
PatrilinealcultureinruralChina.Thetraditionalkin-keepertheoryoffamilyrelations
isusefulforexaminingwhethermaternalorpaternalgrandparentshavedifferentinteractions
withgrandchildren.Accordingtothistheory,genderplaysacentralroleingrandparent–
grandchildinteractions(Bahr,1976;Rosenthal,1985).Comparedtomen,womenaremore
28
activelyinvolvedinmaintainingkinshipties(Lye,1996).Inaddition,Westernstudiesindicated
thatmaternalgrandparentsaremoreinvolvedinintergenerationalsupportexchangeand
interactionthanthepaternalline(Dimmock,Bornat,Pearce,&Jones,2004;Ferguson&
Douglas,2004;Gauthier,2002;Harper,Smith,Lechtman,Ruicheva,&Zeilig,2004).However,
theremaybestrongerfamilytieswithpaternalgrandparentsinsomecontexts,suchasinrural
socialenvironments.ResearchersintheUnitedStatesfoundthatpaternalgrandparent–
grandchildrelationshipsinruralfarmfamiliesfeatureahighdegreeofemotionalcloseness
(King&Elder,1995;King,Silverstein,&Elder,2003;Spence,Black,&Adams,2001).
Moreover,intermsoftheroleofthemiddlegeneration,onerecentstudydemonstratedthat
closerrelationshipsbetweengrandchildrenandpaternalgrandparentsarecontingentonpaternal
relationshipswiththeirparents(Brown,2003).
InruralChina,thetraditionaladageof“Amarrieddaughterislikespilledwater”often
holdstrue.Onlysonscancarryonthefamilynameandinheritthefamilyfortune.Onestudy
showedthatinruralChina,ason’smigrationhasagreatereffectthanadaughter’smigrationon
thelikelihoodofgrandparentandgrandchildco-residence,andgrandparentstendtolivewiththe
childrenoftheirsonsratherthanwiththechildrenoftheirdaughters(Zhang&Li,2004).Studies
alsofoundthatafather’s(middlegeneration)emotionalclosenesswithhiseldersisassociated
withthegrandparent–grandchildrelationship(Cong&Silverstein,2011).Thus,itwas
hypothesizedthatpatrilinealheritage,particularlyofpaternalgrandparentsandfathers,hasa
positiveeffectongrandparent–grandchildsupportexchangeinruralChinesefamilies.
Otherfactors.Otherissuesalsoaffectgrandparent–grandchildsupportexchange.
Accordingtotheneeds-basedmodel,anelderlyindividual’sneedforcareinfluencestheyounger
generation’smotivationtoprovidecare(Becker,1991;Becker&Tomes,1979;Cox,1987;Cox
29
&Rank,1992).Asgrandparentsage,theytendtoneedmoresupportasaresultoffrailtyand
multimorbidity.Conversely,becausemanyofthegrandchildreninAnhuistudyarereaching
adulthood,theymayhavemorecapacitytoprovidesupporttotheirgrandparents.Asa
consequence,therelationshipbetweengrandparentsandgrandchildreninruralChinamight
becomemoreorientedtowardupwardsupportfromgrandchildren.Inaddition,themodelof
bargainingpowerhadbeenvalidatedintheliteratureandindicatesthatolderpeoplereceive
supportfromchildrenaslongastheyremainincontrolofimportantresourcessuchasmoney,
land,andknowledge(Goode,1963;Luo,2009;McElroy&Horney,1981;Sun,2002).The
powerofbothgenerationsmayalsoplayaroleingrandparent–grandchildsupportexchange.
Grandparentswithmoreresourcesorcapital(education,income,beingmarried)maytendto
providemoresupporttograndchildrenwithfewerresources.Incontrast,grandchildrenwith
moreresourcesmayprovidemoresupporttotheirgrandparents.Therefore,inthepresentstudy,
needsandpower-basedvariableswerecontrolledatbothgrandparentandgrandchildlevels,
includingtheage,maritalstatus,education,income,andhealthstatusofelders;theageand
educationofgrandchildren;andthegeographicalproximitybetweenthetwogenerations.
ThePresentStudy
Inthisstudy,IexaminedsupportexchangeanditsdeterminantsamongruralChinese
eldersandtheirgrandchildren.Thisstudyextendedpreviousresearchbyfocusingonboth
upward(receiving)anddownward(providing)supportamongeldersintermsofthegrandchild
generation,inadditiontoexaminingwhetherpastgrandparentcaregivingexperiences,emotional
bondswiththemiddlegeneration,patrilinealheritage,andotherdemographicvariablesplaya
roleineachdimensionofsupportexchange.Accordingly,thefollowinghypothesesweretested:
30
(1)Moregrandparentswouldreceivesupportfrom,thanthoseprovidesupportto,
grandchildrenbasedontheincreasedneedofgrandparentsandcapacityof
grandchildren.
(2)Previousgrandparentcaregivingwouldpositivelyaffectbothupwardanddownward
supportexchangebetweengrandparentsandgrandchildren.
(3)Emotionalbondswithadultchildrenwouldpositivelyaffectgrandparent–grandchild
supportexchange.
(4)Strongpatrilinealculturewouldpositivelyaffectgrandparent–grandchildsupport
exchange.
(5)Youngerage,highereducation,beingmarried,havingmoreincome,andfewer
functionallimitationsamonggrandparentswouldbepositivelyassociatedwith
providingsupporttograndchildren.
(6)Olderage,highereducation,andclosergeographicalproximityamonggrandchildren
wouldbepositivelyassociatedwithprovidingsupporttograndparents.
Methods
Data
DataforthisstudycamefromWave4oftheAnhuistudyconductedin2009,as
describedearlier.ApilotstudywasconductedinApril2009totesttheadequacyofthequestions.
Informationaboutintergenerationalsupportexchange,whichdidnotexistinpreviousdata
collection,wasaddedtotheWave4survey.Thus,thesurveyincludedassessmentsofolder
adults’relationsandsupportexchangewithadultchildrenandgrandchildren,theirphysical
healthstatus,andtheirpsychologicalwell-being.
31
Questionsaboutintergenerationalrelationshipsandsupportbetweengrandparentsand
grandchildrenconcernedonlythegrandchildrenofthegrandparents’eldestchild;apilotstudy
conducted2monthsbeforetheformalsurveyshowedthatolderadultshavethemostclearand
detailedmemoriesaboutthisparticularrelationship.ThispracticeisconsistentwithChinese
culture,i.e.,olderadultsinChinatypicallyshowthemostconcernandpayspecialattentionto
theirfirstchild,andinturn,thechildorchildrenoftheirfirstchild.Thereisempiricalevidence
tosupportthisrationale.Euler,Hoier,andRohde(2001)foundthatgrandparentshave
significantlymorecontactwithgrandchildrenoftheirfirst-bornchildthanwithgrandchildrenof
subsequentchildren.First-bornchildrenoftenshowmoreattachmenttotheirparentsthan
subsequentchildren(Salmon&Daly,1998).Researcherswerealsoconcernedaboutlimitedtime
andresourcesfortheWave4surveybecauseeacholderadultsurveyedhadanaverageoffourto
fivechildren,eachofwhomhadanaverageoftwoorthreechildren.Itwasconsidered
unrealistictoaskolderadultsforinformationabouteachgrandchild.
Afterexclusionofparticipantswhodidnotpasscognitivetests,whodidnothave
children,whosefirstchildwasdead,andwhosefirstchildhadnochild,thefinalsamplesizefor
thisstudywas926,ofwhom47.5%wasmaleand52.5%wasfemale.Inaddition,totestsupport
exchangebetweengrandparentsandeachgrandchild,ananalyticsampleconsistingof
grandparent–grandchilddyadswasconstructed.Therefore,thedatawererestructuredinto1,843
observationsofgrandchildrenaged16yearsandolderwithin926familysets.
Measurement
Dependentvariable.Thedependentvariablewassupportexchangebetweenthe
grandparentandeachgrandchildoftheireldestadultchild. Thisincludedbothupward(receiving)
anddownward(providing)instrumental,financial,andemotionalassistance.Receiving
32
instrumentalsupportwasmeasuredbyaskinggrandparentswhethertheyreceivedanyhousehold
(e.g.,cleaninghouse,washingclothes)orpersonalcare(e.g.,takingabath,puttingonclothes)
supportfromtheirgrandchildduetopoorhealthduringthepast12months(1=yes,0=no).
Receivingfinancialsupportwasmeasuredbyaskinggrandparentswhethertheirgrandchildsent
them(ortheircohabitatingspouses)money,food,orgiftsduringthepast12months(1=yes,0=
no).Receivingemotionalsupportwasmeasuredbyaskingtherespondentswhethertheyfeltthat
theirgrandchildwasfilialandrespectfultowardthemduringthepast12months(1=yes,0= no).
Provisionofinstrumental,financial,andemotionalsupportbygrandparentstograndchildrenwas
measuredinasimilarfashion.Allsixdimensionsofsupportweredichotomizedvariables.
Independentvariables.Therewerethreekeyindependentvariablesinthisstudy.
Intensityofcareprovidedbygrandparentstotheirgrandchildinthepastwasmeasuredona7-
pointscale(0=never,1=seldom,2=aboutonceamonth,3=severaltimesamonth,4=at
leastonceaweek,5=aportion ofaday,6=thewhole dayfrommorningtoevening).
Thesecondmainindependentvariablewastheemotionalbondbetweengrandparentsand
theirfirstadultchild.Thisvariablewasmeasuredbythreequestionsassessingthequalityoftheir
relationshipusingascaleadaptedfromtheIntergenerationalSolidarityInventory(Mangen,
Bengtson,Landry,&Sage,1988).Thethreequestionswere:(1)Takingeverythinginto
consideration,howclosedoyoufeeltoyoureldestchild?(2)Howmuchdoyoufeelthatyour
eldestchildwouldbewillingtolistenwhenyouneedtotalkaboutyourworriesandproblems?
(3)Overall,howwelldoyouandyoureldestchildgetalongtogether?Foreachquestion,scores
rangedfrom0to2(0=notatallclose/notatall/notverywell,1=somewhat
close/somewhat/somewhatwell,2=veryclose/verymuch/verywell).Anadditiveemotional
33
cohesionscorewascomputed,rangingfrom0to6,withhigherscoresrepresentingcloser
parent–childrelationships.Thereliabilitycoefficientforthisscalewas.82.
Thethirdkeyindependentvariablewaspatrilinealheritage,whichwasrepresentedbythe
genderofgrandparents,adultchildren,andgrandchildren,measuredas1=maleor0=female.
Controlvariables.Controlvariablesweregroupedintothreecategoriesbasedon
generation.Grandchild-levelvariablesincludedageatthetimeofthesurvey(inyears),gender,
educationlevel,andgeographicalproximitytograndparents.Educationlevelwasoriginally
measuredusingsevencategories:1=illiterate,2=elementaryschool,3=juniorhighschool,4
=seniorhighschool,5=vocationalcollege,6=communitycollege,and7=collegeorabove.
Thisvariablewasrecodedasacategoricalvariable(1=elementaryschoolorbelow,2=junior
highschool,3=seniorhighschoolor above).Geographicalproximitybetweenthegrandparent
andgrandchildwasmeasuredbyaskingthegrandparentwhereeachgrandchildlivedatthetime
ofthesurvey(1=thesamecity,2=thesameprovince,3=anotherprovince,4=another
country).
Controlvariablesattheparentlevelwereageatthetimeofthesurvey,gender,living
arrangement,andfinancialremittancetoelders.Livingarrangementwasmeasuredbyasking
grandparentswhethertheylivedwiththeirfirstchildduringthepast12months(1=yes,0=no).
Grandparentswerealsoaskedwhethertheyreceivedmoney,food,orgiftsfromtheirfirstadult
childduringthepast12months(1= yes,0=no).
Controlvariablesatthegrandparentlevelwereage,gender,education,maritalstatus,
income,andhealthstatus.Educationwasmeasuredasdescribedforgrandchildren.Duetoa
skeweddistributionofdataoneducationlevel,thisvariablewasrecodedasadichotomous
variableforgrandparents(1=illiterate,or2=someformaleducation).Maritalstatuswas
34
measuredusingfivecategories(1=marriedandlivingwithspouse,2=marriedbutnotliving
withspouse,3= widowed,4=divorced,5=nevermarried)andrecoded(1=married,or2=not
married).Incomewasmeasuredasthelogofthetotalincome(+1)thattherespondentreceived
fromworkorpensionduringthepreviousyear.Healthstatuswasmeasuredastheabilityto
perform15dailyactivities(1=nodifficulty,2=some difficulty,3=cannotperformwithout
help).Respondentsindicatedtheirlevelofdifficultyperformingactivitiesofdailyliving(ADLs)
suchasbathing,dressingorundressing,gettingoutofbed,walkingaroundtheroom,toileting,
andeating;instrumentalactivitiesofdailyliving(IADLs)suchaspreparingmeals,shopping,
doinghousework,takingthebusortrain,andmanagingmoney;andactivitiesrequiringphysical
strengthandmobility(STM)suchasliftinga10-kgbagofrice,climbingoneflightofstairs,
walking100meters,andstooping,crouching,orkneeling.Reliabilityofthisscalewas.97.The
summedscalerangedfrom15to45.
DataAnalysis
Descriptivestatistics(means,standarddeviations,andpercentagedistributions)forthe
dependent,independent,andcontrolvariableswerecalculatedtoobtainabasicunderstandingof
thecharacteristicsofthesample,aswellastestingHypothesis1.Hierarchicallogisticcluster
regressionmodelsofgrandparent–grandchildsupportexchangewereconductedtotestthesix
hypotheses.Inthelogisticclusterregressionmodels,controlvariablesforallgenerationswere
enteredfirsttotestHypotheses5and6,followedbythefirstmainindependentvariable
(previousgrandparentcaregivingofgrandchildren)totestHypothesis2.Theemotionalbond
betweengrandparentsandthemiddlegenerationwasenteredinthethirdmodel(Hypothesis3),
asweresomeinteractiontermsamongthethreekeyindependentvariables(Hypothesis4).All
analyseswereconductedusingSPSS15.0andSTATA11.0.
35
Becausegrandchildrenwerenestedinhouseholdsandanaverageoftwotothree
grandchildrenhadthesamegrandparentrespondent,theCL(meanscluster)commandin
STATAwasusedtoyieldstandarderrorsthatmoreaccuratelyreflectedtheresultsofthe
analysis,therebyavoidinginflatedstatisticalsignificancecausedbyclusteringofthesample.
Results
CharacteristicsofGrandparentsandGrandchildren
Table3.1
CharacteristicsofGrandparents(N=926)
Variable n % M(SD) Range
Age 71.69(8.24) 60–95
60–74 577 62.3
75+ 349 37.7
Gender
Male 440 47.5
Female 486 52.5
Maritalstatus
Married 587 63.4
Other 339 36.6
Education
Noeducation 711 76.8
Someeducation 215 23.2
Householdincome $379.54($676.44) $0–$11,200
Functionallimitations 18.89(6.81) 15–45
Genderofeldestchild
Male 462 49.9
Female 464 50.1
Ageofeldestchild 47.71(8.60) 26–73
Co-residencewitheldestchild
Yes 52 5.60
No 874 94.4
Emotionalbondwitheldestchild 7.4(1.52) 3–9
Table3.1showsthefrequencydistributionsofvariablesthatdescribedthe926
grandparentrespondentswhosefirstadultchildhadgrandchildren.Ofthesegrandparents,62.3%
werebetween60and74yearsold.Therewereslightlymorewomenthanmen(52.5%vs.47.5%,
respectively).Morethan60%ofthegrandparentsweremarried.Moreover,76.8%ofthe
36
grandparentsdidnothaveanyformaleducation.Themeanloghouseholdincomereceivedfrom
workandpensionduringthepreviousyearwas4.67(SD=4.0,range=0–14.5).Themajorityof
thegrandparentswereingoodhealthandhadfewfunctionallimitations(M=18.89,SD=6.81,
range=15–45).Thesamplewasevenlydividedbetweenpaternalandmaternalgrandparents
(49.9%vs.50.1%,respectively).Theaverageageoftheirfirstchildwas47.71(SD=8.6,range
=26–73)andnearlyallofthemdidnotlivewithrespondentsatthetimeofthesurvey(96.6%).
Finally,55.4%ofthegrandparentsreportedhavingaverycloserelationshipwiththeirfirstadult
child(M=7.4,range=3–9).
Table3.2
CharacteristicsofGrandchildren(N=1,843)
Variable n % M(SD) Range
Age 26.20(7.74) 16–48
Gender
Male 979 53.2
Female 861 46.8
Education
Primaryschoolorlower 364 19.8
Juniorhighschool 875 47.7
Seniorhighschoolorabove 588 32.5
Proximitytograndparents
Samecity 795 43.3
Notinsamecity 1,024 56.7
Grandparentcaregivingfrequency 4.60(2.45) 0–6
Never 675 36.6
Seldom 358 19.6
Onceamonth 21 1.1
Severaltimesamonth 50 2.7
Onceaweek 50 2.7
Severaltimesaweek 396 21.7
Everyday 277 15.2
Table3.2reportsthecharacteristicsofthe1,843grandchildren.Almosthalfofthe
grandchildrenwerebetween19and30yearsold(overallM=26.20,SD=7.74;range=16–48).
Slightlymorethanhalf(53.2%)weremen,andalmosthalf(47.7%)hadajuniorhighschool
37
education.Morethanhalf(56.7%)werenotlivinginthesamecityastheirgrandparents.Finally,
morethanthreefifthsofthegrandparents(62.8%)hadprovidedcaretotheirgrandchildreninthe
past,withameancaregivingintensityscoreof4.6(range=0–6).
Grandparent-GrandchildSupportExchange
Table3.3showsthesupportprovisionandreceptionfrequencybetweengrandparentsand
grandchildrenbasedongenderoftheparentgeneration.Moregrandparentsreceivedupward
supportfromgrandchildrenthanprovideddownwardsupporttograndchildrenintermsof
instrumental( χ
2
=2.03,p=.000),financial( χ
2
=7.53,p=.006),andemotionalsupport( χ
2
=
1.27,p=.000).Specifically,10.4%,43.2%,and93.8%ofeldersreceivedinstrumental,financial,
andemotionalsupportfromtheirgrandchildren,respectively,whereas4.3%,9.9%,and36.7%of
themprovidedthosevariousformsofsupporttotheirgrandchildren,respectively.Inaddition,
significantlyhigherpercentagesofeldersprovidedinstrumental(6.8%),financial(11.4%),and
emotionalsupport(41.3%)tograndchildrenoffirst-bornsonsthanfirst-borndaughters(2.1%,
8.5%,and32.3%,respectively).Slightlymoreeldersreceivedemotionalsupportfrom
grandchildrenoffirst-bornsons(94.8%)thanthoseoffirst-borndaughters(92.8%).
Table3.3
SupportExchangebetweenGrandparentsandGrandchildren
GrandparentSupport GenderofEldestChild
a
Total
b
Male Female
%(%oftotal) %
Supportprovision
Instrumental 6.8(3.2)*** 2.1(1.1) 4.3
Financial 11.4(5.4)* 8.5(4.5) 9.9
Emotional 41.3(19.6)*** 32.5(17.1) 36.7
Supportreception
Instrumental 11.5(5.5)† 9.4(4.9) 10.4***
Financial 43.7(20.8) 42.8(22.4) 43.2**
Emotional 94.8(45.1)* 92.8(48.7) 93.8***
a
Indicatorsinthiscolumnrepresentsignificancebetweengenderofeldestchild.
b
Indicatorsinthiscolumnrepresentsignificancebetweentotalsupportprovisionand
reception.
†p<.10,*p<.05,**p<.01,***p<.001
38
DeterminantsofGrandparent-GrandchildSupportExchange
Table3.4showstheresultsofclusterlogisticregressionsondeterminantsof
grandparent–grandchildsupportexchange.Previousgrandparentcaregivingwaspositively
associatedwithbothprovidingandreceivinginstrumental,financialandemotionalsupportat
statisticallysignificantlevels.Theassociationbetweenprovidingcareinthepastandcurrent
provisionofsupporttograndchildrenwasgreaterthanitsrelationshipwithreceivingsupport
fromgrandchildren.Themiddlegenerationplayedanimportantroleingrandparent–grandchild
supportexchange.Specifically,theemotionalbondbetweenthetwooldergenerationswas
positivelyassociatedwithgrandparents’providingemotionalsupport(OR=1.27,p<.05)and
receivingfinancialsupport(OR=1.47,p<.001)andemotionalsupport(OR=3.03,p<.001)
fromtheirgrandchildren.Patrilinealheritagealsoplayedanimportantrole;grandparents
generallyweremorelikelytoprovideemotionalsupporttograndchildrenoffirst-bornsons(OR
=1.95,p<.001)andprovideinstrumentalsupport(OR=1.90,p<.01),financialsupport
(OR=1.68, p<.01),andemotionalsupport(OR=1.98,p<.001)tograndchildreninfirst-bornsons’
families.Inaddition,emotionalclosenessbetweengrandparentsandfirst-bornsonswas
significantlypositivelyassociatedwithupwardanddownwardexchangeofallthreeformsof
supportwithgrandchildren.
Amongcontrolvariables,grandparents’agewasassociatedwithalowerprobabilityof
providingemotionalsupport(OR=0.72,p<.05)andreceivinginstrumentalsupport(OR=0.58,
p<.05)andemotionalsupport(OR=0.35,p<.01).Olderparticipantswhoweremarried,more
educated,andwhohadmoreincomeweremorelikelytoexchangesupportwiththeir
grandchildren.Ontheotherhand,grandparentswerelesslikelytoprovideallthreeformsof
supporttooldergrandchildrenandsignificantlymorelikelytoreceivefinancial(OR=1.08,p
39
<.001)andemotionalsupport(OR=1.07,p<.01)fromoldergrandchildren.Olderadultswho
livedclosertotheirgrandchildrenwerealsomorelikelytoprovideinstrumental(OR=1.52,p
<.001),financial(OR=1.22,p<.01)andemotional(OR=1.14,p<.01)support,andreceive
instrumentalsupport(OR=1.21,p<.05).However,elderswerelesslikelytoreceivefinancial
supportfromgrandchildrenwholivedcloser(OR=0.86,p<.05).
Table3.4
DeterminantsofSupportExchangebetweenGrandparentandGrandchild
Variable SupportProvision(OR) SupportReception(OR)
Inst Fin Emo Inst Fin Emo
Grandparent
Age 1.05 0.79 0.72* 0.58* 1.10 0.35**
Male 0.73 0.71 1.18 1.41 0.81 1.67*
Married 1.22 1.18 1.81*** 1.11 1.40** 0.84
Educated 1.27 1.70* 1.14 2.00*** 0.87 0.87
Functionalhealth 0.99 1.03 0.98* 1.02 0.99 0.99
Income(log+1) 1.15** 1.11*** 1.02 1.03 1.00 0.92*
Grandchild
Age 0.96 0.99 0.99 1.02 1.08*** 1.07**
Male 0.49 1.45 1.11 0.79 0.86 1.14
Education 0.97 1.08 1.10* 0.99 0.96 1.43***
Proximitytograndparents 1.52*** 1.22** 1.14** 1.21** 0.86* 1.07
Eldestchild
Age 1.00 0.91** 0.99 0.97 1.02† 0.97
Male 1.65 1.72 1.95*** 1.90** 1.68** 1.98***
Co-residence 1.69 0.61 1.45 3.81*** 1.53 1.10
Emotionalcloseness 1.00 1.15 1.27* 1.26 1.47*** 3.03***
Caregiving 1.32** 1.20** 1.17*** 1.10† 1.09* 1.09*
Interactionterms
Closenessxson 2.13** 1.94** 3.66*** 2.15*** 1.76*** 3.55***
Caregivingxson 1.06 0.92 1.00 1.08 0.97 1.00
Caregivingxgrandson 1.21 0.91 1.00 1.00 0.98 0.93
Note.Emo=emotional;Fin=financial;Inst=instrumental.
†p<.10,*p<.05,**p<.01,***p<.001
Discussion
Theresultsofthisstudyindicatedthatareciprocalexchangeofsupportoccursbetween
grandparentsandgrandchildreninruralChina.Thisisconsistentwithfamilysolidaritytheory,
whichpositsthatmutualityofsupportprovisionisanecessarypreconditionoffamilysolidarity
40
andtheremustbereciprocityofsupportbetweenthetwogenerations.Thisfindingalso
confirmedsocialexchangetheory,especiallythedirectexchangemodel,whichpositsthat
grandparentsandgrandchildrensimultaneouslyexchangedifferentgoodsandservices.These
findingsarealsoconsistentwithsocialcapitaltheory;bothgrandparentsandgrandchildren
representedpotentialsourcesofsupporttoeachotherduringdailylife,especiallyinruralChina,
wherethemiddlegenerationisoftenabsentandcannotengageintraditionalsupportexchange
withtheirelderlyparents.
Regardingthesupportbalancebetweenthetwogenerations,moregrandparentsinthis
studyreceivedinstrumental,financial,andemotionalsupportfrom,ratherthanprovidedsupport
to,grandchildren.Thus,Hypothesis1wasconfirmed.Thisresultisconsistentwithanearlier
studybyLanger(1990)butcontradictsfindingsreportedbyHoff(2007).Thismayreflectthe
uniquecontextofruralChina;whenformalsupportresourcesareverylimited,grandchildren
mayserveasimportantsourcesofsupport.Findingsalsoconfirmedtheneeds-basedmodel,
whichstatesthatolderadultshaveagreaterneedforsupportandgrandchildrenhaveincreased
capacitytoprovidesupportastheyreachadulthood.
Accordingtodatainthisstudy,moreintensecareprovidedbygrandparentsinthepast
wasassociatedwithanincreasedlikelihoodofprovidingsupporttotheirgrandchildren.This
findingindicatedthatoldergenerationscontinuetheirgivingbehaviortowardyounger
generations,eveninverylateadulthood(Kalmijn,2008),especiallytowardthosewhomtheyhad
caredforinthepast.Furthermore,resultsshowedthatelderswhoprovidedcaretograndchildren
inthepastwerealsomorelikelytoreceiveinstrumental,financial,andemotionalsupportfrom
thosegrandchildren,althoughthestatisticalsignificanceofthoseassociationswasnotstrong.
Therefore,thesecondhypothesiswaspartiallyconfirmed.Nonetheless,thisfindingoffered
41
supportforgeneralsocialexchangetheory,indicatingthatgrandchildrenreciprocatedsupport
theyhadreceivedearlier.Italsoreflectsthealtruisticorself-sacrificialaspectsofsocialsupport
theoryproposedbySilversteinandcolleagues(2002).Altruismorself-sacrificeamongelders
occurswhensupportisprovidedtograndchildrenwithnoexpectationofreceivingsupportfrom
theminreturn.Silverstein(2006)notedthatresearchersoftendepictintergenerationalsupport
transferbygrandparentsasanexpressionofaltruismandself-sacrifice.Thisholdstrueinrural
China,wherecaringforgrandchildrenisaculturallymandatedactivityviewedbyeldersasan
obligationandduty.Inthisstudy,theelderlyparticipantsprovidedcaregivingtotheir
grandchildrenwhentheywereyoungandcontinuedthesupportivebehaviortothose
grandchildrenevenaftertheygrewup.Yet,findingsrelatedtoreciprocalsupportexchangemay
alsoreflectthenormsofkinshipobligationamongyoungergenerationsinChina,orthe
expectationthatindividualsshouldcarefortheirfamilymembersintimesofneedregardlessof
thenatureoftheirrelationshipsinthepast.GrandchildreninruralChinahavebeentaughtor
socializedregardingthisnormofobligationorfilialpiety.Theymayprovidesupporttotheir
grandparentswhenthereisaneed,withlittleconsiderationofwhethertheyhadpreviously
receivedsupportorcarefromtheirgrandparents.
Theresultsofthisstudyindicatedthatthemiddlegenerationservesasabridgebetween
thefirstandthirdgenerations.Hypothesis3wasalsoconfirmed.Thisfindingwasconsistent
withpreviousresearchfindings(i.e.,Brown,2003;Hodgson,1992;Monserud,2008a,2008b;
Whitbeck,Hoyt,&Huck,1993).Findingswerealsoinaccordancewithfamilysystemstheory,
indicatingthatsupportexchangebetweengrandparentsandtheirgrandchildrenareembeddedin
familyties.Asindicatedbythisstudy,theemotionalclosenesswiththemiddlegenerationwas
positivelyassociatedwithgrandparent–grandchildsupportexchange;itisimportantforolder
42
adultstomaintainstrongtieswiththemiddlegenerationinordertohaveharmoniousinteractions
withtheirgrandchildren.Therefore,tobetterunderstandfamilydynamicsingeneral,itis
necessarytoconsidergrandparent–grandchildsupportexchangewithinthecontextofthefamily
unitandcombinethethreegenerationstogether.
Thisstudyalsoreflectedthefunctionofpatriarchallineageongrandparent–grandchild
supportexchangeinruralChina,wherefamiliesaremale-centeredandpatternsofresidence,
work,andinheritancegenerallypromoteapatrilinealpreference.Thisstudyshowedthat
grandparentsweremorelikelytoexchangesupport(bothprovidedandreceived)with
grandchildrenoftheirfirst-bornsonsthanfirst-borndaughters.Thus,ourfourthhypothesiswas
confirmed.BecausegrandparentsinruralChinatendtopaymoreattentiontoandcaremorefor
thechildrenofsonsratherthandaughters,thisresultsingreatersupportexchangewith
grandchildreninsons’households.Emotionalclosenessbetweenfirst-bornsonsandtheirparents
hadaverysignificantinfluenceongrandparent–grandchildsupportexchange.Thus,thisstudy
confirmedthesignificantrolesthatthemiddlegenerationandpatrilinealheritageplayin
determiningtheexchangeofsupportbetweengrandparentsandgrandchildren.
Amongotherfactors,grandparents’age,education,maritalstatus,andincome,aswell
grandchildren’sage,education,andgeographicalproximitywereassociatedwithgrandparent–
grandchildsupportexchange.Thisfindingagainconfirmedtheneeds-basedmodel;younger
generationsweremorelikelytoprovidecareandsupportwhentheirgrandparentsbecomeolder,
frail,ill,orinfinancialneed.Bythetimegrandchildrenreachadulthood,theirgrandparentstend
toneedmoresupportasaresultoffrailtyandmultimorbidity,andgrandchildrenhaveincreased
capacitiestoprovidethatsupport.Therefore,Hypotheses5and6wereconfirmedinthisstudy.
Itshouldbenotedthatthisstudyhassomelimitations.First,theanalysisreliedoncross-
43
sectionaldata;therefore,itwasnotpossibletodeterminecausalityorhowtheexchangeof
supportbetweengrandparentsandgrandchildrenchangedovertime.Alongitudinalstudywould
enhanceunderstandingofhowgrandparent–grandchildsupportchangesoverthelifestagesof
familymembers.Second,themeasuresofeachdimensionofintergenerationalsupportexchange
betweengrandparentandgrandchildrenwerebasedonsinglequestionnaireitems.Third,it
shouldalsobenotedthattheanalyseswerelimitedtothechildrenfromeachrespondent’seldest
child.Supportexchangedbetweengrandparentsandeachoftheiradultchildren,aswellasthe
childrenofeachoftheiroffspring,candiffer.Becausetheeldestchildandhisorherfamily
generallyreceivethemostattentionfromgrandparents,itisnotadvisabletogeneralizefindings
inthisstudyregardingtheroleoftheeldestmemberofthemiddlegenerationin
intergenerationalsupportexchangetootheradultchildren.Finally,thestudysamplewasdrawn
fromalow-income,ruralregioninChinawherestrongpatriarchaltraditionspersist.Therefore,
resultsmaynotbegeneralizabletourbanelders,whoseinvestmentoftimeandresourcesintheir
grandchildrenmaydeviatefrompatternsobservedinAnhuiProvince.
44
Chapter4
ChangePatternsofInstrumentalSupportfromAdultChildrenandGrandchildrenamong
RuralChineseGrandparents
Introduction
Socialsupportisimportantforhealthandwell-beingacrossthelifecourse,butitis
especiallyimportantinlatelifeduetophysicalandlife-coursechanges(Zimmer&Kwong,
2003).AlthougheldersinWesterncountriesreceivesocialsupportfromavarietyofsources,
mostolderadultsinChinareceivesupportprimarilyfromtheiradultchildren(Cheng&Chan,
2006;Pei&Pillai,1999;Sun,2002).ThissituationisevenmorepronouncedinruralChina,
whereolderresidentsreceivefewerpublicsocialservicesandlessfinancialsupportthantheir
urbancounterpartsandhavetorelyonfamilymembersforsupport(Silverstein,2006;Zimmer&
Kwong,2003).
However,formanyruralresidents,supportingolderparentshasnotbeeneasybecauseof
lowerstandardsoflivingresultingfromlaggingeconomicandinfrastructuredevelopmentin
China’sruralareas.Inadditiontothelackoffinancialresourcesamongruralfamilies,alackof
affordablehealthcarecoupledwithurbanmigrationcouldfurtherthreatenfragilefamilysupport
systemsinruralcommunities(S.Chen,2009).Labormigrationamongtheruralworking-age
populationhasledtoagreatergeographicalseparationbetweengenerations,resultingin
increasingopportunitiesforolderpeopletolivewithandreceivesupportfromtheir
grandchildren(T.Z.Chen,2009;Silversteinetal.,2006).Therefore,grandchildrenrepresent
importantsocialsupportresourcesandcapitalforolderadultsinneed.Olderresidentsinrural
Chinamayconcurrentlyreceivesupportfrombothadultchildrenandgrandchildren.Support
45
fromgrandchildrenmayplaydifferingrolesthansupportfromtheparentgeneration.For
example,grandchildrencanprovidesupporttooldergrandparentstosupplanttheabsenceof
adultchildrenwhocannotprovidesupporttotheiragingparentsduetomigration,illness,or
death.Supportfromgrandchildrencansupplementsupportfromthemiddlegenerationthatistoo
weakorinsufficienttomeettheneedsofolderadults.Supportfromgrandchildrencanalso
complementsupportfromthemiddlegenerationtoenhanceorimprovethequalityoflifeof
olderadults.However,thesehypotheticalfunctionsofsupportfromgrandchildrenhavenotbeen
testedbypreviousstudies.Theobjectivesofthisstudyweretoexaminepatternsof
intergenerationalsupportfrombothadultchildrenandgrandchildren;testhowmany
combinationsortypesofinstrumentalsupportareprovidedtoolderadultsbythetwoyounger
generations;toexplorewhethergrandchildrensubstitute,supplement,orcomplement
instrumentalsupportprovidedtoolderadultsbytheparentgeneration;andexaminefactors
associatedwiththeseformsofsupport.
Increasinglifeexpectanciescorrespondinglyincreasetheaveragenumberofyearsthat
eachpersonspendsasagrandparentandgrandchild.Decreasingfertilityratesresultindecreased
familysizeandanincreaseinnuclearfamilies.Withthesedemographicchangesinfamily
structures,verticalfamilystructureswillbecomemoreprevalentthanhorizontalstructures(Hoff,
2007).Supportexchangeamongthreegenerationswillbecomeincreasinglyimportantasthese
familyarrangementsbecomemorecommonandthereforeisdeservingofincreasedresearch
attention(Grundy&Henretta,2006).Thisstudyofintergenerationalsupportbetween
grandparents,parents,andgrandchildrenhasspecialsignificanceincontemporarysociety.
46
InstrumentalSupportfromYoungerGenerations
Althoughtheconstructofintergenerationalsupporthasbeenoperationalizedinavariety
ofways,instrumentalsupporthasremainedanunderlyingdimension(Antonucci,1990;Chen&
Silverstein,2000;Cong&Silverstein,2004).Instrumentalsupportreferstopracticalortangible
formsofsupport,suchashelpwithpersonalcareandhouseholdchores.Instrumentalsupport
withhouseholdchoresandpersonalcareisoftencriticalforpeopleinolderlifebecauseof
physicalandlife-coursechangesandhasbeenviewedasabasicformoffilialpiety(Chow,
2001).Instrumentalsupporthasbeenfoundtosustainorimprovethepsychologicalandphysical
well-beingofolderChineseadults,especiallyintimesofillnessanddistress(Cheng&Chan,
2006;Cong&Silverstein,2004;Silverstein&Bengtson,1994).Peoplewhoreceive
instrumentalsupportlivelongerthanthosewhodonot,especiallyiftheyhavephysical
impairments(Ross&Mirowsky,2002).Therefore,receivinginstrumentalsupportisvery
importantforolderadultsanddeservesattentionsfromresearchers.
Althoughseveralstudieshaveexaminedsupportexchangebetweenolderadultsandtheir
children,fewhaveaddressedsupportamongthreegenerations.Amongstudiesonthree
generations,mostfocusedsolelyonintergenerationalrelationships.Forexample,Monserud
(2008a,2008b,2010)examinedwhetherearlierpatternsofclosenessbetweengrandchildrenand
grandparents,currentrelationshipsbetweenthemiddlegenerationandgrandchildrenor
grandparents,andchangesinparentaltieswithothergenerationswereassociatedwithcurrent
grandparent–grandchildcloseness.Goodman(2003)examinedaffectivebondsamong
grandmothers,parents,andgrandchildrenandtheireffectsonthewell-beingofgrandmothers
(depressionandlifesatisfaction).Althoughthisstudygeneratedeighttriadtypesthatinvolved
theintricatedynamicsofcoalitionsandallianceswithinathree-personsocialsystem,thisstudy
47
andothers(i.e.,Monserud,2008a,2008b,2010)didnotaddresssupportexchangeamongthree
generations.
Accordingtoanexhaustivereviewoftheliterature,onlytwostudieshaveexploredthree-
generationsupportinthefamilycontext.GrundyandHenretta(2006)investigatedhowmiddle-
agedwomenintheparentgenerationresolvedconflictingexpectationsfromthetwoouter
generationsintheUnitedStatesandGreatBritain.Theyfoundthatapproximatelyone-thirdof
thewomenprovidedhelptomembersofbothgenerations.AnotherstudybyHoff(2007)
examinedpatternsofmutualinstrumentalandfinancialsupportbetweenolderGermanadults
andboththeirchildrenandgrandchildren.Thisstudyfoundthetwoyoungergenerationswere
farlesslikelytoreciprocatefinancialsupporttheyreceivedfromoldergenerationwith
instrumentalsupport,especiallygrandchildren.Infact,grandparentsreceivedhardlyany
instrumentalassistancefromtheirgrandchildren(Hoff,2007).
Althoughbothstudiesinvolvedsupportacrossthreegenerations,theydidnotexamine
therelationshipofsupportprovidedbyadultchildrenwithsupportprovidedbygrandchildren.
Astheolderadultsaged,howdidthetwoyoungergenerationsaltertheirsupportbehaviors?Was
theprovisionofinstrumentalsupportbygrandchildrentotheirgrandparentsbasedonsupport
providedbytheparentgeneration?Thisstudyaddressedtheseresearchquestionsbyexamining
howthetwoyoungergenerationsconcurrentlyprovidedinstrumentalsupporttoolderadultsto
determinepotentialtypesofsupportpatternsandassociatedfactors.Giventheincreasingnumber
offamilieswiththreelivinggenerations,thistypeofresearchiscriticaltogainadeeper
understandingoffamilydynamicsinmodernsociety.
48
ThePresentStudy
Thescarcityofpreviousresearchondynamicsupportinvolvingthreegenerations
highlightstheneedforanexplorationofhowthetwoyoungergenerationsprovidesupportto
olderadults,aswellashowchangesinsupportpatternsamongtheyoungergenerationsaffect
eachother.Inthisstudy,thefirstresearchpurposewastoexaminethepatternofchangein
instrumentalsupportfromgrandchildrenandtheparentgeneration.Itwaspossibleforsupport
frombothadultchildrenandgrandchildrentoincrease,decrease,remainthesame,ornotexist.
Therefore,therewere16potentialpatternsofsupportchangeovertime.Todeterminethemajor
typologyorpatternofsupportchangeamongthetwoyoungergenerations,thisstudytested
whetherinstrumentalsupportfromadultchildrenandgrandchildrensubstituted,supplemented,
orcomplementedeachother.Apreviousstudyhadshownthatwhenthemiddlegenerationhad
healthissuesorwascriticallyill,grandchildrenbecamethenextsourceofsupportforolder
adults(Smith,2011).Thesecondresearchpurposewastoexplorewhatelementswereassociated
withmembershipineachtypologythatwasgeneratedbasedonpatternsofinstrumentalsupport
providedbyadultchildrenandgrandchildrentoolderadultsover9years.
Methods
Data
FourwavesofdatafromtheAnhuistudywereusedforthisstudy.Thedataandsampling
informationhavebeendescribedintheintroductorymethodportionofthisdissertation.Given
thatmissingvalueswererelativelyrareinthedata(amaximumoflessthan1%amongall
variables),Iusedmeansubstitutionasanimputationstrategy.Changescoresbetweenadjacent
waveswerecalculatedandpulledtogetherforfinalanalysis.Afterexcludingindividualswho
failedthecognitivetestatallfourtimepoints,thefinalpooledsampleconsistedof2,248
49
observationsofolderadultswhoreceivedinstrumentalsupportfromtheirchildrenor
grandchildren.CharacteristicsoftheoperatingsampleatbaselinearedescribedinTable4.1.
Table4.1
SampleCharacteristicsofOlderAdultsinRuralChinaatBaseline(N=731)
Variable % M(SD) Range
Age 68.03(6.19) 60–95
60–74 62.5
75+ 37.5
Gender
Male 49.1
Female 50.9
Maritalstatus
Married 63.4
Notmarried 36.6
Education
Noeducation 75.9
Someeducation 24.1
Income
Yes 57.0
No 43.0
Self-ratedhealth 2.95(0.86) 1–4
Functionallimitations 20.89(7.38) 15–45
Livingarrangement
Alone 21.3
Withothers 78.7
Numberofchildren 3.88(1.73) 0–10
Geographicalproximitytochildren 2.81(1.42) 1–7
Emotionalbondswithchildren 7.59(3.53) 3–9
Numberofchildrenhelpedwithbabysitting 0.94(1.18) 0–7
Measurement
Receptionofinstrumentalsupportfromadultchildrenandgrandchildrenwasmeasuredat
eachwave.Thisvariablewasmeasuredbyaskingrespondentswhethertheyreceivedany
householdhelpfromtheirgrandchildren(e.g.,cleaninghouse,washingclothes,andwashing
dishes)duringthepast12monthsduetopoorhealth(1=yes,0=no).Ifrespondentsanswered
affirmatively,theywereaskedfromwhomtheyreceivedinstrumentalsupport.Participantscould
identifyuptonineindividualswhoprovidedsupport.Respondentswerethenaskedhowoften
50
theirgrandchildrenprovidedhouseholdsupport(1=seldom,2=severaltimespermonth,3=at
leastonceperweek,4=everyday).Aftercombiningscoresforthesetwoquestions,anoverall
scoremeasuringthelevelofinstrumentalsupportfromchildren(sons,daughters,daughters-in-
law,sons-in-law)andgrandchildren(includinggrandchildren-in-law)rangedfrom0=neverto4
=everyday.Changescoresofinstrumentalsupportfromadultchildrenorgrandchildrenwere
calculatedbetweenadjacentwavesandrecodedintofourcategories:nosupport,increasein
support,decreaseinsupport,andstaticsupport.Eachcategorywasmeasureddichotomously(1
=yes,0= no).Allchangescoresbetweenadjacentwaveswerethenmergedtogether.The
distributionofitemsmeasuringinstrumentalsupportpatternsisshowninTable4.2.
Onaverage,37.0%,31.0%,28.7%,and3.3%ofolderadultsreceivednosupport,
increasedsupport,decreasedsupport,andstaticsupport,respectively,fromtheiradultchildren.
Changesininstrumentalsupportfromgrandchildrenaveraged83.8%,7.3%,8.6%,and0.2%,
respectively.
Table4.2
DistributionsofChange ofSupportfromAdultChildrenandGrandchildrenover
9YearsinRuralChinaby Percentage(N=2,248)
ChangePattern 20012003 20032006 20062009 M
Supportfromadultchildren
None 30.49 38.18 44.21 37.00
Increase 32.86 31.72 27.61 31.00
Decrease 33.19 26.62 25.18 28.70
Static 3.46 3.48 3.00 3.30
Supportfromgrandchildren
None 79.24 86.69 86.55 83.80
Increase 8.22 5.72 7.87 7.30
Decrease 12.00 7.46 5.58 8.60
Static 0.54 0.12 0.00 0.20
Personalcharacteristicswereincludedascontrolvariablesattwolevels:grandparents
andparents.Grandparent-levelvariableswereage,gender,maritalstatus,educationlevel,self-
51
ratedhealth,functionalhealth,income,livingarrangement(aloneorwithothers),numberof
children,emotionalbondwithchildren,andnumberofgrandchildrentheycaredfor.Parent-level
variablesweregeographicproximitywitholderadults(cohabitatedinthesamevillageornot)
andgender(sonsonly,daughtersonly,bothsonsanddaughters).Mostofthesemeasurements
havebeendescribedinthefirststudy.Becausemostofthosevariableschangedacrossfour
waves,theirmeanvalueswereusedforfinalanalysis.
DataAnalysis
Todeveloppotentialtypologiesofpatternsofchangeininstrumentalsupportprovidedto
olderadults,latentclassanalysis(LCA)wasusedtogroupsupportfromadultchildrenand
grandchildrenintotypes.LCAisausefulstatisticalstrategyforexaminingmultidimensional
combinationsofsocialrelationships.Itcanberegardedasthecategoricalcounterpartoffactor
analysis.However,thetwodifferinthatfactoranalysisismoreconcernedwiththestructureof
variables(i.e.,variable-centered)whereasLCAfocusesonthestructureofcases(i.e.,person-
centered).LCAallowsresearcherstomeasureassociationsamongobservedcategorical
indicatorsthroughasetofunobservedlatentclasses(McCutcheon,1987).Thebasicpremiseof
thestudyoflatentclassesisthatthecovariationobservedamongthemanifest(observed)
variablesisduetoeachmanifestvariables’relationshipwiththelatentvariable—acondition
knownaslocalindependence.
Foragivenlatentclassstructure,twotypesofstatisticsaregenerated:conditionallatent
classprobabilities(CLCP)andlatentclassprobabilities(LCP).CLCPsandLCPsaretwo
importantparametersestimatedinlatentclassanalysis.LCPsdescribethedistributionoflatent
classesacrossapopulationandsumto1,indicatingthatinadditiontobeingmutuallyexclusive,
theclassesarealsoexhaustive.CLCPsdescribecharacteristicsofidentifiedlatentclassesbased
52
onthedistributionsofobservedvariablesamongindividualsassignedtoeachclass.Those
probabilitiesareusedtoprofileanddefinethelatentclasses.Theyarecomparabletofactor
loadinginfactoranalysis,representingameasureofthedegreeofassociationbetweeneach
observedvariableandeachlatentclass.CLCPsenableresearcherstocharacterizethenatureof
classtypesbyindicatingtheprobabilitythatanindividualinalatentclasswillscoreaparticular
wayonanobservedmeasure.HighCLCPsindicatethatmembersofagivenclassaremore
likelytoendorseaparticularresponse(SeeM.Guo,2011).Inthisstudy,bothsetsofstatistics
weretestedfortheirinvarianceacrossthepopulationbeingstudied.
Thegoodness-of-fitofLCAmodelswasevaluatedusingthelikelihoodratiochi-square
teststatistic(L2),theBayesianinformationcriterion(BIC),andentropyscores.TheL2test
examinesdiscrepanciesbetweenatheoreticalmodelandtheobserveddata.Aswiththestandard
chi-squaretest,anonsignificantL2indicatesagoodfitofthemodeltodata.TheBIC(Raftery,
1986)isusedwhenselectingthebest-fittingmodelamongcompetingmodels.AsmallerBIC
indicatesabettermodelfit.Higherentropyscores(nomorethan1)indicatebettermodelfit.
Becausechangescoresbetweenadjacentwaveswerepooledtogether,thefinal2,248
observations(changescores)inthisstudywerenestedwithin731olderadults.Asaresult,LCA
wasconductedusingcomplexmixtureanalyticproceduresinMPlus.Thecommand
“Type=ComplexMixture”wasusedtoobtainstandarderrorscorrectedfornon-independenceof
observationsinmixturemodels(SeeM.Guo,2011).Toidentifytypesofinstrumentalsupport
providedtoolderadults,alatentclassmodelwasdevelopedwithonlyonelatentclass(no
relationbetweenmanifestindicators),andadditionalclasseswereaddedoneatatime,with
checksformodelfitandsignificance.Afteranoptimalmodelwasidentified,Iexaminedthe
numberandmeasurementstructureofthelatentclassesandtheprobabilitydistributionofclass
53
representation.Theselectionofthebestmodelwasbasedonthegoodness-of-fitindicators
discussedearlier.Multinomialregressionwasusedtotestforsociodemographicdifferencesin
theoddsofclassmembership.Themultivariatemodelswereestimatedintwostages,with
individualsociodemographicfactorsatthegrandparentlevelandtheparentleveladded
separatelyintothemodel.Thisallowedadeterminationofwhethertheinstrumentalsupport
changepatternswereexplainedsolelybytheneedsofolderadultsorbyfactorsatboth
generationallevels.
Results
ChangePatternsofInstrumentalSupportfromTwoYoungerGenerations
Table4.3
ModelFitforLatentClassAnalysisofInstrumentalSupportamong
RuralOlderChineseAdults
Classes L2 df p AIC BIC Entropy
1 6243.56 247 .000 14611.1 14657.4
2 4298.09 238 .000 12683.6 12781.9 .482
3 1053.05 229 .000 8397.6 8508.9 .811
4 1595.39 220 .000 8405.9 8561.3 .547
5 1582.00 211 .000 8492.6 8747.1 .542
Note.L2=likelihoodratiochi-square;AIC=Akaikeinformationcriterion;BIC=
Bayesianinformationcriterion.
Thefirstgoalofthisstudywastodevelopawell-defined,descriptivelymeaningful,and
generalizabletypologyofchangepatternsininstrumentalsupportfromadultchildrenand
grandchildren.Aseriesoflatentclassstructuresweretestedbyaddingclassesandobserving
changesingoodness-of-fitstatisticsforeachsuccessivemodel.Table4.3showsthatL2statistics
weresignificantforallfivemodels.Addinguptoeightmoreclassesdidnotyieldnonsignificant
L2statistics(resultsnotshown).Severalscholarshavearguedthatalthoughlog-likelihood
valuescanindicatetheappropriatenumberofclasses,theL2testshouldnotbeusedasthesole
decisioncriterionbecauseitdoesnotfeaturetheusuallarge-samplechi-squaredistributionand
54
theobtainedp-valuemaynotbeaccurate(Muthén,2004;Nylund,Asparouhov,&Muthén,
2007).
Asalternatives,BICandentropyscoresweregivenmoreweightwhencomparingnested
latentclassmodelsandchoosingthebestmodel.Table4.3showssuccessivedecreasesinBIC
valueswitheachadditionallatentclass,indicatingrelativeimprovementsinfituptothree
classes.TheBICincreasedinthefour-classmodel,indicatingthesuperiorityofthethree-class
model.Further,entropyscoresincreaseduntiladdingthethirdclass,indicatingrelative
improvementinfit.Asaresult,thethree-classmodelwasselectedforfurtheranalysis.
IusedtheCLCPsoftheacceptedmodeltoprofileeachlatentclass.Assuggestedby
McCutcheon(1987),probabilitieswereinterpretedinbothabsoluteandrelativetermswhen
definingtheclasses.Table4.4displaystheestimatesoftheconditionallatentclassprobabilities
ofthethree-classmodel.Basedonthepatternsoftheseprobabilities,Iassignedthelabels
definedinTable4.4todescribethethreelatentclasses.Theywere:decreasedsupport(13.1%),
increasedsupport(19.8%),andlittlesupport(67.1%).
Table4.4
ConditionalLatentClassProbabilitiesofaThree-ClassModelof
IntergenerationalSupportPatternsinRuralChina
Pattern Class1
(decrease)
Class2
(increase)
Class3
(little)
Supportfromadultchildren
None .00 .00 .42
Increase .21 .81 .27
Decrease .73 .12 .28
Static .06 .07 .03
Supportfromgrandchildren
None .04 .30 .93
Increase .16 .54 .03
Decrease .78 .15 .04
Static .01 .02 .00
Note.Latentclassprobabilitiesgreaterthan.4(inbold)areconsideredrelativelyhigh.
55
AsshowninTable4.4,thefirstclasshadhighprobabilitiesregardingadecreasein
instrumentalsupportfrombothadultchildrenandgrandchildren,indicatingthatinstrumental
supportfromgrandchildrendecreasedwhensupportfromthemiddlegenerationdecreased.It
alsoindicatedthattherelationshipsbetweenolderadultsandthetwoyoungergenerationswere
somewhatremote.Thesecondclasshadhighprobabilitiesregardingincreasesininstrumental
supportprovidedbythetwoyoungergenerations,suggestingacloserelationshipbetweenolder
adultsandthetwoyoungergenerations.Thethirdclasshadhighprobabilitiesforreceivinglittle
instrumentalsupportfrombothadultchildrenandgrandchildren,eventhoughtherewerelow
probabilitiesofincreasing(27%)ordecreasing(28%)supportfromthemiddlegeneration.
Generallyspeaking,olderadultsinthisclassdidnotreceivemuchinstrumentalsupportfrom
youngergenerations,suggestingindifferentrelationshipsbetweenolderadultsandtheir
grandchildren.
PredictorsofInstrumentalSupportChangePatterns
Predictorsoflatentclassmembershipwereexploredusingmultinomiallogistic
regressions;thefirstclass(decreaseinsupport)wasthereferencegroup.Thepurposewastotest
whetherthesociodemographicfeaturesofolderadultsandadultchildrenpredictedthederived
changepatternsofintergenerationalinstrumentalsupportforolderadults.Individual
characteristicsofolderadultswereaddedinthefirststep,followedbyfactorsattheparentlevel.
Table4.5showstheresultsofthemultinomiallogisticregression.
Theresultsshowedthatolderparticipantsweremorelikelytobeinthe“increasein
support”class(OR=1.97,p<.01),indicatingthatolderadultsweremorelikelytoreceive
increasedinstrumentalsupportfrombothadultchildrenandgrandchildren.Incontrast,older
participantswerelesslikelytobeinthe“littlesupport”class(OR=0.97,p<.001).Olderadult
56
womenwere130%morelikelytobeinan“increaseinsupport”andreceivedincreased
instrumentalsupportfromthetwoyoungergenerations(OR=2.31,p<.000).Grandmothers
werealsolesslikelytoreceivelittlesupportfromtheirchildrenandgrandchildren(OR=0.89,p
<.01).Thehealthstatusofolderadultshadsomeinfluenceonclassmembership.Olderadults
withworseself-ratedhealthandmorefunctionallimitationsweremorelikelytobeinthe
“increaseinsupport”group,indicatingtheyreceivedincreasedinstrumentalsupportfromtheir
childrenandgrandchildren.Thoseinthe“lessinsupport”group,whoreceivedlittlesupport
fromyoungergenerations,weremorelikelytobeinbetterhealth.Grandparentswithmore
childrenwerealsomorelikelytobeinthe“increaseinsupport”group(OR=1.03,p<.05).
Table4.5
MultinomialOdds RatiosPredictingClassMembershipinaPartiallyConstrainedThree-
ClassModelofIntergenerationalSupportPatternsinRuralChina
Predictor Class2
(increaseinsupport)
Class3
(littleinsupport)
G1Age 1.97** 0.87***
G1Female 2.31** 0.89**
G1Educated 0.85 1.07
G1Married 0.97 1.03
G1Livewithothers 1.02 0.73
G1Self-ratedhealth 0.96** 1.12*
G1Functionallimitations 1.07* 0.91*
G1Numberofchildren 1.03* 0.41
Emotionalbondwithadultchildren 1.11* 0.34**
Geographicproximitywithadultchildren 1.32** 0.97*
Numberofchildrenhelpedwithbabysitting 1.49* 0.97
*p<.05,**p<.01,***p<.001
Intermsofparent-levelvariables,olderadultswithstrongeremotionalbondswiththeir
adultchildrenweremuchmorelikelytobeinthe“increaseinsupport”class(OR=1.11,p
<.05).Inaddition,iftheylivedinthesamevillageastheiradultchildren,theywerealsomore
likelytoreceiveincreasedinstrumentalsupportfromtheiradultchildrenandgrandchildren(OR
=1.32,p<.01).Onthecontrary,participantswithweakeremotionalbondsorwholivedfarther
57
awayfromadultchildrenweremuchmorelikelytoreceivelessinstrumentalsupportfromtheir
childrenandgrandchildren.Inaddition,olderparticipantswhoprovidedcaretotheir
grandchildrenweremorelikelytoreceiveincreasedinstrumentalsupportfrombothadult
childrenandgrandchildren(OR=1.49,p<.01).
Discussion
Thisstudyinvestigatedthenatureandextentofinstrumentalsupportchangepatterns
amongolderadultsincontemporaryruralChina.Thelatentclassstructureoftwocoresourcesof
instrumentalsupportforolderadultswithinthefamilycontext—fromadultchildrenand
grandchildren—wereexaminedtodescribeandpredictthechangepatternsofinstrumental
supportamongruralolderChineseadults.Usingchangescoresfromafour-wavelongitudinal
datasetoftheAnhuistudy,Idemonstratedthatathree-classmodelbestrepresentedthe
dimensionsofsupportfromadultchildrenandgrandchildren.Onemajorpurposeofthisstudy
wastodeterminetheunderlyingpatternsofchangeininstrumentalsupportprovided
concurrentlytoolderadultsbyadultchildrenandgrandchildren.Thisstudyrevealedthreebasic
typesofsupportexchange—decreasesinsupport,increasesinsupport,andlittlesupport.These
threetypesofsupportchangepatternsalsoreflectedthepossiblerelationshipbetween
grandparentsandgrandchildren.Forexample,“increaseinsupport”typeindicatedaclose
relationshipbetweengrandparentsandgrandchildren,and“decreaseinsupport”and“little
support”suggestedaremoteandindifferentgrandparent-grandchildrelationshiprespectively.
Althoughthesepatternsregardingconcurrentprovisionofinstrumentalsupportfromthetwo
youngergenerationswereidentified,comparisonswithsupportpatternsinWesternsocietiesor
othercountriesisdifficultbecauseofthelackofpreviousresearchonthistopic.Nonetheless,
58
thisstudycontributedtotheknowledgebasebyidentifyingthreetypesofinstrumentalsupport
changepatternsamongruralolderChineseadults.
Regardinginstrumentalsupportfromgrandchildrenversusadultchildrenovertime,this
studyshowedthatinstrumentalsupportfromgrandchildrenwasnotprevalentwhenthe
grandparentshavesuchneeds,suchasduringillness.Thisfindingisconsistentwithfindingsby
Hoff(2007)thatgrandparentsreceivedlittleinstrumentalassistancefromtheirgrandchildren.In
China,culturalandsocietalnormsemphasizefilialpietyorobligationsofchildrentotheir
parents;similartotheircounterpartsinWesternsocieties,grandchildrenmaynotfeelas
compelledtoprovidehelptotheirgrandparentsastotheirparents(Smith,2011).Inaddition,
adultgrandchildrenmayfaceseveralcompetingdemands.Adultgrandchildrenmaybe
beginningtoformtheirownfamilies,oftenwithyoungchildrenorrelativelynewmarriages,and
mayhavetoprovidecaretoanailingparent.Adultgrandchildrenarealsomorelikelytobe
enteringthejobmarket,withlessvacationtimeandlesssecurityorjobtenurepreventingthem
fromtakingtimeawayfromworktocareforgrandparents.
Interestingly,patternsofinstrumentalsupportfromgrandchildrenandadultchildren
mirroredeachother.Inotherwords,asinstrumentalsupportfromadultchildrendecreasedor
increased,supportfromgrandchildrencorrespondinglydecreasedorincreased.Supportfrom
grandchildrenmaystemfromfeelingsofresponsibilitytowardtheirparentswhoaskedthemto
helpcarefortheirelders.Grandchildrenmayserveashelperstotheirparents.Anotherpossible
explanationmightbethatadultchildrenandgrandparentsmaylivegeographicallyclosetoolder
adultsandthusserveasateamoffamilycaregivers,exhibitingsimilarbehaviorsandpatterns
regardinginstrumentalsupport.However,thisfindingmayalsoindicatetheimportanceofthe
middlegeneration.Themiddlegenerationservesasanimportantmediatinglinkbetween
59
grandchildrenandgrandparents,eitherfacilitatingorhinderingtheinteractionandqualityof
theirrelationship(Smith,2011).Previousstudieshaveshownthatmembersofthemiddle
generationwhoareclosetothegrandparentgenerationaremorelikelytohavechildrenwhoare
alsoclosetotheoldergeneration(Fruhauf,Jarrott,&Allen,2006;Fruhauf&Orel,2008;King
&Elder,1995;Rossi&Rossi,1990).Thisstudyconfirmedthatsupportfromthemiddle
generationservedasamodeloranexamplethatinfluencedinstrumentalsupportfrom
grandchildren.
Anotherpurposeofthisstudywastotestthefunctionofinstrumentalsupportfrom
grandchildrenbycomparingitwithsupportfromadultchildren.Iexpectedtofindthat
grandchildrenplayedacomplementaryroleinintergenerationalsupportexchange,ahypothesis
confirmedbythisstudy.Ialsoexploredwhethergrandchildrensupplementedorsubstituted
insufficientorabsentinstrumentalsupportfromthemiddlegeneration.However,thisstudydid
notconfirmthisfunctionofgrandchildren.Wheninstrumentalsupportfromthemiddle
generationdecreasedduetovariousreasons,supportfromthegrandchildrendidnotincrease,at
leastnotenoughtoconstituteaseparateclassinLCA.Onthecontrary,instrumentalsupport
fromgrandchildrendecreasedalongsidedecreasingsupportfromthemiddlegeneration.Many
reasonsmightexplainthisfinding.Olderadultsmaynothaveneededinstrumentalsupportin
termsofpersonalcareorassistancewithhouseholdchores;bothyoungergenerationsmayhave
livedfartherawayfromparticipants;ortheemotionalrelationshipsbetweenolderadultsand
youngergenerationsmayhavebeenpoor.
Thisstudyalsoexaminedassociationsbetweenthesociodemographiccharacteristicsof
thetwooldergenerationsanddifferentsupportpatterns.Findingsshowedthatgrandparents’age,
gender,healthstatus,andnumberofchildren,aswelltheiremotionalclosenessandgeographical
60
proximitywiththeiradultchildrenandthenumberofchildrentheyhelpedwithbabysitting,were
associatedwithreceivingincreasedsupportorlittlesupport.Morespecifically,participantswho
wereolder,female,inworsehealth,whohadmorechildren,strongeremotionalbondsorcloser
geographicalproximitywithadultchildren,andwhohelpedmoreadultchildrenbytakingcare
ofgrandchildrenweremorelikelytoreceiveincreasedsupportovertime.Thisindicatedthat
olderage,beingfemale,andhavingmorehealthlimitationsputolderadultsinamorevulnerable
positionandincreasedtheirneedforinstrumentalsupportfromyoungergenerations.This
findingisconsistentwiththeneeds-basedmodeldescribedinChapter3,indicatingthattheneeds
ofolderadultsmaybeanimportantfactordrivingtheprovisionofsupportbyadultchildrenand
grandchildren.Geographicproximityandemotionalbondswerealsofoundtobeassociatedwith
patternsofintergenerationalsupport.Geographicproximitymayreduceisolationamongolder
adultsandtheiradultchildren,whereasemotionalbondsmayserveasacatalystfortheyounger
generationintermsofwillingnesstoprovidesupport.Reciprocityalsoplaysarolein
intergenerationalsupportexchange.Adultchildrenwhosechildrenwerecaredforbytheir
elderlyparentsandgrandchildrenwhohadahistoryofbeingcaredforbytheirgrandparents
duringchildhoodmayfeelobligatedtoreciprocatethatcarewhenolderadultsareinneed.
Limitations
Thisstudyhadsomelimitationsthatmustbeaddressedbeforeanyconclusionsare drawn.
First,usingtwodimensionsofinstrumentalsupportfromadultchildrenandgrandchildrenmight
havelimitedthisstudy.Addingmoreitemsmayincreasethevalidityoflatentclassanalysis,for
example,provisionofconcurrentinstrumentalsupporttoadultchildrenandgrandchildrenby
grandparents.Receivingandprovidinginstrumentalsupporttospouses,anotherimportantform
ofsupportexchangeamongolderadults,shouldalsobeconsidered.Inaddition,changescoresof
61
adjacentwavesaresometimeshardtoexplainifthechangesover9yearsarenotlinear.Future
studycanfurthertestsuchchangepatternsbyusingoriginallongitudinaldataandlatent
transitionanalysis.
Second,variablesforthetwoyoungergenerationswerelimited.Therewasno
informationabouttheincome,health,ormigrationofthemiddlegeneration.Thus,Iwasunable
totestwhethersupportfromthemiddlegenerationwaslimitedorunavailableduetopoorhealth
conditions,migration,orotherfactors.Inaddition,ifparticipantsdidnothavelivingchildren,
theywereexcludedfromthestudy,regardlessofwhethertheyhadlivinggrandchildren.Thelack
ofinformationaboutdeaths,migration,orhealthconditionsamongthemiddlegenerationlimited
testingofwhethergrandchildrensupplementedorsubstitutedmissingorinsufficientsupport
fromthemiddlegeneration.Moreover,variablesatthegrandchildlevelwerenottested.Inthis
dataset,therewasnospecificinformationaboutthecharacteristicsofgrandchildren,suchas
education,gender,income,andhealth.Thus,Iwasunabletotestwhetherthesocialresourcesor
capacityofgrandchildrenaffectedintergenerationalsupportpatterns.
Third,alladultchildrenandgrandchildrenweretestedasagroup.Informationregarding
whichadultchildorgrandchildprovidedsupportwasnotavailable.Thispreventedanalysisof
associationsbetweensupportprovisionandindividualbackgroundcharacteristics.Italso
preventedmefromdevelopingconclusionsaboutwhetherandhowtheparentsofeach
grandchildfacilitatedorhinderedinstrumentalsupportfromthatgrandchild.However,asmore
dataarecollected(includinginformationaboutsupportexchangebetweenolderadultsandeach
grandchildintheireldestchild’sfamily),futurestudiescouldtestpatternsofsupportfrom
individualgrandchildren.
62
Conclusions
Byusingchangescoresfromafour-wavelongitudinaldataset,thisstudycontributedto
existingknowledgebyexaminingchangepatternsofinstrumentalsupportprovidedbyadult
childrenandgrandchildrentoolderadultsinruralChina.Thisstudyfoundthreetypesofsupport
patternsovertime:decreasingsupport,increasingsupport,andlittlesupport.Generallyspeaking,
ruralolderChineseadultsmainlyreceivedinstrumentalsupportfromtheiradultchildren,and
onlyoccasionallyreceivedsupportfromtheirgrandchildren.Themiddlegenerationplayeda
mediatingroleinfacilitatingorhinderinginstrumentalsupportfromgrandchildrentoolder
adults.Individualsociodemographicvariablesofthetwooldergenerationsindicatedthatthe
personalneedsofolderadults,theirgeographicproximityandemotionalbondswithadult
children,andsocietalnormsofreciprocitywerethemainunderlyingfactorsthatdetermined
patternsofinstrumentalsupportfromthetwoyoungergenerationsovertime.
63
Chapter5
TheDynamicRelationshipbetweenInstrumentalSupportfromGrandchildrenand
DepressionamongRuralChineseGrandparents:ALongitudinalStudy
Introduction
Depressionhasbecomeaninternationalpublichealthproblemaffecting121million
peopleworldwide(BioMedCentral,2011).TheincidencerateofdepressioninChinahasbeen
reportedas12%overall(Brometetal.,2011).Lowersocioeconomicstatus,fewersocial
activities,andapoorerhealthcaresystemplaceolderruralChineseadults,whocomprise80%of
China’solderpopulation,athigherriskfordepressivesymptomscomparedwiththeirurban
counterparts(Chen,Hu,Qin,Xu,&Copeland,2004;Ma,Qin,Chen,&Hu,2006).The
prevalenceofdepressivesymptomsamongolderadultsishigherinruralregions(29.2%)than
urbanareas(20.5%;Zhang,Xu,Nie,Zhang,&Wu,2012).Depressioncanleadtosuicide,which
canhaveadevastatingeffectonfamiliesandcommunities(BioMedCentral,2011;Gottfries,
2001).Therefore,itisveryimportanttoexaminethispopulation’smentalhealthanddevelop
strategiestoaddressthedepressivesymptomsofruralolderadults.
Previousstudieshaveshownthatfamilysupportfromadultchildrenhaseitherharmful
(e.g.,Reinhardt,Boerner,&Horowitz,2006)orbeneficialeffects(e.g.,Cong&Silverstein,
2008a;Wolff&Agree,2004)ondepressionamongolderadults.However,fewstudieshave
examinedtheuniquefunctionofsupportfromgrandchildrenondepressionamongolderadults.
Amongthefewstudiesonthistopic(Lou,2010,2011),allofthemwerecross-sectionaldesigns.
Tomyknowledge,nostudieshaveusedalongitudinalresearchdesigntotracktheeffectsof
supportfromgrandchildrenongrandparents’depression.Additionally,therearefewstudiesthat
64
haveprovidedfirmconclusionsregardingthetemporalsequenceofsupportandpsychological
well-beingoutcomes.Basedontheliterature,itisunclearwhethertherearereciprocaleffects
betweensupportfromgrandchildrenanddepressionorwhethereithervariableisaleading
predictorifcausalitybetweenthesetwovariablesisreciprocal.
Exploringtheseissuesisbothcriticalandreasonable.Ononehand,instrumentalsupport
isveryimportantforolderadultswhoincreasinglyrelyonhouseholdassistanceorpersonalcare
duetoalackofmobilityandphysicaldeclineinoldage.Ontheotherhand,grandparentsinrural
Chinaareverylikelytointeractwithandreceiveinstrumentalsupportfromtheirgrandchildren
duetomigrationofthemiddlegeneration.TogetherwithtraditionalChineseculture,which
emphasizesfilialpietyandsupportfromyoungergenerations,thismaygeneratelower
depressionlevelsamongolderChineseadults.Moreover,itisunknownwhetherthispossible
beneficialfunctionofsupportfromgrandchildrendisappearsorisweakenedaftertakinginto
considerationsupportfromthemiddlegeneration.Thisstudyaimedtoaddressthisgapinthe
researchbyexamininghowinstrumentalsupportfromgrandchildrenhaschangedandinfluenced
depressionamongolderChineseadultslivinginruralareasduring9years.
InstrumentalSupportandDepression
Voilsandcolleagues(2007)examinedthebaselinestatusandtrajectoryofsocialsupport
over5yearsamongolderadultswithdepressionbyusinglongitudinalmethodsandstructural
equationmodels.Thisstudyfoundthatinstrumentalsupportprovidedtoolderadultswashighat
baselineandincreasedovertime.Instrumentalsupportinthisstudywasmeasuredasassistance
fromfamilymembersorfriendsduringillness.Intermsofthetrajectoryofdepressioninolder
age,previous studieshaveshownanincreasinggrowthcurveamongChinese olderadults(Hsiao,
Xu,Chi,&Silverstein,inpress;Huangetal.,2011),althoughsomeWesternstudiesyielded
65
differentfindingsshowingadecreasinggrowthcurveofdepressionovertime(Yang,2007)orno
significantchangeofdepressionovertime(Taylor&Lynch,2004)amongolderadults.
Studiesoftheeffectofinstrumentalsupportonthepsychologicalwell-beingofolder
adultsarerelativelycommonbuthaveproducedmixedresults.Intheindividualistsocietiesof
theWest,receivinginstrumentalsupportmayimplyalossofautonomyandmayendangerself-
esteemandotherformsofpsychologicalwell-being(Brown,Nesse,Vinokur,&Smith,2003;
Reinhardtetal.,2006;Taylor,Welch,Kim,&Sherman,2007).Forexample,onestudyfound
instrumentalsupportfromfamilymembersandfriendswasharmfultodepressionandadaptation
tovisionimpairmentamong570olderadultsintheUnitedStates(Reinhardtetal.,2006).
Accordingtothisstudy,instrumentalsupportmaybenegativeforsomeolderadultsbecauseit
mayemphasizetheirinabilitytoaccomplishdailytasks.However,anotherWesternstudyfound
positiveaspectsofinstrumentalsupportforpsychologicalwell-being.Usingasampleof420
womenwithdisabilitiesaged65andolderwhoreceivedinformalcare,WolffandAgree(2004)
foundthatinstrumentalsupportwasassociatedwithalowerlikelihoodofbeingdepressed.
InEasternsocieties,inwhichfilialpietyplaysacriticalroleintheintergenerational
contract(Croll,2066),oldergenerationsarewillingtoreceivesupport,assistance,andcarefrom
youngergenerationsasaformofreciprocationandgratitudeforprevioussacrifices.Therefore,it
maynotundermineanolderadult’ssenseofindependencetoreceiveassistance.Usingtwo
wavesofdataonolderruralChineseadults,CongandSilverstein(2008a)foundthatdepressive
symptomswereusuallyreducedbyreceivingassistancewithhouseholdchoresandpersonalcare
fromdaughters-in-law,yetincreasedattimeswhensuchsupportwasprovidedbysons.Using
surveydatafromurbanareasofChina,SongandChi(2001)foundthatreceivinginstrumental
supportfromfamilymembers,includingchildrenandgrandchildren,couldhavepositiveeffects
66
onpsychologicaldistress.Thesestudiesshedlightontheeffectsofinstrumentalsupportonthe
well-beingofolderadults.However,thesestudies,bothinWesternsocietiesandChina,
generallyhavefocusedonfamilyand/orpeersupport.Theydidnotdistinguishbetweenthe
sourceofsupport(childrenorgrandchildren)andthereforedidnotaddressthespecificwell-
beingfunctionsofinstrumentalsupportfromgrandchildren.Instrumentalsupportfromthe
currentgenerationofgrandchildrenanditseffectonthepsychologicalwell-beingofolder
grandparentshasnotbeenwellstudied.
However,twostudieshavefocusedspecificallyonsupportfromgrandchildrenandits
influenceonthepsychologicalwell-beingofolderadultsinHongKong.Usingacross-sectional
surveywith215olderparticipantsinHongKong,Lou(2010)exploredsocialsupportfrom
grandchildrenandpsychologicalwell-being(lifesatisfaction)andfoundthatemotionalsupport
andappraisalsupportfromgrandchildrenweresignificantcontributorstolifesatisfaction.Lou
(2011)alsoexploredtherelationshipbetweendepressivesymptomsandgrandparentreward
(relation-oriented,emotionalfulfillment,andlearningopportunityrewards)amongolderadults
inHongKong.Thisstudyfoundthatrelation-orientedgrandparentrewardwassignificantly
negativelyassociatedwithdepressivesymptomsamongolderrespondents.Thesetwostudies
haveopenedthedoorforadditionalstudiesonthistopicandprovidedinsightintograndchild
supportanditsinfluenceonwell-being.However,itremainsunknownwhetherinstrumental
supportfromgrandchildrenaffectswell-beingoutcomes.Inaddition,ChinaandHongKongvary
intermsofmodernizationandothercharacteristics,whichmightaffecttheperceptionand
practiceoffilialpietyinintergenerationalrelationships(Brian&Logan,2001).Studieshave
shownthattheyoungergenerationinHongKongidentifieslesswithfilialpietycomparedwith
thosefrommainlandChina(Lou,2010).Inaddition,thetwoHongKongstudiesemployed
67
cross-sectionaldesignswithsmallsamplesizes;thereisnoevidencetoyieldclearconclusions
regardingtherelationshipbetweensupportfromgrandchildrenandpsychologicalwell-being.
Therefore,itremainsunclearhowgrandchildsupportaffectsthedepressionlevelofgrandparents
inmainlandChina,especiallyinruralareas.Ifatemporalsequenceexistsbetweenreceiving
supportanddepression,itremainsunknownwhetheroneplaysaleadingroleintherelationship.
Thewell-establishedmodeloffamilysupportandmentalhealthprovidesatheoretical
basetodescribe,explain,andunderstandthepervasiveinfluenceoffamilyrelationsonmental
health.Duringthepast30years,researchershaveinvestigatedhowfamilysupportfromadult
childrenimprovesthementalhealthofolderadults,bothasadirecteffectandasabufferof
stress(Lincoln,2000;Lincoln,Chatters&Taylor,2003),However,mostoftheseprevious
studiesthatinvestigatedtheeffectsoffamilysupportonthehealthorwell-beingofolderadults
weremainlyfocusedonsupportfromadultchildren.Thepresentstudyappliedthisgeneral
modeltotestthefunctionsofsupportfromgrandparentsondepressionamongolderadults.
ThePresentStudy
Areviewoftheliteratureindicatedascarcityofresearchoninstrumentalsupportfrom
grandchildrenanditseffectonthepsychologicalwell-beingofolderadults,includingissues
suchasdepression.Toaddressthisresearchgap,thepresentstudyaddressedthefollowing
researchquestions:
(1)Whatisthegrowthcurveofinstrumentalsupportfromgrandchildrenovertime?
(2)Whatisthegrowthcurveofdepressionovertime?
(3)Doeschangeininstrumentalsupportfromgrandchildrenaffectchangeindepression
overtime?
68
(4)Isthereacross-laggedeffectbetweeninstrumentalsupportfromgrandchildrenand
depression?
(5)Istherealeadingpredictorbetweeninstrumentalsupportfromgrandchildrenand
depression?
Basedontheseresearchquestions,thefollowinghypothesesweremade:
(1)Instrumentalsupportfromgrandchildrenwouldincreaseovertimeduetothe
increasingneedsofolderruralChineseadultsastheyage.
(2)DepressionlevelsamongolderruralChineseadultswouldincreaseovertime.
(3)Anincreaseininstrumentalsupportfromgrandchildrenwouldbenegatively
associatedwithanincreaseindepressionovertimeamongruralChinesegrandparents.
(4)Therewouldbeacross-laggedeffectofinstrumentalsupportfromgrandchildrenand
depression,i.e.,higherlevelsofinstrumentalsupportfromgrandchildrenwouldpredictslower
changesindepressionovertime,whereashigherlevelsofdepressionwouldpredicthigherlevels
ofinstrumentalsupportfromgrandchildren.
(5)Instrumentalsupportfromgrandchildrenwouldbealeadingpredictorofdepression.
Methods
Data
FourwavesofdatafromtheAnhuistudywillbeusedforthisstudy.Thedataand
samplinginformationhasbeendescribedintheintroductoryportionofthisdissertation.After
excludingparticipantswhofailedthecognitivetestatallfourtimepoints,thefinallongitudinal
sampleconsistedof731olderadultswhoparticipatedinsurveysin2001,2003,2006,and2009.
Thequalityofexistingdatahasbeenshowntobegood,asevidencedbyhighparticipationand
follow-upratesandexcellentmeasurementproperties.
69
Measurement
Dependentvariables.DepressionlevelwasmeasuredusingnineitemsfromtheCenter
forEpidemiologicStudiesDepressionScale(Radloff,1977)thatweretailoredtoAsian
populations(Hermalin,2002):threeitemsassessedfeelingsofpositiveaffect(happiness,
enjoyment,pleasantness);twoitemsassessedfeelingsofnegativeaffect(lonely,upset);two
itemsassessedasenseofmarginalization(useless,nothingtodo);andtwoitemsassessed
somaticsymptoms(poorappetite,trouble sleeping).Respondentsnotedhowoftenthey
experiencedeachsymptomduringthepastweek(1=rarelyornoneofthetime,2=someofthe
time,3=mostofthetime).Thereliabilitycoefficientsforthenineitemswere.77,78,.78and.80
acrosswaves.Afterreversecodingofthepositiveaffectitems,thesumoftheninevariables
resultedinadepressionscorerangingfrom9to27,withhigherscoresindicatinggreater
depression.
Independentvariables.Theindependentvariableinthisstudywasreceptionof
instrumentalsupportfromgrandchildren,whichwasmeasuredbyaskingrespondentswhether
theyreceivedanyhouseholdhelporpersonalcarefromtheirgrandchildrenorgrandchildren-in-
law(e.g.,cleaninghouse,washingclothes,andwashingdishes)duringthepast12monthsdueto
poorhealth(1=yes,0=no).Ifrespondentsansweredaffirmatively,theywereaskedaboutthe
frequencyofhouseholdsupport(1=seldom,2=severaltimespermonth,3=atleastonceper
week,4=everyday).Resultsforthesetwoquestionswerecombinedintoonescorerepresenting
thefrequencyofinstrumentalsupportfromgrandchildren,rangingfrom0=neverto4=every
day.
Controlvariables.Severalsociodemographicvariablesthathavebeenshowntobe
potentialpredictorsofsupportfromgrandchildrenordepressioninlaterlife(Cheng&Chan,
70
2006;Chou&Chi,2005)werecontrolledforinthisstudy.Agewasmeasuredasacontinuous
variableatthetimeofthesurvey.Educationlevelwasoriginallymeasuredusingseven
categories:1=noeducation,2=elementaryschool,3=juniorhighschool,4=seniorhigh
school,5=vocationalcollege,6=communitycollege,and7= collegeormore.Duetothe
highlyskeweddistribution,thisvariablewasrecodedasadichotomousvariable(1=educated,0
=noeducation).Gender(1= female,0= male),maritalstatus(1=married,2= notmarried),
andhealthstatus(1=verygood,2=good,3=fair,4=notsogood)werealsocontrolledforin
thisstudy.Incomeisacrucialfactorindeterminingthewell-beingofagrandparent(e.g.,Musil
&Ahmad,2002)and,thus,wasincludedasacontrolvariableinthisstudy.Incomewas
representedasthelog(+1)ofthetotalincomethattherespondentreceivedfromworkorpension
duringthepreviousyear.Becauseinstrumentalsupportfromthemiddlegenerationhasbeen
foundtosignificantlyaffectthepsychologicalwell-beingofruralolderadults(Cong&
Silverstein,2008a),thisvariablewasalsoincludedasacontrolandmeasuredinasimilarfashion
toinstrumentalsupportfromgrandchildren,butfocusedonsupportfromsons,daughters,
daughters-in-law,andsons-in-law.Becausehealth,income,andinstrumentalsupportfromthe
middlegenerationchangedacrossfourwaves,themeansofthesevariableswereusedas
invariantvariablesinthisstudy.Allothercontrolvariablesweremeasuredatbaseline.All
controlvariableswerecenteredatgroundmeanfortheanalysis.
Inthisstudy,cross-sectionalcorrelationsbetweenthemainvariablesanddemographic
variables(age,gender,education,maritalstatus,health,andincome)wereanalyzedfirst.
Educationwasconsideredbutexcludedasacontrolvariableduetoitssignificantassociation
withincome.Educationalsohadaskeweddistribution;approximately76%ofrespondentshad
receivednoeducation.Theotherdemographicvariablesofage,gender,maritalstatus,income,
71
andinstrumentalsupportfromthemiddlegenerationwereallsignificantassociatedwithsupport
fromgrandchildrenordepression.Therefore,theywereincludedinthefinalanalysis.
DataAnalysis
StructuralequationmodelingwasconductedusingAmos18totesttheintra-individual
changeacrossmeasurements,takingintoaccountboththecovariancestructureandmean
structureofthevariables.Inthefirstpartoftheanalysis,weusedlatentgrowthcurve(LGC)
modelstoinvestigatethecurvesofboththesupportfromgrandchildanddepressioninolder
adultsseparatelyandtherelationsbetweenthecurves.Then,inthesecondpartoftheanalyses,
latentdifferencescoremodels(LDS)wereusedtoexaminethetemporalsequencebetween
supportfromgrandchildrenanddepression(McArdle,2001;McArdle&Hamagami,2001).
BivariateLGCanalysiswasusedtotestHypotheses1to3,whereasbivariateLDSanalysiswas
usedtoexploreHypotheses4and5.
TheadvantageofbivariateLDSmodelsoverbivariateLGCmodelsisthattheyallowfor
testingofcross-lagged(coupling)effectsbetweenconstructs,i.e.,theeffectofonevariableona
subsequentintra-individualchangeinanothervariable,controllingforconstantandproportional
changes(Orth,Berking,Walker,Meier,&Znoj,2008).Inthecurrentstudy,theyhelped
determinewhetherreceivingsupportfromgrandchildrenpredictedsubsequentchangesin
depressionofgrandparentsorwhethergrandparents’depressionpredictedsubsequentchangesin
supportfromgrandchildren.Inaddition,usingLDSmodelsratherthanatraditionalcross-legged
panelmodelallowsforconsiderationofboththecovarianceandmeanstructuresofthevariables,
andLDSmodelsaccountforintra-individualchangesinabsolutevalues(Berking,Neacsiu,
Comtois,&Linehan,2009;Orthetal.,2008).However,bivariateLDSmodelsdonotallowfor
72
clear-cutinterpretationofthecorrelationsbetweenthegrowthcurves,whichisthereasonwhy
LGCanalyseswereconductedfirst.
Figure5.1.LatentGrowth CurveModelwithoutControlVariables
Note.Sintercept=interceptofintrumentalsupportfromgrandchildren;Dintercept=interceptofdepressionofolder
adults;Sslope=slopeofinstrumentalsupportfromgrandchildren;Dslope=slopeofdepressionofolderadults.
Figure5.1showsthebivariategrowthcurveanalysisthatconnectstwounivariatemodels.
Thetwounivariatemodelsofinstrumentalsupportfromgrandchildrenanddepressionof
grandparentswereillustratedintheupperhalfandlowerhalfofFig.5.1,respectively.Each
73
univariatemodelwasbasedontherepeatedassessmentofindicatorsacrossfourmeasurement
occasions,withthemodelaccountingforthecovariancestructure,aswellasthemeanstructure.
Theindicatorswereexplainedbytwocorrelatedlatentvariables:theinterceptandtheslopeof
thegrowthcurve,aswellasbyuncorrelatederrors.Theloadingsoftheinterceptwerefixedto1,
andtheloadingsoftheslopewerefixedtovaluescorrespondingtothelengthofthetime
intervals,staringwith0forthefirstassessment.Inthisstudy,assessmentsweredistributed
regularlyovertime.Therefore,wefixedtheloadingsofthefourrepeatedassessmentsto0,1,2,
and3.
Figure5.2showsthebivariateLDSmodels.InLDSmodeling,indicatorsareexplained
bylatenttruescorevariables(denotedasLsupportandLdpreinFig.5.2)anderrors.Thetrue
scorevariables(exceptforTime1)areexplainedbythelatenttruescoresonthepreceding
assessmentandbylatentdifferencescores(denotedasDsupportandDdpreinFig.5.2).Thetrue
scoresofsupportfromgrandchildrenanddepressionatTime1wereentirelyexplainedbylatent
interceptvariables,giventhatnoautoregressorwasavailable.Inthepresentstudy,thepath
coefficientsfromlatentslopevariablestolatentdifferencescoreswerefixedto1(McArdle,
2001).LDSmodelingalsoaccountsforself-feedbackorproportionaleffect,representedbythe
effectsofthetruescoreatoneassessmentonthedifferencescoreatthenextassessment
(coefficientscand dinFig.5.2).Mostimportantly,LDSmodelingindicatescross-laggedeffects
betweenthetruescoreofonevariableandthelatentdifferencescoresoftheothervariable
(coefficientsaandbforHypotheses4and5,respectively,inFig.5.2).Couplingeffectsshow
whethertheintra-individualchangeinonevariableisexplainedbythetruescoreoftheother
variablemeasuredduringtheprecedingassessment,whilecontrollingforconstantchangeand
self-feedback.ThemeansandvariancesofslopesinLDSmodelsaredifferentfrommeansand
74
variancesofslopesinLGCmodels.Unstandardizedcoefficientsarereportedbecause
standardizedpathcoefficientsdonotapplytoLDSmodels.
Figure5.2.LatentDifferenceScoreAnalysisModelwithoutControlVariables
Note.Lsupport=latenttruescoreofinstrumentalsupportfromgrandchildren;Ldepre=latenttruescoreof
depression;Dsupport=latentdifferencescoreofinstrumentalsupportfromgrandchildren;Ddepre=latent
differencescoreofdepression;a=couplingeffectoflatenttruescoreofinstrumentalsupportfromgrandchildren
andlatentdifferencescoreofdepression;b=couplingeffectoflatenttruescoreofdepressionandlatentdifference
scoreofinstrumentalsupportfromgrandchildren;c=self-feedbackofinstrumentalsupportfromgrandchildren
reflectingeffectsoftruescoreofinstrumentalsupportatoneassessmentondifferencescoreatnextassessment; d=
self-feedbackofdepressionreflectingeffectsoftruescoreofdepressionatoneassessmentondifferencescoreat
nextassessment.
75
Missingdataisapotentialprobleminlongitudinalresearchduetodeathorotherreasons.
Toaddressmissingvalues,weusedthefullinformationmaximumlikelihoodprocedureincluded
inAmosandMplus;missingvalueimputationyieldsresultsthatarelessbiasedandmore
reliablethanthosegeneratedbyconventionalmethodssuchaslistwiseorpairwisedeletion
(Allison,2003;Schafer&Graham,2002).Modelfitwasassessedusingthreefitindexesthat
havebeenrecommendedasbeingthemostuseful(Hu&Bentler,1999;MacCallum&Austin,
2000):thenormedfitindex(NFI),thecomparativefitindex(CFI),andtherootmeansquare
errorofapproximation(RMSEA).HuandBentler(1999)suggestedthatgoodfitisindicatedby
valuesgreaterthanorequalto.95fortheNFIandCFIandlessthanorequalto.06forthe
RMSEA.Inadditiontotheseindexes,wereported χ
2
statisticsandconfidenceintervalsfor
RMSEAvalues.
Results
SampleCharacteristicsatBaseline
Table5.1featuresdescriptivestatisticsforsamplecharacteristics.Participantsinthis
studyhadanaverageageof68.03(SD=6.19),rangingfrom60to95.Genderwasevenly
dividedinthissample,with50.9%womenand49.1%men.Morethanhalfoftheparticipants
(63.6%)weremarried,andapproximately76%ofrespondentshadnoeducation.Aboutone-
quarter(25.7%)oftheolderadultsinthisstudyreportedtheirgeneralhealthtobeexcellentor
good,andnearlyhalfofallparticipants(46%)reportedtheywereinfairhealth.Respondentshad
anaverageincomeof839.50RMBor$123.40(SD=1526.70RMBor$224.50),rangingfrom0
RMBto14,000RMB($2,058.80).Theaverageinstrumentalsupportreceivedbyolderadults
fromadultchildrenwas2.16(SD=0.97),indicatingthatparticipantsreceivedhouseholdhelpor
personalcarefromtheiradultchildrenapproximately“atleastseveraltimesamonth”onaverage.
76
Table5.1
SampleCharacteristicsatBaseline(N=731)
Variable nM(SD) % Range
Age 731 68.03(6.19) 60–95
Female 372 50.9
Someeducation 176 24.1
Married 465 63.6
Self-ratedhealth 731 2.95(0.86) 1–5
Income 731 $123.40($224.50) $0–$2,058.80
Instrumentalsupportfromadultchildren 704 2.16(0.97) 0–4
Table5.2showsthemeansandstandarddeviationsofinstrumentalsupportfrom
grandchildrenanddepression.Theresultsindicatedthatdepressiongenerallyincreasedovertime,
althoughsomeindividualsreportedaconstantordecreasinglevelofdepression.Meanvaluesof
supportfromgrandchildrenalsoincreasedbetween2001and2009.However,thecritical
questionwasnotwhetherthereischangeinsamplemeansbutwhetherthereisinterindividual
variabilityinintraindividualchangeornot(Orthetal.,2008).Therefore,LGCanalyseswere
conductedbasedonstructuralequationmodeling.
Table5.2
MeansandStandardDeviationsofAge,ProvisionofSupportto
Grandchildren,andDepressionOverTime
Year Age InstrumentalSupport
fromGrandchildren
Depression
n M SD n M SD n M SD
2001 1,696 70.83 7.39 1,444 0.88 1.25 1,677 15.55 4.05
2003 1,325 72.41 6.95 1,106 1.20 1.15 1,322 15.36 4.13
2006 1,041 74.68 6.63 837 1.23 1.16 1,041 15.85 4.03
2009 752 76.11 6.23 548 1.36 1.04 752 15.63 3.84
BivariateLatentGrowthCurveAnalysis
77
ThefitofthebivariateLGCmodelwasgood( χ
2
=52.69,df=22,p<.000,RMSEA
=.044,NFI=.944,CFI=.966).Thecrucialstructuralparametersofthebivariatemodelsare
providedinTable5.3.Theresultsshowedthatthemeaninterceptsofbothinstrumentalsupport
fromgrandchildrenanddepressionweresignificantatp<.01.Thisindicatedthattherewere
significantintraindividualdifferencesinthebaselinemeasuresofsupportfromgrandchildren
anddepression.However,onlythemeanslopeofdepressionwassignificant,whichindicated
intraindividualdifferencesinsubsequentchangesindepressionbutnotinstrumentalsupport
fromgrandchildren.Inaddition,thevariancesoftheinterceptsandslopesofsupportfrom
grandchildrenanddepressionwerealsosignificant.Thus,theresultsindicatedthattherewere
significantinterindividualdifferencesatbaselineandovertimeregardingthesetwovariables.
Table5.3alsoreportstherelationshipbetweentheinterceptsandslopesofthesame
variables,showingthattheinterceptandslopewassignificantlynegativelycorrelatedonlyfor
supportfromgrandchildren( β=-0.94,p<.05).Thenegativecorrelationindicatedthat
individualswhostartedwithhighvaluesatWave1(2001)tendedtohavelowerslopes(i.e.,a
moregradualincreaseinsupportreceivedfromgrandchildren).Thesupportinterceptand
depressioninterceptweresignificantlynegativelycorrelated( β=-0.13,p<.000),indicatingthat
ahigherstartingvalueofinstrumentalsupportfromgrandchildrenwasassociatedwithaslower
increaseindepressionovertime.Conversely,higherlevelsofdepressionatWave1were
correlatedwithaslowerincreaseininstrumentalsupportfromgrandchildrenovertime.The
resultsalsoshowedthatchangesininstrumentalsupportfromgrandchildrenwerealsonegatively
associatedwithchangesindepressionlevelsovertime( β=-1.64,p<.01).Thisindicatedthatan
increaseininstrumentalsupportfromgrandchildrenwasassociatedwithaslowerincreasein
depressionlevelsovertime,andanincreaseindepressionwasassociatedwithaslowerincrease
78
ininstrumentalsupportfromgrandchildren.Thecorrelationsbetweentheinterceptofone
variableandtheslopeoftheothervariablewerealsosignificant,indicatingthatthestartingpoint
ofonevariablewascorrelatedwiththegrowthcurveoftheothervariable.Ahigherstarting
pointofinstrumentalsupportwasnegativelyassociatedwithincreaseddepressionover9years
( β=-0.25,p<.05),andhigherlevelsofdepressionatWave1werenegativelyassociatedwith
increasesininstrumentalsupportfromgrandchildren( β=-1.34,p<.01).
Table5.3
BivariateLatentGrowthCurveModel:EstimatesandCoefficients
withoutControlVariables
Parameter Coeff.
a
tp
Means
Supportintercept 1.31 6.06 .000
Depressionintercept 15.53 17.35 .000
Supportslope 1.02 1.87 .081
Depressionslope 3.10 4.52 .000
Covariances
Supportintercept ↔supportslope -0.94 -2.12 .034
Depressionintercept ↔depressionslope 0.07 0.82 .414
Supportintercept ↔depressionintercept -0.13 -4.76 .000
Supportslope ↔depressionslope -1.64 -1.80 .004
Depressionintercept ↔supportslope -1.34 -2.89 .001
Supportintercept ↔depressionslope -0.25 -2.30 .020
Variances
Supportintercept 1.18 13.25 .000
Depressionintercept 12.55 8.61 .000
Supportslope 0.15 4.41 .000
Depressionslope 0.39 2.51 .012
SupportT1 1.08 8.80 .000
SupportT2 1.15 13.73 .000
SupportT3 1.22 12.55 .000
SupportT4 2.37 11.35 .000
DepressionT1 12.85 10.44 .000
DepressionT2 18.01 15.53 .000
DepressionT3 18.59 15.72 .000
DepressionT4 16.26 10.61 .000
a
Unstandardizedcoefficient
BivariateLatentDifferenceScoreAnalysis
79
TheresultsofbivariateLGCanalysesdidnotprovideabasisfordrawingaconclusion
aboutwhetheroneofthevariablespredictedsubsequentchangesintheothervariablebecause
theparametersofcurveswerebasedontheidenticaltimeinterval.Therefore,LDSanalyseswere
usedtoinvestigatethetemporalsequenceofsupportfromgrandchildrenanddepression.
Table5.4
BivariateLatentDifferenceScoreAnalyses:EstimatesandCoefficients
withoutControlVariables
Parameter Coeff.
a
tp
Regressioncoefficients
Support →Dsupport(self-feedbackc) -0.34 -1.46 .160
Depression →Ddepression(self-feedbackd) -0.17 -2.86 .004
Support →Ddepression(couplinga) -0.04 -2.12 .012
Depression →Dsupport(couplingb) -0.15 -2.72 .008
Variances
Supportintercept 0.22 4.67 .000
Depressionintercept 4.86 7.67 .000
Supportslope 3.00 3.08 .007
Depressionslope 0.57 2.10 .036
Supporterror 0.53 11.23 .000
Depressionerror 8.39 13.95 .000
Dsupporterror 0.21 3.16 .002
Ddepressionerror 3.21 3.12 .002
Note.Dsupport=latentdifferencescoreofinstrumentalsupportfromgrandchildren;
Ddepression=latentdifferencescoreofdepressionamongolderadults
a
Unstandardizedcoefficient
ThefitoftheLDSmodelwasgood( χ
2
=98.50,df=22,p<.000,RMSEA=.045,NFI
=.91,CFI=.94).ThemainparametersarereportedinTable5.4.Theproportionalorself-
feedbackeffectsweresignificantfordepression( β=-0.17,p<.01)butnotforinstrumental
supportfromgrandchildren.Thus,inadditiontotheconstantchangeexplainedbythegrowth
curveportionoftheLDSmodeling,therewasevidenceforproportionalchangeinindicatorsof
depression.Thecross-laggedorcouplingeffectsofinstrumentalsupportfromgrandchildrenand
depressionwererelevantfortestingHypothesis4.Theresultsshowedthatthelevelof
instrumentalsupportfromgrandchildrenwasnegativelysignificantlyassociatedwithsubsequent
80
increasesindepression( β=-0.04,p<.05),andthelevelofdepressionwasalsonegatively
significantlyassociatedwithsubsequentincreasesinsupportfromgrandchildren( β=-0.15,p
<.01).Thesefindingsindicatedthatreportingmoreinstrumentalsupportfromgrandchildrenina
previoussurveyledtoaslowerincreaseinsubsequentdepression,andhigherlevelsof
depressionatprevioustimepointsledtoaslowerincreaseinsubsequentinstrumentalsupport
fromgrandchildren.Inaddition,theresultsshowedthatdepressionhadaleadingpredictiverole
amongthetwovariables,whichwasindicatedbyitshigher βortvalue.
Amongdemographiccontrolvariables,ageatbaselinehadsignificanteffectsonthe
supportfromgrandchildrenintercept( β=0.03, p<.001),depressionintercept( β=0.09, p<.05),
instrumentalsupportfromgrandchildrenslope( β=0.01,p<.001),anddepressionslope( β=-
0.01,p<.05).Thisindicatedthatolderagecorrespondedwithmoreinstrumentalsupportfrom
grandchildrenandhigherlevelsofdepressionatbaseline.Also,olderagewasassociatedwith
morechangeinsupportfromgrandchildrenandlesschangeindepressionacrossthefourwaves.
Femaleparticipantshadhigherlevelsofdepressionatbaseline( β=0.46,p<.05)andasteeper
increaseinratesofdepressionovertime( β=0.19p<.05).Marriedparticipantshadlowerlevels
ofdepressionatWave1( β=-0.78,p<.001).Incomewassignificantlyassociatedwiththe
intercept( β=-0.03,p<.001)andslope( β=-0.01,p<.05)ofdepression,butnotinstrumental
supportfromgrandchildren.Thisindicatedthatmoreincomewasassociatedwithlessdepression
atbaselineandovertime.Receivingmoreinstrumentalsupportfromthemiddlegenerationwas
positivelyassociatedwithbaselinelevelsofsupportfromgrandchildren( β=0.22,p<.001)and
depression( β=-0.94,p<.001).Inaddition,self-ratedhealthwaspositivelyassociatedwith
baselinesupportfromgrandchildren( β=0.11, p<.001)andbaselinedepression( β=1.63,p
<.000).
81
Discussion
Inthisstudy,Iinvestigatedtherelationshipbetweeninstrumentalsupportfrom
grandchildrenanddepressionusingalongitudinaldatasetwithfourmeasurementsover9years.
Thislongitudinalstudyrevealedthatalthoughinstrumentalsupportfromgrandchildrenincreased
overtime,thisincreasewasnotstatisticallysignificant.Thus,Hypothesis1wasnotconfirmed.
Thismaybeduetotheinfluenceofthemiddlegeneration,whichprovidedhigherlevelsof
instrumentalsupporttoolderadultsthangrandchildren,particularlywhentheolderadultsneed
suchsupport.AsthestudyshowedinChapter4,grandchildrencomplementedsupportfromthe
middlegenerationandfollowedsupportbehaviorofparentgeneration;therefore,changesin
instrumentalsupportfromgrandchildrenovertimemaynotbesignificant.Thislongitudinal
studyalsorevealedthatdepressiongenerallyincreasedovertime,athoughitsmeanlevel
decreasedsomewhatatthesecondmeasurement.Thus,Hypothesis2wasconfirmed.This
findingisconsistentwithpreviousstudiesondepressionamongolderChineseadults,whichalso
foundthatdepressionworsenswithage(Hsiaoetal.,inpress;Huangetal.,2011).
Theresultsofthelongitudinalmodelsofthisstudy,especiallythosederivedfromLGC
analysis,showedthatchangesininstrumentalsupportfromgrandchildrenwerenegatively
correlatedwithchangesindepressionatstatisticallysignificantlevels.Thisindicatedthatan
increaseininstrumentalsupportfromgrandchildrenwasassociatedwithamoregradualincrease
indepressionovertime.Thus,Hypothesis3wasconfirmed.Thisfindingconfirmedthegeneral
beneficialeffectofsocialsupportonmentalhealthconditions.Byincludinginstrumentalsupport
fromgrandchildren,thepresentstudyalsoextendedthebroadtheoryofsocialsupportand
previousresearchthatgenerallyfocusedonsupportfromadultchildren.
82
Theresultsfromthisstudyalsoyieldedasignificantfindingregardingthecoupling
effectsofsupportfromgrandchildrenanddepression.Inonerespect,supportfromgrandchildren
wasfoundtobesignificantlyandnegativelyassociatedwithsubsequentchangesindepression.
Thesefindingsindicatedthatmoresupportfromgrandchildrenatapreviousmeasurementpoint
ledtoaslowerincreaseinsubsequentdepression.Thisfindingonceagainconfirmedthat
receivingsupporthasbeneficialeffectsonthementalhealthconditionofolderadults.However,
incontrasttoexpectations,thepresentstudyfoundthatdepressionhadanegativesignificant
effectonsubsequentchangesinsupportfromgrandchildren,i.e.,higherpreviouslevelsof
depressionledtoaslowerincreaseinsubsequentinstrumentalsupportfromgrandchildren.Here,
depressionservedasathreattoreceivingsupportintermsofinfluencingthewillingnessor
abilityofolderadultstoreceivesupport,ratherthanasaneedforinstrumentalsupportfromthe
youngergeneration.
Inaddition,depressionplayedaleadingpredictiveroleinthebivariaterelationship
betweenthesetwovariables.Thatis,depressionwasmorelikelytoaffectinstrumentalsupport
fromgrandchildreninthissamplethantheopposite.Here,thementalhealthcondition,suchas
depression,isanimportantresourceforolderadultsandismorelikelytoaffectolderadults’
willingnessandcapacitiestointeractwithothers.ThisfindingechoedMorrow-Howell,
Hinterlong,Sherraden,andRozario(2001),whonotedthatwell-beingplaysasignificantrolein
determininganindividual’scapacityforengaginginactivities.Thisfindingisalsoconsistent
withthesocialresourcestheory(Lin,1982,1983),whichclassifiesresourcesintotwocategories:
personalandsocialresources.Psychologicalwell-beingormentalhealthconditions,suchas
depression,representimportantpersonalresourcesthatmayaffectcapacitytointeractwith
others,suchasreceivingsupportfromgrandchildren.
83
Limitations
Theresultsofthisstudyshowingupwardtrajectoriesofinstrumentalsupportfrom
grandchildren(althoughnotstatisticallysignificant)anddepressionshouldbeinterpretedwith
cautionbecausetheoverallchangesinthetwovariablesovertimewereminimal.Nevertheless,
theobservedchangesduring9yearsindicatedsignificantintraindividualdifferencesinthe
trajectoryofdepressionamongparticipantslivinginruralareas,whicharehometotwothirdsof
the1.34billionmembersoftheagingpopulationinChina.Althoughthisstudywasbasedon
solidtheoreticalmodelsandemployedLGCandLDSmodelingtoanalyzepaneldatadrawn
fromrandomsamples,thusincreasingconfidenceinfindings,somemethodologicalpitfallslimit
interpretationsandgeneralizability.
First,thedatawerenotcollectedusinganexperimentaldesignanddonotofferadequate
controlofunobservedfactors.Therefore,definiteconclusionsofcausalitycannotbedrawn.The
effectsofinstrumentalsupportfromgrandchildrenondepressionandviceversaovertimehave
beenrarelystudiedamongAsianpopulationsanddeservemoreattentioninthefuturestudies
withexperimentaldesigns.Nevertheless,longitudinalanalysesareusefulbecausetheycan
indicatewhetherthedataareinaccordancewithacausalmodel.Second,thesingle-item
measurementofinstrumentalsupportfromgrandchildrenwassomewhatcrude.Moreover,the
secondarydatasetdidnotincludeinformationonwhichgrandchildorgrandchildrenprovided
instrumentalsupport,towhichmiddle-generationfamilyeachgrandchildbelonged,orthe
emotionalclosenessbetweeneacholderadultandeachadultchild.Futurestudiesshouldexplore
howthosefactorsmoderateormediatetheassociationbetweeninstrumentalsupportfrom
grandchildrenanddepressionamongolderadults.Third,thestudysamplecamefromAnhui
Provinceineastern-centralChina,aregioninwhichmorethan80%ofthepopulationislocated
84
inruralareas.Therefore,theabilitytogeneralizeresultstotheentirepopulationofolderadults
inChinaissomewhatcompromised,especiallyinurbanChina,giventhesignificantregionaland
urban–ruralvariations.
ImplicationsforPolicyandPractice
By2010,China’sagingpopulationhadgrownto1.34billion,representing13.3%ofthe
totalpopulation(Bristow,2011).Morethan80%ofolderadultsinChinaliveinruralareasthat
lackuniversalpublicpensionsandlong-termcareprograms(Cong&Silverstein,2011).Hence,
thispopulationhasbeenfoundtobeathigherriskofdepression(Zhang,Xu,Nie,Zhang,&Wu,
2012).Findingsofthisstudyhighlighttheimportantroleofofficialsinwelfareinstitutionsand
nonprofitorganizationsinruralareasintermsoftrainingmorequalifiedmedicalpersonnelto
providementalhealthservicesorinterventionstoolderadultstoimprovetheirpsychological
well-being.RuralChinesecommunitiescanprovideprogramstoimprovethementalhealth
statusofolderadults,whichcouldincreasetheirinteractionswithgrandchildren,suchas
receivingsupportfromgrandchildren.Thissupportcouldleadtoaslowerincreaseorevena
decreaseindepression,creatingapositivecyclethatwouldbenefitsociety,thefamily,andthe
psychologicalwell-beingofolderadults.
Furthermore,supportfromadultchildreninlaterlifehasdrawnsignificantattentionfrom
researchersandpolicymakersduringthepasttwodecadesacrossallcultures.However,the
conceptandfunctionofsupportfromgrandchildreninChinesesocietyisrelativelynewand
deservesmoreattentionfromresearchers,practitioners,andpolicymakers.InruralChina,
grandparentsandgrandchildrenareimportantsourcesofsupport.Whenformalsocialresources
orprogramarelimitedandsupportfromadultchildrenisunavailableduetomigrationofthe
middlegeneration,grandchildrenbecomeimportantsocialassetsorcapitalforolderadultsin
85
ruralChina.TheChineseculturalvalueoffilialpietythatemphasizestheobligationofyounger
generationstotheirelders,aswellastherealitythatbetween60%and90%ofolderrural
Chineseadultscarefortheiryounggrandchildren(Liuetal.,2009),makesreceivingsupport
fromgrandchildrenreasonableforgrandparentsinolderage.Becausethisstudyhasindicated
thatsupportfromgrandchildrenleadstolessdepression,communitiesshouldencourageand
advocateforyoungergenerations,suchasgrandchildren,tofunctionassupportproviderstotheir
grandparentsforthebenefitofthefamilyandsocietyasawhole.
Conclusions
ResearchonsupportfromgrandchildrenanddepressionamongolderruralChineseadults
issorelyneeded.Drawingonsolidsocialsupporttheoreticalmodels,thisstudycontributedtothe
literatureonsupportanddepressioninthreeimportantways.First,LDSmodelingusedinthe
presentstudyprovidedforamorevaliddescriptionoflongitudinaldatathancross-laggedmodels
(Ferrer&McArdle,2003).Thisstudyrevealedtheunderlyingmechanismsbetweeninstrumental
supportfromgrandchildrenanddepressionbyshowingthetemporalsequenceoftheconstructs.
Second,findingshighlightedtheimportanceofpromotingpsychologicalwell-beingamongolder
adultsinAsiancountries,particularlyinruralareaswithalackofmedicalresourcesandmental
healthservices.Moreover,thefindingsofthisstudyhaveimplicationsforpolicyplannersand
practitionersinofficialandnonprofitinstitutionsintermsofassumingmoreproactiverolesin
creatingprogramsthatenhanceawarenessofmentalhealthissues,aswellasencouraging
grandchildrentoserveasanotherimportantsupportresourcetohelptheagingpopulation
maintainpositivementalhealthandwell-being.
86
Chapter6
ConclusionsandImplications
Socialsupportisespeciallyimportantforpeopleinlaterlifeduetophysicalandlife-
coursechanges.ThemajorityofolderadultsinChinareceivesocialsupportprimarilyfromtheir
adultchildren.However,solerelianceonadultchildrenforoldagesupporthassomeproblems
duetolowerstandardsoflivingandmigrationofmiddle-agedlaborersamongruralresidents.
TheolderadultsinruralChina,then,mustseeksupportfromothermemberswithintheirfamily
networks.Duetoincreasedlongevity,alargeproportionofolderadultshaveadultgrandchildren
whomayserveasimportantsourcesofsupport.However,fewstudieshaveexploredthe
dynamicsofsupportprovidedtoolderadultsbygrandchildreninruralChina:acriticalissue
giventhatdemographicshiftsindicatemostgrandparentsandgrandchildrenwillshare
overlappinglifecourses.
ThisdissertationfocusedonfamilysupportinruralChinaandinvestigatedthedynamics
ofintergenerationalsupportexchangebetweenolderparentsandtheirgrandchildren—
specifically,howadultgrandchildrenprovidedsupporttotheirgrandparentsandhowthis
supportaffectedthewell-beingofolderadultsinruralChina.Thisdissertationfeaturedthree
studiesthataddresseddifferentaspectsofintergenerationalsupportinruralChina:(1)
characteristicsofsupportexchangebetweengrandparentsandgrandchildrenandassociated
factors;(2)theunderlyingpatternsofinstrumentalsupportthatolderadultsreceivedfromadult
childrenandgrandchildren;and(3)howsupportfromgrandchildrenaffectedlevelsof
depressionamongolderadultsover9years.
87
Thesethreestudieswereconductedusingdatafromafour-wavesurveyconductedin
AnhuiProvince,China.OnenotableadvantageoftheAnhuisurveyisitsmultipledimensionsof
supportandrelationshipcharacteristics.Tothebestofmyknowledge,theAnhuisurveyisthe
largestdatasetavailablewithdetailedinformationregardingmonetary,instrumental,and
emotionalsupportexchangesamongruralChineseelders,theiradultchildren,andsomeoftheir
grandchildren.ThelongitudinaldesignoftheAnhuisurveyenabledmetodrawmorerigorous
conclusionsregardingcurrentsupportexchange,aswellaschangingpatternsofintergenerational
supportbetweengrandparentsandgrandchildrenovertime.
SummaryofMajorResearchFindings
Thefirststudyfocusedonthecurrentsupportexchangebetweengrandparentsand
grandchildrenusingcross-sectionaldatafromthefourthwaveoftheAnhuistudy.Specifically,
boththerolesofprovidingsupporttoandreceivingsupportfromgrandchildrenintermsof
instrumental,financial,andemotionalsupportwereinvestigated.Theresultsindicatedthatmore
grandparentsreceivedsupportfromthanprovidedsupporttotheirgrandchildren.Thisfindingis
inoppositiontoapreviousstudy(Hoff,2007)thatfoundeldersoftensacrificebygivingmore
supportthantheyreceive.However,thisfindingisconsistentwithtraditionalfamilialnormsin
ruralChina,deemingthatyoungergenerationsshouldfeelobligatedtocare foroldergenerations.
ThisisbasedonboththetraditionalConfucianethicsoffilialpietyandtheneedsofolderadults,
whooftenhavemorephysicallimitationsinoldage.Thisstudyalsofoundthathigher
percentagesofeldersexchangedsupportwiththegrandchildrenoftheirfirst-bornchildifthat
childwasmale.ThisfindingisconsistentwiththepatrilinealtraditioninruralChinathatsupport
exchangesareexpectedwithsonsandtheirfamilymembers(Cong&Silverstein,2012).
Analysisofsixdimensionsofsupportexchangeindicatedthathavingprovidedcaretotheir
88
grandchildrenduringthepastwaspositivelyassociatedwithcurrentsupportexchangesbetween
grandparentsandgrandchildren.Thisconfirmedthereciprocalfilialpietyofyoungergenerations;
grandchildrenaremotivatedtoprovidesupporttotheirgrandparentsbecauseofgratitudefor
theirprevioussacrifices,aswellastheirlong-terminvestmentinthefamily.Moreover,thisalso
reflectedtheauthoritarianfilialpietyornormsofresponsibilitythatyoungergenerationsfeel
towardolderadults,whicharelinkedtotheauthoritarianstatusorhierarchyofolderadultsinthe
familysystem(Yeh&Bedford,2003).Emotionalbondswiththemiddleparentgenerationwere
alsosignificantlyassociatedwithprovidingemotionalsupporttoandreceivinginstrumental,
financial,andemotionalsupportfromgrandchildren.Emotionalbondswithsonswereespecially
significantforbothreceivingandprovidingsupport.Thisalsoconfirmedthepatrilinealtradition
inruralChinaandthemediatingroleofthemiddlegeneration.
Thesecondstudyofthisdissertationuseddataregardinginstrumentalsupportfromadult
childrenandgrandchildrentoexploretheunderlyingpatternsofinstrumentalsupportfromthe
twoyoungergenerationsovertime.UsingfourwavesoftheAnhuistudy,Irevealedthree
underlyingclassesofinstrumentalsupportprovidedbyadultchildrenandgrandchildrentothe
olderadultsover9years:decreasedsupport,increasedsupport,andlittlesupport.Findings
indicatedthatapproximately67.1%ofolderadultsbelongedtothethirdgroup,meaningthey
receivedlittlesupportfromtheirgrandchildrenandmixedpatternsofsupportfromadultchildren
(withthelargestgroupofferinglittlesupport).Thismightbeexplainedbythesharedfamily
unityofadultchildrenandgrandchildren,theinfluenceandauthorityoftheparentgeneration,or
competingdemandsamonggrandchildrensuchasemergingcareersorburgeoningfamilies.
Instrumentalsupportfromgrandchildrenmirroredsupportfromadultchildren.Thisfinding
89
highlightedthemediatingroleofthemiddlegeneration,whichfacilitatedorhinderedsupport
fromgrandchildren.
Thethirdstudyofthisdissertationexploredintergenerationalsupportbetween
grandparentsandgrandchildrenbyexaminingtheeffectsofsuchsupportonthepsychological
well-beingofolderadults.Thefindingsofthisstudyshowedthatbothinstrumentalsupportfrom
grandchildrenanddepressionincreasedovertime,althoughtheformerwasnotstatistically
significant.Findingsalsoindicatedthatincreasesinsupportfromgrandchildrenwerenegatively
associatedwithincreasesindepression,suggestingthatinstrumentalsupportfromgrandchildren
hadbeneficialeffectsondepression.Cross-laggedeffectsbetweeninstrumentalsupportfrom
grandchildrenanddepressionwerealsofound.However,grandparents’depressionlevelsplayed
aleadingroleinpredictingsupportfromgrandchildren,ratherthanviceversa,indicatingthe
significantinfluenceofmentalhealthconditionsinlaterlife.
Overall,thefindingsofthisdissertationrevealedthatgrandchildrenareimportantsources
ofsocialsupportandcapitalforolderadultsinruralChina.Moreparticipantsreceivedsupport
fromgrandchildrenthanprovidedsupporttograndchildren,asindicatedbythefirststudy,and
instrumentalsupportfromgrandchildrenwasassociatedwithslowedprogressionofdepression,
asfoundinthethirdstudy.However,thesecondstudyshowedthatgrandchildrengenerally
complementedthesupportprovidedbythemiddlegenerationratherthansubstitutingor
supplementingabsentorinsufficientsupport.Supportfromgrandchildrengenerallyfollowedthe
behavioroftheparentgeneration.Thisdissertationcontributedtofillingintheresearchgap
regardingthecaregivingroleandfunctionsofthegrandchildreninrelationtoolderadultsin
ruralChina.
90
ContributionsofthisDissertation
Asaresultofsociodemographicshifts,relationshipsandsupportexchangewithinthree-
generationfamilynetworks,aswellasbetweengrandparentsandgrandchildrenhavebecome
relativelycommon(Harper,2005;Hoff,2007;Hoff&Tesch-Römer,2007).Unfortunately,little
attentionhasbeengiventothisissue;fewstudieshaveaddressedthesupportdynamicsbetween
grandparentsandgrandchildren.Thisdissertationisoneofonlyahandfulofstudiesthat
investigatedthesupportfromgrandchildrentotheirgrandparentstogainabetterunderstanding
offamilydynamicsinlaterlife.Itadvancedknowledgeconcerningthecomplexityinvolvedin
familyrelationsbyexaminingthecurrentgrandparent-grandchildmultidimensionalsupport
exchangeatonetimepoint,longitudinalpatternsofinstrumentalsupportfromgrandchildrenand
adultchildren,andthedynamicrelationshipbetweensupportfromgrandchildrenandthe
psychologicalwell-beingofolderadults.
Theoreticallyspeaking,thisdissertationconfirmedthesocialcapitaltheoryinthefamily
context,expandedtraditionalmodelsoffamilysupportandwell-being,andaddedtothe
caregivingliterature.Thesocialcapitaltheory,or,morespecificallyfamilycapitaltheory,refers
tosocialresourcesembeddedinanindividual’srelationshipswithotherfamilymembers.It
recognizessocialresourcesfromintergenerationalinteractions(Coleman,1988;Furstenberg&
Kaplan,2004).Theseresourcescanbeusedtocopewithfamilystressorsandpromotethewell-
beingoffamilymembers(Coleman,1990).Fromtheperspectiveofolderadults,immediate
relationshipswithadultchildrenandgrandchildrenarebasiccomponentsofintergenerational
interactions(Monserud,2008a).However,previousstudiesgenerallyfocusedonadultchildren
asthemajorsourceofsocialsupportwithinthefamilyforolderadults(Songetal.,2007).The
roleofsupportivefamilyresourcesbeyondparent-childdyadshasbeenlargelyignoredinthe
91
literature.Asmentionedearlier,longerlifeexpectanciesandbetterhealthmakeolderadults
morelikelytoassumegrandparentingrolesforlongerthananypointinhistory(Marshall&
Bengtson,2011;Xu&Chi,2011),andinteractionswithingrandparent-grandchilddyads
representasignificantsocialresourceforolderruralChineseadults(Lin&Yi,2011;Xu,2001;
Xu&Chi,2011).Thisstudyconfirmedtheexistenceofgrandparent-grandchildfamilycapitalin
ruralChina.Grandchildrenprovidedinstrumental,financial,andemotionalsupporttotheir
grandparents,whichhadbeneficialeffectsondepressionamongolderadults.
Previousstudieshaveverifiedtheimportanceoffamilysupportforthehealthandwell-
beingofolderadultsduringthepastthreedecades.Ingeneral,familysupportcanhavedirect
beneficialandindirectmoderating(buffering)ormediatingeffectsonhealthandwell-being
(Lincoln,2000;Lincolnetal.,2003).Yet,mostofthesepreviousstudiesinvestigatingtheeffects
offamilysupportonthehealthorwell-beingofolderadultsfocusedmainlyonsupportfrom
adultchildren.Thisstudynotonlyconfirmedthebeneficialeffectsofinstrumentalsupporton
thepsychologicalwell-beingofolderadults,butalsoextendedthetheoryofsocialsupportand
healthbyaddingsupportfromanotherimportantfamilygroup—grandchildren.
Thisdissertationalsoaddedtotheliteratureoncaregiving.Adultgrandchildrenhave
beenallbutoverlookedincaregivingresearch,althoughdemographicandcaregivingtrends
suggestthatadultgrandchildrenarefillingsociety’sneedforadditionalinformalcaregivers
(Smith,2011).TheFamilyCaregiverAlliance(2003)estimatedthatbetween59%and75%of
caregiversfortheelderlyintheUnitedStatesarewomen.Inaddition,researchersfoundthat
22%ofcaregiversareundertheageof35years(Dellman-Jenkins,Blankemeyer,&Pinkard,
2000).Noresearchtodatehascapturedhowmanyofthesecaregiversareadultgrandchildren.
Thisgenerationisonlybeginningtoreceiveconsiderationincaregivingparadigms,suchasthe
92
experiencesofadultgrandchildrenwhenagrandparentisill(Boon,Shaw,&MacKinnon,2008;
Fruhauf&Orel,2008;Ruiz&Silverstein,2007).Intergenerationalstudiestypicallyfocusonthe
adultchildgeneration,minimizingorignoringtherolethatgrandchildrenoccupyinthese
families(e.g.,Bengtson,Schaie,&Burton,1995;Thiele&Whelan,2008).Thisdissertation
showedthatgrandchildrenprovidedpersonalcareorhouseholdchoreassistancetotheir
grandparentsovertimewhentheirgrandparentsdidnotfeelwellorwereinneed.
ImplicationsforFutureResearch
Thelackofinclusionofgrandchildreninstudiesoffamilysupportandcaregivingandthe
overalllackofresearchonsupportexchangebetweengrandchildrenandgrandparentsrepresents
alimitationofthisstudy,giventheinabilitytocompareandcontrastfindings.Basedonthe
findingsofthisdissertation,futurestudiesshouldfocusonthefollowingareasregardingsupport
dynamicsbetweengrandparentsandgrandchildren.
First,whyorunderwhatcircumstancesdoadultgrandchildrenprovidesupportfortheir
grandparents?Grandchildrenplayanintegralsupportroleforgrandparentsintoday’ssociety.
However,theymaybeactivelysupportingboththeirparentsandgrandparents.Asaresult,our
understandingofwhatsupportgrandchildrenprovide,whoplaystheroleofsupportprovider,
andunderwhatcircumstancestheyprovidesupportdeservesadditionalattentionandexploration.
Thefirststudyofthisdissertationrevealedthatsupportfromgrandchildrenwasbasedonthe
needsofelderlygrandparents,aswellasthecapacityofgrandchildrentoprovidesupport.Future
researchshouldexaminethephysical,financial,andemotionalneedsofolderadultstodetermine
howeachfactorinfluencessupportfromgrandchildren.Thesecondstudyofthisdissertation
revealedsomecharacteristicsofgrandparentsthatwereassociatedwithreceivinginstrumental
supportfromthetwoyoungergenerations.However,fewfactorsorcharacteristicsofadult
93
childrenandgrandchildrenwereexamined.Death,sickness,migration,orrefusaltoprovide
supportamongthemiddlegenerationmaypromptgrandchildrentoprovidesupporttotheir
grandparents.Inaddition,furtherexaminationofgenderdynamicswilllikelydeepenour
understandingofthepatrilinealinfluenceonthesupportbehaviorsofgrandchildren.
Second,howdoemotionalbondsandrelationshipsinteractwithsupportexchangeinthe
familycontext?Thisdissertationfoundthatemotionalbondsbetweenolderadultsand
grandchildrenplayimportantrolesingrandparent-grandchildsupportexchangeinruralChina.
StudiesinWesternsocietiesalsorevealedthatclosenessbetweenolderadultsandtheir
grandchildren,aswellasbetweenadultchildrenandgrandchildren,hadpositiveinfluenceson
thesupportprovisionbygrandchildren(Monserud,2008a,2008b,2010).Futurestudiesshould
exploreintergenerationalrelationshipsandsupportexchangeacrossthreegenerationsinmore
detail.Forexample,futureresearchcanexamineintergenerationalfamilyrelationshipsandhow
theyareaffectedwhenthethirdgeneration(grandchildren)providessupporttoagrandparentin
need.Inaddition,interventionresearchinvolvingcaregivingdyadshasindicatedthatcaregivers,
inadditiontocarerecipients,shouldbeincludedinfuturestudies(Lingler,Sherwood,Crighton,
Song,&Happ,2008;Smith,2011).Furtherresearchshouldalsoexaminethecomplexityof
intergenerationalrelationshipsandhowsupportfromgrandchildrenchangesthesebondsfrom
theperspectivesofallfamilymembers.Thesestudieswouldprovideinsightintowhether
grandchildrenhelpcreatepeaceorconflictwithinfamilies.
Third,whatfamilydynamicsacrossthreegenerationsinfluencesupportprovisionto
olderadults?Theresultsofthisdissertationrevealedthatthemiddlegenerationplayedan
importantroleintheprovisionofsupportbygrandchildrentoolderadults.Futurestudiesshould
expandonthisfindingbyexploringintergenerationaldynamicsregardingsupportprovisionto
94
olderadultsbyincludingtheinsightsofgrandchildren,theirsiblings,theirparents,andother
membersofthemiddlegeneration,ifnecessary.Becausethisdissertationexaminedthemiddle
generationandthegrandchildgenerationasgroupsratherthanindividuals,itisunknown
whethergrandchildrenfollowedthebehavioroftheirownparentsspecificallyintermsof
providingsupporttoolderadults.Futurestudiesshouldexaminewhetherstressisexhibited
amongsiblingsinthemiddlegenerationastheresultofagrandchildprovidingsupportorcareto
agrandparent.Theseintergenerationaldynamicsamonggrandchildren,aunts,uncles,and
grandparentsneedfurtherexaminationtodeterminewhyagrandchildmayassumeasupport-
providerrole.
Finally,howdootherformsofsupportfromgrandchildrenaffectthewell-beingofolder
adults?Thisdissertationonlyexploredtheeffectsofinstrumentalsupportonthepsychological
well-beinganddepressionofolderadults.Ononehand,futurestudiesshouldinvestigatethe
effectsofotherformsofsupportthatgrandparentsreceivefromgrandchildren,suchasemotional
orfinancialsupport,onphysical,functional,andpsychologicalwell-beingmeasures,suchas
self-ratedhealth,activitiesofdailyliving,instrumentalactivitiesofdailyliving,andlife
satisfaction.Thisisfeasiblegiventhestrongmeasurementpropertiesofthosevariablesinthe
Anhuidataset.Ontheotherhand,theeffectsofprovidingsupporttograndchildrenonthehealth
andwell-beingofolderadultsalsoneedstobefurtherexamined,becausesupportexchangesare
bidirectional.Futurestudiescouldfurthertesttheclassicalmodeloffamilysupportandhealth,in
additiontoexaminingwhetherthephysicalormentalhealthconditionsofolderadultsplaya
leadingpredictiveroleintermsofreceivingfromandprovidingsupporttograndchildren.
95
ImplicationsforPoliciesandSocialWorkPracticeinRuralChina
FamilialsupportinChinawillencounterevengreaterchallengesinthefuturebecauseof
trendsofrapidaging,decliningfamilysizes,andongoingsocialupheaval.Studiesaboutthe
dynamicsoffamilystructureandintergenerationalsupportandtheirimplicationsforindividual
well-beingareespeciallyimportanttomaximizetheeffectivenessofinformalandformalsupport
forolderadults.Thefindingsofthisdissertationthatrevealedtheimportantsupportroleplayed
bygrandchildrenalsohavepracticeandpolicyimplicationsforintergenerationalfamilysupport
andcaregiversofolderruralChineseadults.
Becausesocialservicesandcommunitycareforolderadultsarelargelyunderdeveloped
inruralcommunities(Yipetal.,2007),thefamilynetworkrepresentsthemainsourceofsupport
forolderruralChineseadults(Leung,1997;Xu,2001).ComparedwithurbanChina,the
shortageofpublicservicesfortheelderlyinruralChinahasresultedinincreasedpressuretofind
solutionswithinfamilies.Althoughthemajorcaregiversorsupportersofolderadultsremain
adultchildren,grandchildrenrepresentanotherimportantfamilyresourceforolderadultsin
termsofsupportinlaterlife.Oneimportantlessonlearnedfromthisstudyisthatolderadultscan
seeksupportfromotherfamilymembersinadditiontotheiradultchildren.Thefindingsofthis
dissertationalsosuggestedthatthefamilycanuniteandworktogethertoeasesomeofthe
caregivingorsupporthardshipsfacedbytheprimarycaregiver,whoismostlikelyamemberof
themiddlegeneration.Practitionersinofficialwelfareinstitutionsandnonprofitorganizationsin
ruralareas,suchassocialworkers,needtounderstandthefullrangeofresourcesfromwhich
familysupportcanbedrawnandhowtheycanaffectthehealthandwell-beingofolderadults
differently.Todevelopmoreviablefamilialsourcesofeldercare,itisimportanttohavea
comprehensiveunderstandingofallpossiblecaregiversinafamilyunitwhenassessingthe
96
informalresourcesforagingresidents.Forexample,practitionerscouldencourageandadvocate
foryoungergenerations,suchasgrandchildren,toworktogetherwithparentgenerationsto
functionassupportprovidersfortheirgrandparentsforthebenefitofthefamilyandsocietyasa
whole.
However,solerelianceonfamilymembersforsupportinoldagemayplacetoomuch
pressureonyoungergenerations.Attheformalsupportlevel,practitionersandsocialwork
professionalshaveconsiderablepotentialtoeffectivelyaddressproblemswithincaredyadsin
multiplesettings.Becauseolderadultsoftenhavemanydifferentneeds,multidisciplinaryand
multifacetedserviceswouldbeidealtomeetthoseneeds.Healthandsocialservicesforolder
adultslivingwithlong-termchronicillnessesareincreasinglybeingprovidedinthehome.A
moresophisticatedunderstandingofcaredyadswouldleadtomoretargetedassessmentsand
interventionsthatareapplicabletoin-homesettings.Itisalsonecessarytofocusonolderadults
withfewerpersonalresources,suchasthosewithpoorphysicalfunctioningandfewerfamily
members,tobettermeettheneedsofmorevulnerableolderpopulations.
Anotherimportantlessonlearnedfromthisstudyisthatmacro-levelsocioculturaland
policyfactorsshapeparentalexpectations,familyrelations,andindividualwell-being.The
rapidlychangingsocialcontextinChinaisasourceofbothopportunityandobstruction.Itis
importanttoassessthestrengthsandchallengesoffamiliesinthelargersocioculturalandpolicy
context,suchasthepreferenceformaleoffspring,theone-childpolicy,thehousehold
registrationsystem,andgreatergeographicmobility.Theoverwhelmingimportanceofhaving
sonsinvolvedineldercaresuggeststhatpolicymakersshouldeitherprovideincentivesfor
intergenerationalcaregivingamongdaughtersandthechildrenofdaughtersorallocatenew
97
resourcesfortheprovisionofinstitutionalandagency-basedeldercareamongolderparentswith
fewornosons.
Associetiescontinuetoevolve,changesinfamilystructureandtraditionalvaluesmay
alterboththewillingnessandabilityofyoungergenerationstoprovidesupporttoolder
generations.TheeffectofthesechangeshasbeenparticularlystronginfamiliesintheAsia-
Pacificregion(Yue&Ng,1999;Zhan&Montgomery,2003).Toaddressthesechallenges,
socialworkersneedtobearmedwithknowledgeaboutthetraditionalvalueoffilialpietyandits
implicationsinfamilylife.Educationprogramsneedtobedevelopedtoenhancetheyounger
generations’understandingoftheneedsofolderadults,enablingthemtomoreeffectively
performtheirfilialroles.Intergenerationalprogramsshouldfocusoncommunicativestrategies
thatpromoteintergenerationalcontactandfacilitatepositiveinteractionbetweenyoungerand
oldergenerations.Allgenerationscanbenefitfromlearningtogetherandunderstandingthe
differentexpectationsandneedsofyoungerandolderfamilymembers.Itiscriticalthatplanners
ofintergenerationalprogramsareawareofthecontextualfactorsinvolvedindesigningtheir
programs.Attentionshouldbedirectedtowardidentifyingandtargetingspecificcontextsin
whichrelationsbetweengenerationsarelikelytobemostproblematic.Practitionersalsoneedto
developprogramsthatarecongruentwithlocalculturalnormsofsupportinlaterlife.
Finally,althougheverycountry’sapproachtosocialpolicyisuniqueandundoubtedly
influencedbydifferenthistoricalandculturalfactors,countriesintherelativelyearlystagesof
populationaging,suchasChina,canlearnfromdevelopedcountriesintermsofthearrangement
offormalcareasasupplementforfamilialsupport.Coordinatedresearchactivitiesbetween
developinganddevelopedcountriescancompoundthecontributionsofinvestmentsmadeby
individualcountriesinbothresearchandpolicyreform.
98
References
Allison,P.D.(2003).Missingdatatechniquesforstructuralequationmodeling.Journalof
AbnormalPsychology,112,545–557.
Agree,E.M.,Biddlecom,A.E.,Valente,T.W.,&Chang,M.(2001).Socialnetworkmeasures
ofparent-childexchange:ApplicationsinTaiwan&thePhilippines.Connections,24(2),
59–75.
Antonucci,T.C.(1990).Socialsupportandsocialrelationships.InR.H.Binstock&L.K.
George(Eds.),Handbookofagingandthesocialsciences(3rded.,pp.205–227).San
Diego,CA:AcademicPress.
Antonucci,T.C.,Jackson,J.S.,&Biggs,S.(2007).Intergenerationalrelations:Theory,research,
andpolicy.JournalofSocialIssues,63(4),679.
Attar-Schwartz, S., Tan, J.-P., & Buchanan, A. (2009). Adolescents’
perspectives on relationships with grandparents: The contribution of
adolescent, grandparent, and parent-grandparent relationship variables.
Children and Youth Services Review, 31, 1057–1066.
Ashton, V. (1996). A study of mutual support between Black and White
grandmothers and their adult grandchildren. Journal of Gerontological
Social Work, 26(1–2), 87–100.
99
Attias-Donfut,C.(2003).Familytransfersandculturaltransmissionsbetweenthreegenerations
inFrance.InV.L.Bengtson&A.Lowenstein(Eds.),Globalagingandchallengesto
families(pp.214–252).Hawthorne,NY:AldinedeGruyter.
Bahr,H.(1976).Thekinshiprole.InF.I.Nye(Ed.),Rolestructureandanalysisofthefamily
(pp.61–79).BeverlyHills,CA:Sage.
Barnett,M.,Scaramella,L.,Neppl,T.,Ontai,L.,&Conger,R.(2010).Intergenerational
relationshipquality,genderandgrandparentinvolvement.FamilyRelations,59,28–44.
Baker,L.A.,&Silverstein,M.(2012).Thewell-beingofgrandparentscaringforgrandchildren
inruralChinaandtheUnitedStates.InS.Arber&V.Timonen(Eds.),Contemporary
grandparenting:Changingfamilyrelationshipsinaglobalcontext.Bristol,England:Policy
Press.
Becker,G.S.(1991).Atreatiseonthefamily.Cambridge,MA:HarvardUniversityPress.
Becker,G.S.,&Tomes,N.(1979).Anequilibriumtheoryofthedistributionofincomeand
intergenerationalmobility.JournalofPoliticalEconomy,87,1153–1189.
Bengtson,V.L.(2001).Beyondthenuclearfamily:Theincreasingimportanceof
multigenerationalbonds.JournalofMarriageandFamily,63(1),1–16.
Bengtson,V.L.,Olander,E.B.,&Haddad,A.A.(1976).Thegenerationgapandagingfamily
members:Towardaconceptualmodel.InJ.F.Gubrium(Ed.),Time,roles,andselfinold
age(pp.237–263).NewYork:HumanSciencesPress.
Bengtson,V.L.,Schaie,K.W.,andBurton,L.M.(Eds.).(1995).Adultintergenerational
relations:Effectsofsocietalchange.NewYork:Springer.
100
Berking,M.,Neacsiu,A.,Comtois,K.A.,&Linehan,M.M.(2009).Theimpactofexperiential
avoidanceonthereductionofdepressionintreatmentforborderlinepersonalitydisorder.
BehaviorResearchandTherapy,47,663–670.
BioMedCentral.(2011,July26).Globaldepressionstatistics.ScienceDaily.Retrievedfrom
http://www.sciencedaily.com/releases/2011/07/110725202240.htm
Bromet,E.,Andrade,L.H.,Hwang,I.,Sampson,N.A.,Alonso,J.,deGirolamo,G.,…Kessler,
R.C.(2011).Cross-nationalepidemiologyofDSM-IVmajordepressiveepisode.BMC
Medicine,9,90.doi:10.1186/1741-7015-9-90
Bowen,M.(1966).Theuseoffamilytheoryinclinicalpractice. ComprehensivePsychiatry,7,
345–374.
Boon,S.,Shaw,M.,&MacKinnon,S.(2008).Grandparenthealthandyoungadults'judgments
oftheirgrandparent-grandchildrelationships.JournalofIntergenerationalRelationships,
6(2),155–173.
Brian,J.S.,&Logan,J.(2001).Intergenerationalrelations:AcomparisonbetweenChineseand
Americanfamilies.JournalofSociology,2,85–95(inChinese).
Brown,L.H.(2003).Intergenerationalinfluencesofperceptionsofcurrentrelationshipswith
grandparents.JournalofIntergenerationalRelationships,1(1),95–112.
Brown,S.L.,Nesse,R.M.,Vinokur,A.D.,&Smith,D.M.(2003).Providingsocialsupport
maybemorebeneficialthanreceivingit:Resultsfromaprospectivestudyofmortality.
PsychologicalScience,14,320–327.
Chen,R.,Hu,Z.,Qin,X.,Xu,X.,&Copeland,J.R.M.(2004).Acommunity-basedstudyof
depressioninolderpeopleinHefei,China:TheGMS-AGECATprevalence,case
101
validationandsocio-economiccorrelates.InternationalJournalofGeriatricPsychiatry,
19(5),407–413.doi:10.1002/gps.1103
Chen,S.(2009).AgingwithChinesecharacteristics:Apublicpolicyperspective.Ageing
International,34,172–188.
Chen,T.Z.(2009).ThelivingstatusofleftoverolderadultsinruralareasofChina:Asurveyof
258olderadultsinHunanProvince. HubeiSocialScience,8,57–60(inChinese).
Chen,X.,&Silverstein,M.(2000).Intergenerationalsocialsupportandthepsychologicalwell-
beingofolderparentsinChina.ResearchonAging,22,43–60.
Cheng,S.,&Chan,A.C.M.(2006).Filialpietyandpsychologicalwell-beinginwellolder
Chinese.JournalsofGerontology,61B(5),262.
Chi,I.,&Chou,K.L.(2001).SocialsupportanddepressionamongelderlyChinesepeoplein
HongKong. InternationalJournalofAgingandHuman Development,52,231–252.
ChinaCensusBureau(2004).Thenumbersofout-migrationsamongrurallaborers.Survey
World,4,11(inChinese).
Chou,K.-L.,&Chi,I.(2005).PrevalenceandcorrelatesofdepressioninChineseoldest-old.
InternationalJournalofGeriatricPsychiatry,20(1),41–50.doi:10.1002/gps.1246
Chow,N.W.S.(2001).ThepracticeoffilialpietyamongtheChineseinHongKong.InI.Chi,
N.L.Chappell,&J.Lubben(Eds.),ElderlyChineseinPacificRimcountries:Social
supportandintegration(pp.125–136).HongKong:HongKongUniversitiesPress.
Coleman,J.S.(1988).Socialcapitalinthecreationofhumancapital.AmericanJournalof
Sociology,94,95–120.doi:10.1086/228943
Coleman,J.S.(1990).Foundationsofsocialtheory.Cambridge,MA:HarvardUniversityPress.
102
Cong,Z.,&Silverstein,M.(2004).Intergenerationallivingarrangements,socialsupport
exchangeandthepsychologicalwell-beingofolderpeopleinruralchina.Gerontologist,
44(1),377–390.
Cong,Z.,&Silverstein,M.(2008a).Intergenerationalsupportanddepressionamongeldersin
ruralChina:Doin-lawsmatter?JournalofMarriageandFamily,70,599–612.
Cong,Z.,&Silverstein,M.(2008b).Intergenerationaltime-for-moneyexchangesinruralChina:
Doesreciprocityreducedepressivesymptomsofoldergrandparents?ResearchinHuman
Development,5(1),6–25.
Cong,Z.,&Silverstein,M.(2011).Intergenerationalexchangebetweenparentsandmigrantand
nonmigrantsonsinruralChina.JournalofMarriageandFamily,73(1),93–104.
doi:10.1111/j.1741-3737.2010.00791.x
Cong,Z.,&Silverstein,M.(2012).Avignettestudyongenderedfilialexpectationsofeldersin
ruralChina.JournalofMarriageandFamily,74,510–525.
Cox,D.(1987).Motivesforprivateincometransfers,JournalofPoliticalEconomy,95,508–546.
Cox,M.,&Paley,B.(1997).Familiesassystems.AnnualReviewofPsychology,48,243–267.
Cox,D.,&Rank,M.(1992).Inter-vivostransfersandintergenerationalexchange,Reviewof
EconomicsandStatistics,74,305–334.
Croll,E.J.(2006).TheintergenerationalcontractinthechangingAsianfamily.Oxford
DevelopmentStudies,34,473–491.
Dellman-Jenkins,M.,Blankemeyer,M.,&Pinkard,O.(2000).Youngadultchildrenand
grandchildreninprimarycaregiverrolestoolderrelativesandtheirservicesneeds.Family
Relations,49(2),177–186.
103
Dimmock,B.,Bornat,J.,Pearce,S.&Jones,D.(2004).Intergenerationalrelationshipsamong
UKstepfamilies.InHarper,S.(Ed.), Familiesinageingsocieties(pp.82–94).Oxford,
England:OxfordUniversityPress.
Du,P.,Ding,Z.,Li,Q.,&Gui,J.(2004).Nongcunzinvwaichuwugongduiliushoulaorende
yingxiang(Theimpactofoutlabormigrationontheelderlystayersinruralareas)
PopulationResearch,28(6),44–52(inChinese).
Euler,H.A.,Hoier,S.,&Rohde,P.A.(2001).Relationship-specificclosenessof
intergenerationalfamilyties:Findingsfromevolutionarypsychologyandimplicationsfor
modelsofculturaltransmission.JournalofCross-CulturalPsychology,32,147–158.
FamilyCaregiverAlliance.(2003).Womenandcaregiving:Factsandfigures(FactSheet).San
Francisco,CA.Retrievedfrom
http://www.caregiver.org/caregiver/jsp/content_node.jsp?nodeid=892
Feng,N.,&Xiao,N.(2007).PopulationaginginChinaasreflectedbytheresultsofthe2005
populationsamplesurvey.Paperpresentedatthe23rdPopulationCensusConferenceon
Utilizationofthe2000and2005RoundsofAsia-PacificCensusesinChristchurch,New
Zealand.Retrievedfromhttp://www.eastwestcenter.org/events/23rd-population-census-
conferenceonApril6,2013.
Ferguson,N.,&Douglas,G.(2004).Grandparentingindivorcedfamilies.Bristol,Avon:Policy
Press.
Ferrer,E.,&McArdle,J.J.(2003).Alternativestructuralmodelsformultivariatelongitudinal
dataanalysis.StructuralEquationModeling,10,493–524.
Fruhauf,C.,Jarrott,S.,&Allen,K.(2006).Grandchildren'sperceptionsofcaringfor
grandparents.JournalofFamilyIssues,27(7),887–911.
104
Fruhauf,C.,&Orel,N.(2008).Developmentalissuesofgrandchildrenwhoprovidecareto
grandparents.InternationalJournalofAging&Human Development,67(3),209–230.
Furstenberg,F.F.,&Kaplan,S.B.(2004).Socialcapitalandthefamily.InJ.L.Scott,J.K.
Treas,&M.P.Richards(Eds.).TheBlackwellcompanion tothesociologyoffamilies(pp.
218–232).Malden,MA:BlackwellPublishers.
Gauthier,A.(2002).Theroleofgrandparents.CurrentSociology,50(2),295–307.
Giles,J.,&Mu,R.(2007).Elderlyparenthealthandthemigrationdecisionsofadultchildren:
EvidencefromruralChina.Demography,44(2),265–288.
Goldstein,A.,Zhigang,G.,andGoldstein,S.(1997).Therelationshipofmigrationtochanging
householdheadshippatternsinChina,1982–1987.Population Studies,51,75–84.
Goode,W.J.(1963).Worldrevolutionandfamilypatterns.London:FreePressofGlencoe.
Goodkind,D.,&West,L.A.(2002).China'sfloatingpopulation:Definitions,dataandrecent
findings.UrbanStudies,39(12),2237–2250.
Goodman,C.C.(2003).Intergenerationaltriadsingrandparent-headedfamilies.Journalof
Gerontology:SocialScience,58B(5),281–289.
Goodman,C.C.,&Silverstein,M.(2002).Grandmothersraisinggrandchildren:Familystructure
andwell-beinginculturallydiversefamilies.Gerontologist,42(5),676–689.
Gottfries,C.G.(2001).Latelifedepression.EuropeanArchivesofPsychiatryandClinical
Neuroscience,251(Suppl.2),57–61.doi:10.1007/BF03035129
Grundy,E.,&Henretta,J.C.(2006).Betweenelderlyparentsandadultchildren:Anewlookat
theintergenerationalcareprovidedbythe‘sandwichgeneration.’Ageing&Society,26,
707–722.
105
Guo,B.,Pickard,J.,&HuangJ.(2007).Aculturalperspectiveonhealthoutcomesofcaregiving
grandparents.JournalofIntergenerationalRelationships,5(4),25–40.
Guo,M.,Aranda,M.,&Silverstein,M.(2009).Theimpactofout-migrationontheinter-
generationalsupportandpsychologicalwellbeingofolderadultsinruralChina.Ageing&
Society,29,1085–1104.
Guo,M.(2011).IntergenerationalrelationsbetweenChineseruralelders andtheiradult
children:A multifacetedperspective(Doctoraldissertation).AvailablefromProQuest
DissertationsandThesesdatabase.(USCNo. 901469588;UMINo.3477903).
Harper,S.(2005).Understandinggrandparenthood.InM.Johnson(Ed.),Handbook ofageand
ageing(pp.422–428).Cambridge:CambridgeUniversityPress.
Harper,S.,Smith,T.,Lechtman,Z.,Ruicheva,I.,&Zeilig,H.(2004).GrandmothercareinLone
ParentFamilies.Oxford:OxfordInstituteofAgeing.
Harris,L.A.(2002).Grandparentinvolvementinchildcare:Relationsbetweencaregiving
levels,grandparentrolemeaning,rolesatisfaction,andpsychologicalwell-being
(Unpublisheddoctoraldissertation).SouthernIllinoisUniversity,Carbondale.
Harwood,J.,&Lin,M.-C.(2000).Affiliation,pride,exchange,anddistanceingrandparents’
accountsofrelationshipwiththeircollege-agedgrandchildren.JournalofCommunication,
50(3),31–47.
He,W.,Sengupta,M.,Zhang,K.,&Guo,P.(2007)USCensusBureau,internationalpopulation
reports,P95/07-2healthandhealthcareoftheolderpopulationinurbanandruralChina:
2000.Washington,DC:U.S.GovernmentPrintingOffice.
Hermalin,A.I.(Ed.).(2002).Thewell-beingoftheelderlyinAsia:Afour-countrycomparative
study.AnnArbor,MI:UniversityofMichiganPress.
106
Hodgson,L.G.(1992).Adultgrandchildrenandtheirgrandparents:Theenduringbond.
InternationalJournalofAgingandHumanDevelopment,34(3),209–225.
Hoff,A.(2007).Patternsofintergenerationalsupportingrandparent-grandchildandparent-child
relationshipsinGermany.Ageing&Society,27,643–665.
Hoff,A.,&Tesch-Römer,C.(2007).Familyrelationsandaging:Substantialchangessincethe
middleofthelastcentury?InH.W.Wahl,C.Tesch-Römer,&A.Hoff(Eds.),New
dynamicsinoldage:Individual,environmentalandsocietalperspectives(pp.65–84).
Amityville,NY:Baywood.
Hsiao,H.Y.,Xu,L.,Chi,I.,&Silverstein,M.(inpress).Multipleroleoccupancyasaprotective
factorforChineseolderadults’psychologicalwell-being:Alongitudinalstudyusinglatent
growthcurveanalysis.JournalofAgingandMentalHealth.
http://dx.doi.org/10.1080/13607863.2013.771614
Hu,L.,&Bentler,P.M.(1999).Cutoffcriteriaforfitindexesincovariancestructureanalysis:
Conventionalcriteriaversusnewalternatives.StructuralEquationModeling,6,1–55.
Huang,J.F.,Wong,R.H,Chen,C.C.,Mao,I.F.,Huang,C.C.,Chang,W.H.,&Wang,L.
(2011).Trajectoryofdepressionsymptomsandrelatedfactorsinlaterlife:Apopulation
basedstudy.JournalofAffectiveDisorders,133,499–508.
Huang,Q.(2009).Thelivingconditionsofleftoverolderadults:AsurveyinMaobeitown,
XuanhancountyofSichuanProvinceofChina. EconomicandSocialDevelopment,7(6),
112–115(inChinese).
Jiang,L.(1995).Changingkinshipstructureanditsimplicationsforold-agesupportinurbanand
ruralChina.PopulationStudies,49(1),127–145.
107
Kalmijn,M.(2008).Theeffectsofseparationanddivorceonparent-childrelationshipinten
EuropeWelfareStates.InC.Saraceno(Ed.),Families,ageingandsocialpolicy:
IntergenerationalsolidarityinEuropeancountries(pp.170–191).Northampton,MA:
EdwardElgarPublishing,Inc.
Kerr,M.,&Bowen,M.,(1988).Familyevaluation:An approachbasedonBowenTheory.New
York,NY:Norton.
King,V.,&Elder,G.H.,Jr.(1995).Americanchildrenviewtheirgrandparents:Linkedlives
acrossthreeruralgenerations.JournalofMarriageandtheFamily,57(1),165–178.
King,V.,Russell,S.,T.,&Elder,G.H.,Jr.(1998).Grandparentinginfamilysystems:An
ecologicalperspective.InM.E.Szinovacz(Ed.),Handbookongrandparenthood(pp.53–
69).Westport,CT:GreenwoodPress.
King,V.,Silverstein,M.,&Elder,G.H.,Jr.(2003).Relationswithgrandparents:RuralMidwest
versusurbanSouthernCalifornia.JournalofFamilyIssues,24(8),1044–1069.
Langer,N.(1990).Grandparentsandadultgrandchildren:Whatdotheydoforoneanother?
InternationalJournalofAgingandHumanDevelopment,31(2),101–110.
Leung,J.C.(1997).FamilysupportfortheelderlyinChina:Issuesandchallenges.Journalof
AgingandSocialPolicy,9(3),87–101.doi:10.1300/J031v09n03_05
Li,Q.M.(2004).ConcernoftheskippedgenerationhouseholdinruralChina. Forumof
PopulationandDevelopment,28(4),25–36(inChinese).
Lin,J-P.,&Yi,C-C.(2011).FilialnormsandintergenerationalsupporttoagingparentsinChina
andTaiwan.InternationalJournalofSocialWelfare,20,109–120.doi:10.1111/j.1468-
2397.2011.00824.x
108
Lin,N.(1982).Socialresourcesandinstrumentalaction.In:P.V.Marsden&N.Lin(Eds.),
Socialstructureandnetworkanalysis(pp.131–145).BeverlyHills,CA:Sage.
Lin,N.(1983).Socialresourcesandsocialactions:Aprogressreport.Connections,6,10–16.
Lincoln,K.D.(2000).Socialsupport,negativesocialinteractions,andpsychologicalwell-being.
SocialServiceReview,74(2),231–252.
Lincoln,K.D.,Chatters,L.M.,&Taylor,R.J.(2003).PsychologicaldistressamongBlackand
WhiteAmericans:Differentialeffectsofsocialsupport,negativeinteractionandpersonal
control.JournalofHealthandSocialBehavior,44(3),290–407.
Lingler,J.H.,Sherwood,P.R.,Crighton,M.H.,Song,M.K.andHapp,M.B.(2008).
Conceptualchallengesinthestudyofcaregiver-carerecipientrelationships.Nursing
Research,57(5),367–372.
Lou,V.W.(2010).LifesatisfactionofolderadultsinHongKong:Theroleofsocialsupport
fromgrandchildren.SocialIndicatorsResearch,95,377–391.
Lou,V.W.(2011).DepressivesymptomsofolderadultsinHongKong:Theroleofgrandparent
reward.InternationalJournalofSocialWelfare,20,135–147.
Luo,B.(2009).Theimpactofrural-urbanmigrationonfamilialeldercareinruralChina
(Unpublisheddoctoraldissertation).GeorgiaStateUniversity,Atlanta,GA.
Lüescher,K.,&Pillemer,K.(1998).Intergenerationalambivalence:Anewapproachtothe
studyofparent-childrelationsinlaterlife.JournalofMarriageandtheFamily,60(2),413–
425.
Lye,D.N.(1996),Adultchild-parentrelationships.AnnualReviewofSociology,22,79–102.
109
Marshall,V.W.&Bengtson,V.L.(2011).Theoreticalperspectivesonthesociologyofaging.In
R.A.Settersten,Jr.,&JacquelineL.A.(Ed.),Handbookofsociologyofaging(pp.17–
33). NewYork:Springer.
Ma,S.S.,Qin,X.,Chen,R.L.,&Hu,Z.(2006).Investigationoftheincidenceofgeriatric
depressioninaruralcommunityofAnhuiprovince.Chinese Journal ofPublicHealth,22,
385–386(inChinese).
McArdle,J.J.(2001).Alatentdifferencescoreapproachtolongitudinaldynamicanalysis.InR.
Cudeck,S.duToit,&D.Sörbom(Eds.),Structuralequationmodeling:Presentandfuture
(pp.341–380).Lincolnwood,IL:ScientificSoftwareInternational.
McArdle,J.J.,&Hamagami,F.(2001).Latentdifferencescorestructuralmodelsforlinear
dynamicanalyseswithincompletelongitudinaldata.InL.M.Collins&A.G.Sayer(Eds.),
Newmethodsfortheanalysisofchange(pp.139–175).Washington,DC:American
PsychologicalAssociation.
McCutcheon,A.L.(1987).Latentclassanalysis.NewburyPark,CA:SagePublications,Inc.
McElroy,M.,&Horney,M.J.(1981).Nash-bargainedhouseholddecisions:Towarda
generalizationofthetheoryofdemand.InternationalEconomicReview,22,333–349.
MacCallum,R.C.,&Austin,J.T.(2000).Applicationsofstructuralequationmodelingin
psychologicalresearch.AnnualReviewofPsychology,51,201–226.
Mangen,D.J.,Bengtson,V.L.,&LandryP.H.,(Eds.).(1988).Themeasurementof
intergenerationalrelations.BeverlyHills,CA:Sage.
McGowen,M.R.,Ladd,L.,&Strom,R.D.(2006).On-lineassessmentofgrandmother
experienceinraisinggrandchildren.EducationalGerontology,32,669–684.
110
Minuchin,P.(1988).Relationshipswithinthefamily:Asystemsperspectiveondevelopment.In
R.A.Hinde&J.Stevenson-Hinde(Eds.),Relationshipswithinfamilies:Mutual
influences.NewYork,NY:OxfordUniversityPress.
Molm,L.D.,&Cook,K.S.(1995).Socialexchangeandexchangenetworks.InK.S.Cook,G.
A.Fine,&J.S.House(Eds.), Sociological perspectives on social psychology(pp.209–235).
Boston:AllynandBacon.
Monserud,M.A.(2008a).Intergenerationalrelationshipsandaffectualsolidaritybetween
grandparentsandyoungadults.JournalofMarriageandFamily,70(1),182–195.
Monserud,M.A.(2008b).Youngadults’relationshipswithgrandparents:Parents’
intergenerationalties andgrandchildren’sadultroles(Unpublisheddoctoraldissertation).
WashingtonStateUniversity,Pullman,WA.
Monserud,M.A.(2010).Continuityandchangeingrandchildren’sclosenesstograndparents:
Consequencesofchangingintergenerationalties.Marriage&FamilyReview,46,366–388.
Morrow-Howell,N.,Hinterlong,J.E.,Sherraden,M.,&Rozario,P.A.(2001).Advancing
researchonproductivityinlaterlife.InN.Morrow-Howell,J.E.Hinterlong,&M.
Sherraden(Eds.),Productiveaging:ConceptsandChallenges(pp.285–311).Baltimore,
MD:JohnsHopkinsUniversityPress.
Muthén,B.(2004).Latentvariableanalysis:Growthmixturemodelingandrelatedtechniques
forlongitudinaldata.InD.Kaplan(Ed.),Handbookofquantitativemethodologyforthe
socialsciences(pp.345–368).NewburyPark,CA:Sage.
NationalBureauofStatisticsofChina.(2005).http://www.stats.gov.cn/english/#
NationalBureauofStatisticsofChina.(2010).Chinastatisticalyearbook:2010.Beijing:China
StatisticsPress.
111
Nylund,K.L.,Asparouhov,T.,&Muthén,B.(2007).Decidingonthenumberofclassesin
latentclassanalysisandgrowthmixturemodeling:AMonteCarlosimulationstudy.
StructuralEquationModeling,14(4),535–569.
Orth,U.,Berking,M.,Walker,N.,Meier,L.L.,&Znoj,H.(2008).Forgivenessand
psychologicaladjustmentfollowinginterpersonaltransgressions:Alongitudinalanalysis.
JournalofResearchinPersonality,42,365–385.
Pei,X.&Pillai,V.K.(1999).OldageinChina:Theroleofthestateandthefamily.
InternationalJournalofAgingandHumanDevelopment,49,197–212.
Piercy,K.W.(1998).Theorizingaboutfamilycaregiving:Theroleofresponsibility.Journalof
MarriageandFamily,60(1),109–118.
Pruchno,R.,&McKenney,D.(2002).Psychologicalwell-beingandblackandwhite
grandmothersraisinggrandchildren:Examinationofatwo-factormodel.Journalof
Gerontology,57(5),444–452.
Radloff,L.(1977).TheCES-Dscale:Aself-reportdepressionscaleforresearchinthegeneral
population.AppliedPsychologicalMeasurement,1(3),385–401.
Raftery,A.E.(1986).Choosingmodelsforcross-classifications.AmericanSociologicalReview,
51,145–146.
Reinhardt, J. P., Boerner, K., & Horowitz, A. (2006). Good to have but not
to use: Differential impact of perceived and received support on well-
being. Journal of Social and Personal Relationships, 23(1), 117–129.
Rosenthal,C.J.(1985).Kin-keepinginthefamilialdivisionoflabor.JournalofMarriageand
theFamily,47,965–974.
112
Ross,C.E.,&Mirowsky,J.(2002).Familyrelationships,socialsupportandsubjectivelife
expectancy.JournalofHealthandSocialBehavior,43(4),469–485.
Rossi,A.,andRossi,P.(1990).Ofhumanbonding:Parent-childrelationsacrossthelifecourse.
NewYork:AldinedeGruyter.
Ruiz,S.,&Silverstein,M.(2007).Relationshipswithgrandparentsandtheemotionalwell-beingof
lateadolescentandyoungadultgrandchildren.JournalofSocialIssues,63(4),793–808.
Salmon, C. A., & Daly, M. (1998). Birth order and familial sentiment:
Middleborns are different. Human Behavior and Evolution, 19, 299–
312.
Schafer,J.L.,&Graham,J.W.(2002).Missingdata:Ourviewofthestateoftheart.
PsychologicalMethods,7,147–177.
Silverstein,M.(2006).Intergenerationalfamilytransfersinsocialcontext.InR.H.Binstock&L.
K.George,Handbookofagingandthesocialsciences(6thed.,pp.165–180).Amsterdam:
ElsevierAcademicPress.
Silverstein,M.,&Bengtson,V.L.(1994).Doesintergenerationalsocialsupportinfluencethe
psychologicalwell-beingofolderparents?Thecontingenciesofdeclininghealthand
widowhood.SocialScienceandMedicine,38,943–957.
Silverstein,M.,Cong,Z.,&Li,S.Z.(2006).Intergenerationaltransfersandlivingarrangements
ofolderpeopleinruralChina:Consequencesforpsychologicalwell-being.Journalof
Gerontology:SocialSciences,61(B),256–266.
113
Silverstein,M.,Conroy,S.,WangH.,Giarrusso,R,&Bengtson,V.L.(2002).Reciprocityin
parent-childrelationsover theadultlifecourse.Journal of Gerontology: Social Sciences, 57,
3–13.
Silverstein,M.,Giarusso,R.,&Bengtson,V.L.(1998).Intergenerationalsolidarityandthe
grandparentrole.InM.E.Szinovacz(ed.),Handbook ongrandparenthood(pp.144–158).
Westport,CT:GreenwoodPress.
Smith,T.L.(2011).Adultgrandchildrenprovidingcaretofrailelderlygrandparents
(Unpublisheddoctoraldissertation).UniversityatAlbany,StateUniversityofNewYork.
Song,L.,Li,S.,Zhang,W.,&Feldman,M.W.(2007).Intergenerationsupportandself-rated
healthoftheelderlyinruralChina:AninvestigationinChaohu,Anhuiprovince.InZ.Yi,D.
Poston,D.A.Vlosky,&D.Gu(Eds.). HealthylongevityinChina:Demographic,
socioeconomic,andpsychologicaldimensions.Dordrecht:SpringerPublisher.
Song,X.M.,&Chi,I.(2001).Theroleofsocialsupportintherelationshipbetweenphysical
healthstrainanddepressionofelderlyChinese.InI.Chi,N.Chappell,&J.Lubbern(Eds.),
ElderlyChineseinPacificRimcountries:Socialsupportandintegration(pp.189–200).
HongKong:HongKongUniversityPress.
Spence,S.A.,Black,S.R.,&Adams,J.P.(2001).Grandparentsandgrandparentinginarural
southernstate:Astudyofdemographiccharacteristics,roles,andrelationships.Journalof
FamilyIssues,22(4),523–534.
Sun,R.(2002).OldagesupportincontemporaryurbanChinafrombothparents’andchildren’s
perspectives.ResearchonAging,24(3),337–359.
114
Taylor,J.E.(2001).Microeconomicsofglobalization:EvidencefromMexico,China,El
Salvador,andtheGalapagosIslands.ReporttotheLatinAmericaandCaribbeanRegional
Office.WorldBank,WashingtonDC.
Taylor,M.G.,&Lynch,S.M.(2004).Trajectoriesofimpairment,socialsupport,anddepressive
symptomsinlaterlife.JournalofGerontology(SocialScience),59B(4),238–246.
Taylor,S.E.,Welch,W.T.,Kim,H.S.,&Sherman,D.K.(2007).Culturaldifferencesinthe
impactofsocialsupportonpsychologicalandphysiologicalstressreactivity.
PsychologicalScience,18,831–837.
Thiele,D.,&Whelan,T.(2008).Therelationshipbetweengrandparentsatisfaction,meaning,
andgenerativity.InternationalJournalofAging&Human Development,66(1),21–48.
Tu,E.J.,Freedman,V.A.,&Wolf,D.A.(1993).KinshipandfamilysupportinTaiwan.
ResearchonAging,15,465–486.
UnitedNations,PopulationDivision(2004).Worldpopulationprospects.InThe2002revision:
Comprehensivetables.NewYork:UnitedNations.
Voils,C.I.,Allaire,J.C.,Olsen,M.K.,Steffens,D.C.,Hoyle,R.H.,&Bosworth,H.B.(2007).
Five-yeartrajectoriesofsocialnetworksandsocialsupportinolderadultswithmajor
depression.InternationalPsychogeriatrics,19(6),1110–1124.
Whitbeck,L.B.,Hoyt,D.R.&Huck,S.M.(1993).Familyrelationshiphistory,contemporary
parent-grandparentrelationshipquality,andthegrandparent-grandchildrelationship.
JournalofMarriageandtheFamily,55(4),1025–1035.
Whyte,M.K.(Ed.).(2003). China’s revolutions andintergenerational relations.AnnArbor,MI:
UniversityofMichigan,CenterforChineseStudies.
115
Whyte,M.(2004).FilialobligationsinChinesefamilies:Paradoxesofmodernization.InC.Ikels
(Ed.),Filialpiety:PracticeanddiscourseincontemporaryEastAsia(pp.106–127).
Stanford,CA:StanfordUniversityPress.
Wolff. J., & Agree, E. M. (2004). Depression among recipients of informal
care: The effects of reciprocity, respect, and adequacy of support. Journal
of Gerontology, 59B(3), 173–180.
Xu,Y.(2001).FamilysupportforoldpeopleinruralChina.SocialPolicy&Administration,
35(3),307–320.doi:10.1111/1467-9515.00235
Xu,L.,&Chi,I.(2011).LifesatisfactionamongruralChinesegrandparents:Therolesof
intergenerationalfamilyrelationshipandsupportexchangewithgrandchildren.International
JournalofSocialWelfare,20,148–159.doi:10.1111/j.1468-2397.2011.00809.x
Yang,Y.(2007).Isoldagedepressing?Growthtrajectoriesandcohortvariationsinlatelife
depression.JournalofHealthandSocialBehavior,48(March),16–32.
Yue,X.,&Ng,S.H.(1999).FilialobligationsandexpectationsinChina:Currentviewsfrom
youngandoldpeopleinBeijing.AsianJournalofSocialPsychology,2(2),215–226.
Yeh,K.,&Bedford,O.(2003).Atestofthedualfilialpietymodel. AsianJournalofSocial
Psychology,6,215–228.
Yip,W.,Subramanian,S.V.,Mitchell,A.D.,Lee,D.T.S.,Wang,J.,&Kawachi,I.(2007).
Doessocialcapitalenhancehealthandwell-being?EvidencefromruralChina.Social
Science&Medicine,64(1),35–49.doi:10.1016/j.socscimed.2006.08.027
Zhan,H.J.,&Montgomery,R.J.V.(2003).GenderandeldercareinChina:Theinfluenceof
filialpietyandstructuralconstraints.GenderandSociety,17(2),209–229.
116
Zhang,X.(1998).Some legalandsocialissuesconcerningtherurallabormigrationinChina.
Unpublishedpaper.
Zhang,W.&Li,S.(2004).Thedeterminantsoflivingarrangementforruralolderadultswhen
theiradultchildrenmigratedtourbancity.ChineseJournalofPopulationScience,1,42–49
(inChinese).
Zhang,L.,Xu,Y.,Nie,H.,Zhang,Y.,&Wu,Y.(2012).Theprevalenceofdepressivesymptoms
amongtheolderinChina:Ameta-analysis.International Journal of Geriatric Psychiatry,
27(9),900–906.doi:10.1002/gps.2821
Zimmer,Z.,&Kwong,J.(2003).Familysizeandsupportofolderadultsinurbanandrural
China.Demography,40(1),23–44.
Zuo,D.(2011).AgepatternofintergenerationalsupportoftheruralfamiliesinChina:Alife
courseperspective(Unpublisheddoctoraldissertation).Xi’anJiaotongUniversity,
Shaanxi,China(inChinese).

Abstract (if available)

Abstract Social support is especially important for people in late life due to physical and life-course changes. The majority of older adults in China receive social support primarily from their adult children. However, for many rural residents, supporting older adults has not been easy due to lower standards of living and migration to urban cities. On the other hand, due to increased longevity, a large proportion of older adults have adult grandchildren who may also serve as an important source of support. However, little attention has been paid to the dynamics of grandchild support among older adults in rural China, despite demographic shifts that indicate grandparents and grandchildren are left behind and live with each other in rural China. This dissertation examined how adult grandchildren provide family support to older grandparents and how this support affects the psychological well-being of older adults in rural China. ❧ This dissertation consists of three independent studies. Its main purpose was to add to the existing knowledge regarding intergenerational support by investigating the dynamics of support provided to older adults by their grandchildren in rural China. Using a longitudinal dataset from a group of older adults in a rural region of the Anhui province in mainland China, this dissertation had the following specific aims: (1) to investigate support exchange between grandparents and grandchildren and associated factors

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Family relationships and their influence on health outcomes over time among older adults in rural China

PDF

Children 's migration and the financial, social, and psychological well-being of older adults in rural China

PDF

The effects of time and space on health status of older adults in China

PDF

The immediate and long term legacy of relationships with grandparents for the well-being of grandchildren

PDF

Intergenerational transfers in rural China: Do children's gender and birth order matter?

PDF

Life transitions, leisure activity engagement, and cognition among older adults

PDF

The impact of post-traumatic stress and post-traumatic growth on young adult cancer survivors

PDF

The relationship of expectation and satisfaction of filial piety with stroke-related outcomes of Chinese older adults in stroke rehabilitation

PDF

A study of depression disclosure among Latino older adults in Los Angeles

PDF

Investigating the effectiveness of a social work intervention on reducing hospital readmissions among older adults

PDF

My Life, My Wishes, a death education program for community-dwelling Chinese older adults: development and pilot testing

PDF

COACH: Connecting Older Adults to Community for Health - an evidence-based program to address older adult social isolation and loneliness

PDF

Examining the longitudinal influence of the physical and social environments on social isolation and cognitive health: contextualizing the role of technology

PDF

Intergenerational transmission of values and behaviors over the family life course

PDF

Mechanisms of the effect of involuntary retirement on older adults' health and mental health

PDF

The effects of physical functioning and public stigma on psychological distress as mediated by cognitive and social factors among Korean survivors of childhood cancer

PDF

Internet communication use, psychological functioning and social connectedness at older ages

PDF

Age integration in late life: sociodemographic & psychosocial correlates of intergenerational-only, peer-only, and age-integrated social networks

PDF

A multilevel model of work-family conflict in a global context: a comparative study across 24 countries

PDF

Integrative care strategies for older adults experiencing co-occurring substance use and mental health disorders (I-CARE)…

Asset Metadata

Creator Xu, Ling (author)

Core Title Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work / Gerontology

Publication Date 07/05/2013

Defense Date 04/25/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag China,grandparents and grandchildren,intergenerational support,OAI-PMH Harvest,older adults,psychological well-being

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chi, Iris (committee member), Chou, Chih-Ping (committee member), Palinkas, Lawrence A. (committee member)

Creator Email linderling@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-284613

Unique identifier UC11293575

Identifier etd-XuLing-1742.pdf (filename),usctheses-c3-284613 (legacy record id)

Legacy Identifier etd-XuLing-1742.pdf

Dmrecord 284613

Document Type Dissertation

Format application/pdf (imt)

Rights Xu, Ling

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA