Commongeneticin - PDF document

Commongeneticin”uencesunderpinreligiosity,communityintegration,andexistentialuncertainty GaryJ.Lewis ,TimothyC.Bates , DivisionofPsychology,SchoolofNaturalSciences,UniversityofStirling,StirlingFK94LA,UK DepartmentofPsychology,UniversityofEdinburgh,7GeorgeSquare,EdinburghEH89JZ,UK ExistentialuncertaintyTwinstudy abstractAlthoughgeneticfactorsunderpinindividualdifferencesinreligiosity,thepsychologicalmechanismsthroughwhichsuchin”uencesaremanifestedarepresentlyunknown.Religiosityisassociatedwithcon-cernsforcommunityintegrationandexistentialcertainty,suggestingthatheritablein”uencesunderly-ingsuchsentimentsmayoverlapwithheritablein”uencesunderpinningreligiosity.Herewetestedthishypothesiswithinageneticallyinformativedesign,usingalarge,nationally-representativetwinsample.Aspredicted,heritableeffectsunderlyingcommunityintegrationandexistentialuncertaintywere 1.Introduction Religiousbeliefsandvaluesappearinallhumansocietiesre-cordedtodate(Bulbulia,2004;Swatos,1998),althoughindividuallevelsofreligiosityalsodiffersigni“cantly(e.g.Lewis,Ritchie,&Bates,2011;Norenzayan&Gervais,2013;Saroglou,2010).Assuch,theoriginsofvariationinreligiousbeliefhavebeenofenduringinteresttosocialandbehavioralscientists(c.f.Dennett,2006 =.30…Bouchard,2004;Bouchard,McGue,Lykken,&Tellegen,1999;DOnofrioetal.,1999;Kendleretal.,2003).Thisobservation,how-ever,givesrisetotheimportantquestionofwhatpsychologicalmechanismsmediatethisbiologicalin”uenceonreligiosity.Cur-rently,littleisknownofthemechanism(s)throughwhichgeneticfactorsexerttheireffectsonreligiousbelief;however,inpsycho-logicalresearch,twobroadandrelevantaccountshavebeenpro-posedconcerningneedsforcommunityandexistentialcertainty,andthisresearchcanbeusedtoguidepredictionsatthegeneticlevel.Belowweintroducetheseperspectives,followedbyamulti- religiosity.First,however,weoutlineworkestablishingagenetic componenttoreligiosity. 2.Geneticbasesofreligiosity Religiosityiscommonlyviewedastheproductofenvironmen-talin”uencesdesignedtotransmitandreinforceprevailingcul-turalvalues(Batson,Schoenrade,&Ventis,1993;Dudley&Dudley,1986).Thisassertionhasbeenchallenged,however,bystudiesutilizinggenetically-informativedesigns.Suchdesignsuti-lizeanaturallyoccurringphenomenon…humantwinning…todecomposebehavioralvariationintogeneticandenvironmentalcomponents(Neale&Cardon,1992 Utilizingthisapproach,Martinetal.(1986),inthe“rstgeneticanalysisofreligiosity,reportedsigni“cantheritableeffectsonbeliefsregardingsuchmattersastheimportanceofobservingtheSabbathandtruthfulnessoftheBible.ThisworkwasreplicatedandextendedWaller,Kojetin,Bouchard,Lykken,andTellegen(1990)thatindividualdifferencesinreligiousattitudes,theimportanceof Correspondingauthorsat:DivisionofPsychology,SchoolofNaturalSciences,UniversityofStirling,StirlingFK94LA,UK(G.J.Lewis).E-mailaddresses:(G.J.Lewis),(T.C.Bates). JournalofResearchinPersonality47(2013)398…405 ContentslistsavailableatSciVerseScienceDirectJournalofResearchinPersonalityjournalhomepage:www.elsevier.com/locate/jrp etal.,1999;Bradshaw&Ellison,2008;DOnofrioetal.,1999;Koenig,McGue,Krueger,&Bouchard,2005;Vance,Maes,&Kendler,2010 3.Communityandexistentialneeds:linkswithreligiosity Whilethetwinstudiesdetailedaboveprovidestrongevi-dencethatgeneticfactorsin”uenceindividualdifferencesinreli-gioussentiment,suchworkdoesnotprovideevidencethatsuchheritablefactorsre”ectgenesspeci“callyforreligion.Onealternativepositionwouldpositthatsuchgeneticfactorsoperateatmorebasic,andnon-theological,psychologicallevels.Thesein”uences,inturn,wouldgiverisetoindividualdifferencesinreligiosity,suchaswouldbethecaseifreligionexiststomeetbasichumanneeds(Graham&Haidt,2010;Kay,Shepherd,Blatz,Chua,&Galinsky,2010).Littleiscurrentlyknown,however,ofthemechanism(s)throughwhichgeneticfactorsexerttheiref-fectsonreligiousbelief,althoughtwocandidates…concernsregarding(1)communityintegrationand(2)existentialcertainty…areprominentinthebehavioralliterature,andaredetailedfurtherbelow. The“rstaccountofreligiosityisperhapsbestunderstoodbyadoptingwhatGrahamandHaidt(2010)callarelentlesslysocial(p.140)approach,arguingthatreligionexiststofacilitatecooper-ativecommunitybinding(alsoseeNorenzayan&Shariff,2008Thisfocusonsocialandcommunityfunctionsofreligionhasalonghistoryintheliterature.Forinstance,Durkheim(1915)that,[t]heideaofsocietyisthesoulofreligion(p.433).Thesocialbindingfunctionofreligionisalsosupportedbyresearchshowingthatcommunityintegrationissigni“cantlyassociatedwithhigherlevelsofreligiousbelief(Cavendish,Welch,&Leege,1998).Workonfactorsin”uencingchurchmembershipdropoutsprovidesaddi-tionalsupportforreligiousbeliefasamanifestationofsocialneeds.Forexample,Hartman(1976)askedformerMethodistchurchmemberswhytheydecidedtoleavetheirchurchandfoundthemostfrequentresponseentailedtheirfailuretofeel...loved,orwantedbyothersinthecongregation(p.40). Experimentalworkalsosupportslinksfromsocialbindingtoreligiosity:Forinstance,BirgegardandGranqvist(2004)thatsubliminalprimesre”ectingaseparationthreat(e.g.motherisgone)inducedsigni“cantsubsequentincreasesinthedesiretobeclosetoGod.Similarworkfoundthatindividu-alswhowereexposedtosciencevignettespurportingtoprovetheexistenceofGodreportedsigni“cantlyhigherbeliefinGod,althoughonlyifGodwasperceivedasacceptingandloving(vs.rejecting:Gebauer&Maio,2012).Ofinterest,thisinterac-tionwasattenuatedifparticipantsunderwentanattachmentprime,abriefvisualizationofasigni“cantother(vs.astranger),indicatingthattheobservedincreasedbeliefinGodmayre”ectamoregeneralizedneedtobelong(Gebauer&Maio,2012).Withthese“ndingsinmind,religiousindividualsmaybepredisposedtowardsgrouplifeingeneralratherthantoreligiousbeliefstrictlyconceived.Thismodel,then,wouldlocatespeci“creli-gionswithinthesetofgroupstructuresmeetingthisneed,ratherthanasauniquephenomenonperse. Thesecondaccountofreligiositywhichwedrawonheresug-geststhatreligiosityemergesasasourceofexistentialcertainty,generatingasenseofagencyandcontrol,andthusservingtore-ducefeelingsofanxiety(Inzlicht,McGregor,Hirsh,&Nash,2009Kay,Shepherd,etal.,2010Peterson,1999).Supportingthisidea,religiosityhasbeenshowntoincreasewhenperceivedcontrolisthreatened.Forexample,Kay,Shepherdetal.(2010)foundthatjustpriortoanelection,whengovernmentstabilitywaslow,individualsweremorelikelytobelieveinacontrollingGod,comparedwithimmediatelyafteranelection(whenasenseofgovernmentalstabilityhadreturned).Theseauthorsalso observedthatexperimentallymanipulatingperceivedcontrol throughthepresentationofvignettesdescribinghigh/lowlevelsofgovernmentalstabilityalsoresultedinlower/higherlevelsofbeliefinacontrollingGod.Independentworksupportingthisconclusionhasdemonstratedthatwhenperceivedpersonalcon-trolisundermined(byrecallinganunpleasantlifeeventwhereonelackedcontrol),beliefinGodisenhanced(Rutjens,vanderPligt,&vanHarreveld,2010).Interestingly,intheRutjensetal.study,thisincreasedbeliefinGodwasonlyobservedintheabsenceofotherbeliefframeworksthathelpedtocreateor-derintheworld:individualsprimedtounderstandDarwinismasanorderlyprocesswithinevitableoutcomesdidnotshowen-hancedbeliefinGodwhenpersonalcontrolwasthreatened,althoughenhancedbeliefin(theorderlyformof)Darwinismwasobservedfortheseindividuals. WorkstemmingfromtheterrormanagementliteratureGreenberg,Pyszczynski,&Solomon,1986)…whichstressesthathumanpsychologicalconcernsarestronglyin”uencedbytheawarenessofdeath…hasalsolentsupporttothenotionthatreligiousbeliefservesaroleinreducingexistentialangst(cf.etal.,2010).Forinstance,NorenzayanandHansen(2006)thatmanipulatingmortalitysalienceledindividualstoreportgreaterlevelsofreligiosityandastrongerbeliefinGod.Further-JonasandFischer(2006)reportedthataf“rmingintrinsicreligiosityreduceddeaththoughtaccessibilityfollowingmortal-itysalience.Religiosity,then,canbeseenasaresponsetoactiv-ityinageneralexistentialcertaintysystem,servingtoenhancepersonalcontrolandprovidinganepistemicbufferfromarangeoffactorssuchasunpredictability,instability,andconcernsovermortalitythatexistinthisworld. 4.Thecurrentstudy Inlinewiththeworkdetailedabove,theliteratureisnowatapointwhereextensivepsychologicalresearchhasprovidedtestablehypothesesregardingtheoriginsoftheheritablein”u-encesunderlyingreligiosity.Thedemonstrationthatmanipulat-ingcommunityintegrationorexistentialuncertaintygivesrisetochangeinreligiosity,isconsistentwiththesepsychologicalprocessesshowinglinkswiththeheritablein”uencesonreligios-ity;thatis,itisplausiblethatthesesensitivitiestosocialandexistentialneedsbeingmetarethemselves,inpart,heritable.Insupportofthisposition,whiletheseconstructshavenotre-ceivedgeneticallyinformativestudyperse,analogousvariablestotheseconstructshavebeenobservedtocontainheritablevar-iation,furtherstrengtheningthepositionthatcommongeneticeffectsmaybepresentbetweentheseconstructsandreligiosity.Forexample,socialwarmthandgregariousness(i.e.elementsofextraversion)haveshowngeneticin”uences(e.g.Eid,Riemann,Angleitner,&Borkenau,2003);andarangeoftraitsre”ectinganxiety(e.g.Mackintosh,Gatz,Wetherell,&Pedersen,2006havealsoshownheritableeffects.Ofnote,muchpreviousbehav-ioralworklinkingcommunityandexistentialneedshasimpliedordirectlyarguedforcausallinkstoreligiosity;inthecurrentstudyweareunabletotestcausalpredictions:ratherwearelimitedheretotestsofgeneticandenvironmentalcorrelations.Accordingly,inthecurrentstudywetestedwhethermeasuresofcommunityintegrationandexistentialuncertaintyshowedoverlappingheritableeffectswithreligiositybyutilizingalarge,nationallyrepresentativesampleofidentical,ormonozygotic,andfraternal,ordizygotic,twins.Inlinewiththecorrelationalandexperimentalworklinkingbothcommunityintegrationandexistentialuncertaintywithreligiosity,wepredictedthatge-neticfactorsunderlyingindividualdifferencesintheseconstructswouldbesigni“cantlycorrelated. G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 399 5.Method 5.1.Participants Phenotypicdatawereavailablefor993pairsoftwinsassessedforreligiosity,communityintegration,andexistentialuncertaintyfollowingcontactbytheMacArthurFoundationSurveyofMidlifeDevelopmentintheUnitedStates(MIDUSI;Brim,Ryff,&Kessler,).Ofthemonozygotic(MZ)pairs,178weremale(meanage=44.47,SD=11.53)and192werefemale(meanage=43.56,SD=12.28).Ofthedizygotic(DZ)pairs,143weremale(meanage=44.25,SD=12.35),214werefemale(meanage=45.90,SD=12.53),and266wereopposite-sexpairs(meanage=45.77,SD=11.85).Byrace,93.2%ofthesamplewascomprisedofWhiteindividuals,3.3%ofBlackindividuals,1.7%ofNativeAmerican/Es-kimoindividuals,and1.8%ofindividualswhoreportedtheirraceasother.Byreligion,85.1%participantsidenti“edthemselvesasbelongingtoaChristianreligion,1.3%reportedbelongingtoaJewishdenomination,1.7%reportedbelongingtoaBuddhist,Hin-du,orotherdenomination,2.2%self-reportedasatheistoragnostic,andafurther9.7%reportednoreligiouspreference.By(highestlevelof)education,7.9%hadnotcompletedhighschool,28.2%werehighschoolgraduates,31%hadcompletedsomecollege/universityeducation,18.2%hadacollegedegree,2.9%hadcompletedsomegraduateschooleducation,and11.8%hadcompletedagraduate-leveldegree. 5.2.Measures Religiositywasmeasuredwiththefollowingitems:Howreli-giousareyou?;Howimportantisreligioninyourlife?;importantisitforyou…orwoulditbeifyouhadchildrennow…tosendyourchildrenforreligiousorspiritualservicesorinstruction(1:Notatall,to4:Very).Theseitemswerehighlycorrelated(.54….81,allvaluesCronbachsalpha=.85)andwerethussummedintoacompositereligiosityscore.Althoughadditionalindicatorsofreligiousbelief/commitmentwereavailable,wese-lectedonlythoseitemsthatdidnotmakeexplicitreferencetoso-cialmanifestationsofreligiosity(e.g.attendanceataplaceofworship,identi“cationwithareligiousgroup),ortoitemswherereligiositywasrelatedtoexistentialcertainty(e.g.coping,divinesupport).Thisomissionofsuchitemsservedtoavoidconfoundingthetestofourcorehypothesesthroughcontentoverlap,andthus(speci“cally)providesamoreconservativetestofourhypotheses. Communityintegrationwasmeasuredwiththefollowingthreeitems:IdontfeelIbelongtoanythingIdcallacommunity(reverse-scored);Ifeelclosetootherpeopleinmycommunity;Mycommu-nityisasourceofcomfort(1:Disagreestrongly,to7:Agreestrongly).Theseitemsweresigni“cantlycorrelated(.41….57,allvaluesCronbachsalpha=.74)andwerethussummedintoacompositescore. Existentialuncertaintywasmeasuredwiththefollowingtwoitems:Theworldistoocomplexforme;Icannotmakesenseofwhatsgoingonintheworld(1:Disagreestrongly,to7:Agreestrongly).Theseitemsweresigni“cantlycorrelated(.44,.01:Cronbachsalpha=.61)andwerethussummedintoacomposite 5.3.Analysis Theclassicaltwindesignpartitionsobservedvariationintothreelatentcomponents:Additivegeneticin”uences(A),shared-environmentalin”uences(C;environmentalin”uencesfosteringsimilaritieswithintwinpairs),andunique-environmentalin”u-ences(E;environmentalin”uencesservingtomakeindividuals withinatwinpairlesssimilar).Geneticeffectsareinferredwhen monozygotic(MZ)twinsaremoresimilarthandizygotic(DZ)twins,whereasshared-environmenteffectsareinferredwhenMZtwincorrelationsarelessthantwicethatoftheDZtwins.Un-ique-environmenteffectsareinferredwhenMZtwinsarecorre-latedlessthanatunityforagiventrait,andthisvariancecomponentthusalsocontainsmeasurementerror.Whiletheseheuristicsprovideaninstructiveguidetothepatternofrelativege-neticandenvironmentaleffects,modernapproachestypicallyuti-lizeamulti-groupstructuralequationmodelingframework,whichfacilitatesformaltestsofparametersigni“cance,aswellasfortheestimationofparametersinmultivariatemodels(Neale,2003 Priortoconductingtheanalyses,wecontrolledfortheeffectsofageandsex,andstandardizedresidualswereusedinallsubse-quentanalysesinlinewithstandardpractice(McGue&Bouchard,).Themodelswereestimatedbyfull-informationmaximum-likelihoodanalysisusingOpenMx(Bokeretal.,2010a,2010b)andRDevelopmentCoreTeam,2009 Thehypothesisthatgeneticvarianceinreligiosityisassociatedwiththegeneticin”uencesunderlyingcommunityintegrationandexistentialuncertaintywastestedinamultivariateCholeskydecomposition(Neale&Cardon,1992)ofadditivegenetic,shared-environment,andunique-environmentcovariancebe-tweenthemeasures.Thismodelspeci“esasmanylatentfactorsastherearevariablesforeachsourceofvariance(i.e.A,C,andE),witheachsubsequentfactorhavingonefewerpathwaysthantheprecedingfactor(seeFig.1).Inotherwords,foradditivegeneticef-fects(A)the“rstlatentfactorloadsonallofthemeasuredvari-ables:Thesubsequentlatentfactorsloadon…2,variables.Inthisway,eachfactoraccountsforasmuchoftheremainingvarianceaspossible,untilthelastfactoraccountsforjusttheremainingvarianceinthelastmeasuredvariable.Thisisrepeatedforshared-environmental(C)andunique-environmentalfactors(E).Neatlyforthepresentpurpose,thisarrangementoffac-torsmeansthatenteringameasuredvariabletotherightofothervariablesteststhedegreetowhichthisvariableisre”ectedinthegeneticand/orenvironmentallatentfactorsthatprimarilyaccountforvarianceintheprecedingmeasuredvariables.Themodelalso Fig.1.GraphicalrepresentationoftheCholeskydecomposition.:Shared-environment(C)effectswerealsomodeledbutareomittedhereforclarityof G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 revealsifgeneticand/orenvironmentalcovariancebetweenmea-suresispresent. Forcurrentpurposes,thecriticaltestconcernswhetherherita-blevarianceinreligiosityisoverlappingwithgeneticeffectsoncommunityintegrationandexistentialuncertainty.Ifthisisthecase,thenthespeci“cgeneticeffectsonreligiosity(i.e.thosege-neticin”uencesonreligiositythatdonotco-varywitheithercom-munityintegrationand/orexistentialuncertainty)willbeabletoberemovedfromthemodelwithoutsigni“cantlyworsening“t. 6.Results Thephenotypiccorrelationsbetweenstudyvariablesarede-tailedinTable1.Theseindicatethatatthephenotypiclevel,indi-vidualshighoncommunityintegrationarelowerinexistentialuncertainty,andsomewhatmorelikelytobereligious.Further-more,existentialuncertaintyandreligiosityappearunrelatedatthisobservationallevel.Thisnullassociationbetweenexistentialuncertaintyandreligiosityrunscontrarytoourpredictions;how-ever,itisnoteworthythatthestrengthofthegeneticallyinformedapproachadoptedhereisthatisallowsthedecompositionofthesesummarybehavioralmeasuresintomultiple,potentiallydistinctandcomplexcomponentsofin”uence.Inotherwords,geneticandenvironmentalcorrelationsneednotre”ectphenotypiccorre-lations(Purcell,2008).Accordingly,weincludedexistentialuncer-taintyinourbiometricanalyses. 6.1.Univariateanalyses Wenextexaminedtheunivariateheritabilityforeachofourmeasures.MZcorrelationswerenotablyhigherthanforDZpairsonallmeasures,implyingthepresenceofgeneticin”uencesonvariationineachmeasure(seeTable2).Assumptiontestinginuni-variateandmultivariatemodelssuggestedthatneitherscalarnorgeneralsex-limitationeffectsweresigni“cant(maleandfemaleDZpairs,andsameandopposite-sexgroupscouldbeequatedwithoutsigni“cantlossofmodel“tforanyofthevariables),althoughitisworthbrie”ynotingthatthecurrentsamplepos-sessedlimitedpowertodetectsucheffectsandthusshouldbeinterpretedtentatively.Accordingly,malesandfemaleswerepooledforsubsequentanalyses. Bothcommunityintegrationandexistentialuncertaintywerebestexplainedbymodelscontainingadditivegeneticandun-ique-environmenteffects(shared-environmenteffectscouldbere-movedwithoutsigni“cantlyworseningmodel“t: =0,=1; =0,=1,respectively):Additivegeneticandunique-envi-ronmenteffectsexplained31%and69%ofthevarianceincommu-nityintegration,and36%and64%ofthevarianceinexistentialuncertainty.Forreligiosity,additivegenetic,shared-environment,andunique-environmenteffectswereallsigni“cant,explaining26%,26%,and48%ofthevariance,respectively.FullresultsoftheunivariatemodelingaredetailedinTable2 6.2.Multivariateanalyses Wenextmovedtoatestofourcorehypotheses.Webuiltamul-tivariatemodelcomprisedofcommunityintegration,existentialuncertainty,andreligiosity(alsoseeAppendixA).Inthismodel,thegeneticeffectsspeci“ctoreligiositywereestimatedatjust.10.Thisvalueisconsiderablyreducedfromthegeneticpathesti-mateof.51forreligiosityintheunivariatemodel(i.e.thesquarerootoftheunivariateheritabilityestimateforreligiosity).Thissug-gests,then,thatcommunityintegrationandexistentialuncertaintytothismodelsharemostofthegeneticvariationinourmeasureofreligiosity.Thisissupportedbytherelativelyhighgeneticcorrela-tionsofcommunityintegrationandexistentialuncertaintywithreligiosityinthismodel….67and.33,respectively…againindicat-ingsharedgeneticin”uencesfromcommunityintegrationandexistentialuncertaintytoreligiosity.Thegeneticcorrelationbe-tweencommunityintegrationandexistentialuncertaintywas Toexplicitlytestwhethertheheritableelementofreligiositycouldbeunderstoodentirelyintermsofsharedgeneticlinkstocommunityintegrationandexistentialuncertainty,weexaminedwhetherthespeci“cgeneticeffecttoreligiositycouldberemovedfromthemodelwithoutsigni“cantlyworsening“t.Aspredicted,removingthisspeci“cgeneticeffectforreligiosityincurrednosig-ni“cantlossof“t( (1)=0.002,=.96),indicatingthatgeneticeffectsunderlyingcommunityintegrationandexistentialuncer-taintyweresuf“cienttoexplaintheheritablebasesunderpinningreligiosity. Wethenexaminedwhethergeneticeffectsunderlyingreligi-osityweresharedsolelybyeithercommunityintegrationorexistentialuncertainty:Neitheroftheseeffects,however,couldberemovedwithoutsigni“cantlossof“t.Removingthepathfromexistentialcertainty(a2)toreligiosity(andsoforcingthemodeltoexplainheritableeffectsonreligiositysolelythroughcommunityintegration),signi“cantlyworsened“t (1)=6.09,=.01),indicatingsigni“cantcommongeneticef-fectsbetweenexistentialuncertaintyandreligiosity.Similarly,removingthepathfromcommunityintegration(a1)toreligios-ity,forcingthemodeltoexplainheritableeffectsonreligiositysolelythroughexistentialuncertainty,alsosigni“cantlywors-ened“t( (1)=4.33,=.04).Theseresultsindicatedthatbothcommunityintegrationandexistentialuncertaintywererequiredtoaccountforthegeneticbasesunderpinningreligiosity.More-over,thismodelingshowshowsigni“cantgeneticeffectsbe-tweenexistentialuncertaintyandreligiosity,inspiteofanon-signi“cantbehavioralassociation,couldarise:Heritableeffectsraisingbothcommunityintegrationandreligiosity(a1)actedtodecreaseexistentialuncertainty;simultaneously,heritableef-fectsonexistentialuncertaintyactedtoincreasereligiosity(a2).Insum,thesecountervailingeffectsservetocanceleachotheroutleadingtoagreatlyreducedcorrelationatthephenotypiclevel.Thisobservationcanbeunderstoodwiththefollowingreal-worldexample:Lets“rstpositatheoreticalmodelofauto-mobileperformanceusingtwolatentfactors;“rstly,thatthemorecylindersacarhas,(a)themorehorsepoweritcangener-ate,(b)theworseitsmilespergallon(mpg)willbe,and(c)themoreitwillcosttobuy;secondly,thatstreamlining(a)increasesfuelef“ciency,butalso(b)increasesthecosttobuy.Inthisexample,oneshouldbeabletoseehowmpgandcostwilllikelybeunrelated,despitesharinganunderlyingcommonbasis:thisarisesbecausegreaterstreamliningbothdrivesupcostandmpg,whereasgreaternumberofcylindersdrivesupcostbutdrivesdownmpg.This“ndingiscommentedoningreaterdetailinSection Wenextexaminedthesigni“canceofenvironmentalin”uencesinthemodel.Allshared-environmentpathsexcepttoreligiosity Table1Phenotypiccorrelationsamongthestudymeasures.CommunityintegrationExistentialuncertaintyExistentialuncertaintyReligiosity.27.01.G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 couldberemovedwithoutsigni“cantlyworseningmodel“t (5)=2.03,=.85);however,shared-environmentseffectstoreligiosityweresigni“cant( (1)=15.76,.001)andsowereretained.Finally,weexaminedunique-environmenteffects.Removingthoseunique-environmenteffectsthatwerecommonbetweenthemeasuressigni“cantlyworsened“t( (3)=17.67,.001);however,thiseffectwasexclusivelydrivenbytheun-ique-environmenteffectscommontocommunityintegrationandexistentialuncertainty( (1)=17.08,.001)andsoonlythispathwasretained.The“nalmodelisdetailedinFig.2 7.Discussion The“rst“ndingofthepresentstudywastocon“rmamod-erateheritabilityforreligiosity:geneticfactorsaccountedfor26%ofthevariance.Thisiscomparable,thoughsomewhatsmal-lerthanvaluesreportedpreviously,whichrangedfrom30%to45%(Bouchard,2004).Additionally,religiositypossessedasignif-icantshared-environmentin”uence(perhapsre”ectingparentalcultureorotherfamily-leveleffects),a“ndingthatalsocon“rmspriorwork(Bouchard,2004).Existentialuncertaintyandcom-munityintegrationalsocontainedmoderategeneticin”uences,withapproximatelyathirdofthevarianceineachcaseac-countedforbyheritablefactors.Mostimportantly,andinlinewithourhypotheses,thegeneticeffectsonreligiositywereover-lappingwiththegeneticeffectsoncommunityintegrationandexistentialuncertainty. Communityintegrationshowedastrongrelationshipwithreli- giosity,withageneticcorrelationof.67(indicatingthat,oftheto-talgeneticin”uenceonthesetraits,almost45%ofthisin”uence(i.e.,.67 )issharedbybothtraits).Bycontrast,existentialuncer-taintyandreligiosityshowedamoremoderategeneticcorrelationof.33.Thisindicatesthat,atleastforthemeasureofreligiousbeliefusedinthecurrentstudy,socialneedsappeartobethepredomi-nantcorrelate,supportingthetheorizingofGrahamandHaidt These“ndingsdemonstratethatreligiosityisabiologicallycomplexconstruct,withdistinctheritablecomponents,whichinturnsuggeststhatattemptstolocatespeci“cgeneticvariants…whichhasbecomepossibleandpopularinrecentyearsamongso-cialpsychologists(Ebstein,Salomon,Chew,Zhong,&Knafo,2010…willneedtoaccountforthisheterogeneityatthegeniclevel.Spe-ci“cally,geneassociationstudiesseekingtodelineatethemolecu-larbasesofreligiositywouldbewisetoincludeintheiranalysesmeasuresre”ectingtheseunderlyingfoundationsofreligiosity;namely,communityintegrationandexistentialuncertainty.Thisapproachwill,however,needtoholdinmindthelikelymodestef-fectsizeforanygivengenevariantandthehighlypolygenicarchi-tectureofacomplextraitsuchasreligiosity(e.g.Munafò&Flint, Interestingly,anddeservingoffurthercomment,wasthesig-ni“cantgeneticoverlapbetweenexistentialuncertaintyandreligiositywhichsatincontrasttoanullphenotypiccorrelationforthesevariables.Ourmodelingallowsustounderstandhow Table2Univariatemodelingresults,acrosszygositytwinpaircorrelations,andphenotypiccorrelationsforcommunityintegration,existentialuncertainty,andreligiosity.ACEMZmMZfDZmDZfDZosCommunityintegration.31(.13….41).00(.00….13).69(.61….80).26.46.12.22.12Existentialuncertainty.36(.11….46).00(.00….19).64(.56….74).32.41.22.11.23Religiosity.26(.05….48).26(.07….43).48(.41….56).54.58.36.39.39:A=additivegeneticeffects;C=shared-environmenteffects;E=unique-environmenteffects;95%con“denceintervalsshowninbrackets;MZm=MZmalepairs;MZf=MZfemalepairs;DZm=DZmalepairs;DZf=DZfemalepairs;DZos=DZoppositesexpairs;CI=communityintegration;EU=existentialuncertainty. Fig.2.Finalmodelofadditivegenetic,shared-,andunique-environmenteffectsoncommunityintegration,existentialuncertainty,andreligiosity.:Circlesrepresentlatentadditivegenetic,shared-,andunique-environmentvariables.Pathvaluesarestandardizedpathcoef“cients(squaringthepathloadingindicatestheproportionofvarianceaccountedforinthemeasuredvariablebythelatentfactor). G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 thisarises:Thetwosourcesofgeneticin”uenceonreligiosityhaddistinctrelationshipstoexistentialuncertainty:Factora1onexistentialuncertaintybutreligiosity.Factora2loadedonboth.Thisobservationspeci“callyshedslightonwhyexistentialuncertaintyandreli-giosityarecorrelatedweakly(=.01):Onaverage,thesetworoughlyequalgeneticin”uencescanceloutleadingtoagreatlyreducedcorrelationatthephenotypiclevel.These“ndingscanalsobeunderstoodfromtheperspectiveofregressionmodeling:Forinstance,atthebehavioral,orphenotypic,levelitisnotuncommonto“ndthatagivenpredictorvariableisnotsignif-icantlyassociatedwithagivenoutcomevariable,althoughto“ndthebetacoef“cientbetweenthe“rstpredictorandtheoutcomevariabletobesigni“cantlydifferentfromzeroaftercontrollingfortheeffectofasecondpredictor.Theclassicaltwindesign,asemployedinthecurrentstudy,allowsustoexaminethesepatternsofeffectsatthelevelofvariancecom-ponents(i.e.A,C,andE)ratherthansolelyatthephenotypiclevel:buttheprincipleremainsfundamentallythesameasisthecaseinthestandardregressionframework.Assuch,herewe“ndthatthegeneticeffectsprimarilyaccountingforvaria-tioninexistentialuncertainty(Factora2)aresigni“cantlyandpositively(albeitmodestly)associatedwithreligiositywhencontrollingfortheheritableeffectsunderlyingcommunityinte-gration.These“nding,then,suggestthatpsychologicaltheoriesofreligionmustexpect,andaccountfor,multiple,possiblycountervailingcausesatthebiologicallevel. Additionally,inthefullCholeskymodel(seeAppendixA)weobservedadditionalcountervailingeffectssuchthatinadditiontogeneticeffectsactingpositivelyonbothexistentialuncer-taintyandreligiosity,wealsoobservedshared-environmentef-fectsonthesetwovariablesthatactedpositivelyonreligiositybutactedtodecreaseexistentialuncertainty:Thusenvironmen-talin”uencesmayalsoservetodecreasephenotypicassocia-tions.Thisobservationagainreiteratesthelikelycomplexitiesthatunderliehigher-orderpsychologicaltraitsandillustratesthevalueofthetwindesigninelucidatingthenatureofsuch Speculatively,thecurrent“ndingsareconsistentwithaposi-tionpositingthatreligionmaynotbethesoleorganiza-tionorsystemableto“llthenichecreatedbyhumanneedsforcommunityandexistentialmeaning.Thesuccession,dis-placement,andevolutionofreligionscanbeviewedinthislightastheshapingofreligioussystemsbytheiradherentstomaxi-mizetheextenttowhichtheirneedsaremet.Itmightbepre-dicted,then,thatundercertainconditions,evenasecularapparatus,shoulditbeabletomeetthecommunityandexisten-tialneedsofpeople,maybeattributedthesameimportanceinpeopleslivesasreligionoftenoccupies(e.g.Kay,Shepherdetal.,2010GrahamandHaidt(2010),however,suggestthisexchangeablegoodsnotionofreligionmayfailtoacknowledgethetight“tbetweenreligiousbeliefandhumanpsychology:religiouspracticesandritualsco-evolvedwithreligiouslyinclinedminds,sothattheynow“ttogetherextremelywellbecauseofourbiologicalandcultural-evolutionaryhistory,itshardtocomeupwithanythingthat“tsorsatis“esasmanypeopleasdoesreli-(p.147).Theextenttowhichreligionisawhollyexchangeablegoodthusremainsanopenquestion. Theprinciplelimitationofthepresentstudyisitsuseofasinglemeasureofreligiousbelieffocusedonreligiosity.Religiousbeliefisamultidimensionalconstruct(Hill&Hood,1999;Ken-dleretal.,2003).Assuch,whilethestrengthandimportanceofreligiousbeliefisamajorcomponentofreligiosity,otheras-pectsofreligiouslifemaynotsimilarlybeaccountedforbycommunityandexistentialfactors.Religiousfundamentalism Altemeyer&Hunsberger,2005)andspirituality(Zinnbauer&Pargament,2005),forinstance,showsigni“cant(opposite)asso-ciationswithopennesstoexperience(Saroglou,2010),withopennessinturnnotobviouslyre”ectingcommunityinterestsorexistentialconcerns(McCrae&John,1992).Afullaccountofreligiousfaith,then,willnodoubtbemorecomplexthanthatpresentedhere,likelyincludingadditionalfactorssuchasopen-nesstoexperience. Secondly,ourmeasuresofcommunityintegrationandexisten-tialuncertaintywerenecessarilybriefwithsubsequentrestrictionsinconstructrangeandreliability.Accordingly,futureworkisrec-ommendedtofurtherdelineatetheprecisenatureofthegeneticrelationsbetweenthesenon-theologicaltraitsandreligiositywithbroaderandmorepowerfulmeasures.However,itisworthhigh-lightingthattherelativelyshortscalesusedherewereneverthe-lessabletobemodeledaswhollyoverlappingwiththegeneticcomponentofreligiosity. Thirdly,whilewereportresultsherethatindicatecommunityintegrationandexistentialuncertaintywhollymediatedthege-neticin”uencesonreligiosity,thisisnottosaythatalternativeaccountsarenotplausible.Forexample,anunmeasuredvariablehighlylinkedtocommunityintegrationmaybetheultimatecauseofthisphenotypicandgeneticassociationwithreligiosity.Onequestionthispossibilityraisesiswhethertheseadditionalvariableswouldexplainadditionalheritablevariationinreligios-ity,orshowoverlapswiththegeneticeffectscommonwithcommunityintegrationandexistentialuncertainty.Futureworkisrecommendedtofurtherdelineatethegeneticcovariatesofreligiosity. Fourthly,thecurrentstudydesigndoesnotpermitcausalinferencestobemadefromwhatis,inessence,correlationaldata.Whilethecurrent“ndingsareconsistentwithamodelofreligiosityemergingfrommorebasicsocialandexistentialneeds,itwillbeofsomevaluetodeterminewhetherthisdirec-tionofcausalityiscorrectlypositioned.Longitudinalstudyde-signsthatmeasuresuchsocialandexistentialsentimentbeforereligiousviewshaveemerged,suchasinchildhood,mayshedlightonthisissue. Finally,itispossiblethatparticipantsrespondedtotheitemsconcerningcommunityintegrationwiththeirreligiouscommunityinmind.Ifso,thiswouldpotentiallyhavetheeffectofgeneratingadegreeofcontentoverlapbetweenourmeasuresofreligiosityandcommunityintegration(inspiteofoureffortstoavoidsuchasce-nario).Withthisinmind,futureresearchcouldexplicitlydistin-guishthereligiousandsecularcomponentsofcommunityorientation. Insummary,ourresultsindicatethatthegeneticin”uencesfrombothcommunityintegrationandexistentialuncertaintyaresharedwiththegeneticeffectsunderlyingameasureofreligiosityre”ectingthestrengthandimportanceofreligioninoneslife.Additionalfamilialin”uenceswereobservedforreligiosity,inlinewithpreviouswork.These“ndingsaresup-portiveoftheoriesregardingtheroleofreligionasasystemformeetingbasicsocialandexistentialneeds,andfortheadditionalroleofculturaltransmissioninshapingthestrengthofreligiousbeliefs.Futureworkshouldseektoexaminewhetherthecurrent“ndingsgeneralizetoadditionalaspectsofreligiousbelief,suchasreligiousfundamentalismandspiri-tuality,aswellastoestablishthecausaldirectionsofthese AppendixA. Thefull(unreduced)Choleskydecompositionmodelwithstan-dardizedparameterestimates. G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 403 :Shared-environmenteffectsC2andC3wereestimatedatzeroandsoareomittedhereforclarityofillustration. References Altemeyer,B.,&Hunsberger,B.(2005).Fundamentalismandauthoritarianism.InR.F.Paloutzian&C.L.Park(Eds.),Handbookofthepsychologyofreligionand(pp.378…393).NewYork:Guilford Batson,C.D.,Schoenrade,P.,&Ventis,L.W.(1993).Religionandtheindividual:Asocial…psychologicalperspective.OxfordUniversityPress Birgegard,A.,&Granqvist,P.(2004).ThecorrespondencebetweenattachmenttoparentsandGod:Threeexperimentsusingsubliminalseparationcues.PersonalityandSocialPsychologyBulletin,30,1122…1135 Boker,S.,Neale,M.,Maes,H.,Wilde,M.,Spiegel,M.,Brick,T.,etal.(2010a).OpenMx:Anopensourceextendedstructuralequationmodelingframework.Psychometrika,76,306…317. Boker,S.,Neale,M.,Maes,H.,Wilde,M.,Spiegel,M.,Brick,T.,etal.(2010b).Multipurposesoftwareforstatisticalmodeling(versionRpackageversion0.3.1- Bouchard,T.J.Jr.,(2004).Geneticin”uenceonhumanpsychologicaltraits:ACurrentDirectionsinPsychologicalScience,13,148…151 Bouchard,T.J.,Jr.,McGue,M.,Lykken,D.,&Tellegen,A.(1999).Intrinsicandextrinsicreligiousness:Geneticandenvironmentalin”uencesandpersonalityTwinResearch,2,88…98 Bradshaw,M.,&Ellison,C.G.(2008).Dogeneticfactorsin”uencereligiouslife?Findingsfromabehaviorgeneticanalysisoftwinsiblings.JournalfortheScienti“cStudyofReligion,47,529…544 Brim,O.G.,Ryff,C.D.,&Kessler,R.C.(2004).Howhealthyarewe?:A25nationalstudyofwell-beingatmidlife.Chicago:UniversityofChicagoPress Bulbulia,J.(2004).Thecognitiveandevolutionarypsychologyofreligion.andPhilosophy,19,655…686 Cavendish,J.C.,Welch,M.R.,&Leege,D.C.(1998).SocialnetworktheoryandpredictorsofreligiosityforBlackandWhiteCatholics:EvidenceofaBlacksacredcosmos.JournalfortheScienti“cStudyofReligion,37,397…410 Dennett,D.C.(2006).Breakingthespell:Religionasanaturalphenomenon.London,England:PenguinBooks DOnofrio,B.M.,Eaves,L.J.,Murrelle,L.,Maes,H.H.,&Spilka,B.(1999).Understandingbiologicalandsocialin”uencesonreligiousaf“liation,attitudes,andbehaviors:Abehaviorgeneticperspective.JournalofPersonality,67,953…984 Dudley,R.L.,&Dudley,M.G.(1986).Transmissionofreligiousvaluesfromparentstoadolescents.ReviewofReligiousResearch,28,3…15 Durkheim,E.(1915).Theelementaryformsofthereligiouslife.NewYork:FreePress Ebstein,R.P.,Salomon,I.,Chew,S.H.,Zhong,S.,&Knafo,A.(2010).Geneticsofhumansocialbehavior.Neuron,65,831…844 Eid,M.,Riemann,R.,Angleitner,A.,&Borkenau,P.(2003).Sociabilityandpositiveemotionality:Geneticandenvironmentalcontributionstothecovariationbetweendifferentfacetsofextraversion.JournalofPersonality,71,319…346 Gebauer,J.E.,&Maio,G.R.(2012).TheneedtobelongcanmotivatebeliefinGod.JournalofPersonality,80,465…501 Graham,J.,&Haidt,J.(2010).Beyondbeliefs:ReligionsbindindividualsintomoralPersonalityandSocialPsychologyReview,14,140…150 Greenberg,J.,Pyszczynski,T.,&Solomon,S.(1986).Thecausesandconsequencesofaneedforself-esteem:Aterrormanagementtheory.InR.F.Baumeister(Ed.),Publicselfandprivateself(pp.189…212).NewYork:Springer-Verlag Hartman,W.J.(1976).Membershiptrends:AstudyofdeclineandgrowthintheUnitedMethodistChurch1949…1975.Nashville:DiscipleshipResources Hill,P.C.,&Hood,R.W.(1999).Measuresofreligiosity.Birmingham,Alabama:ReligiousEducationPress Inzlicht,M.,McGregor,I.,Hirsh,J.B.,&Nash,K.(2009).NeuralmarkersofreligiousPsychologicalScience,20,385…392 Jonas,E.,&Fischer,P.(2006).Terrormanagementandreligion:EvidencethatintrinsicreligiousnessmitigatesworldviewdefensefollowingmortalityJournalofPersonalityandSocialPsychology,91,553…567 Kay,A.C.,Shepherd,S.,Blatz,C.W.,Chua,S.N.,&Galinsky,A.D.(2010).ForGod(or)country:Thehydraulicrelationbetweengovernmentinstabilityandbeliefinreligioussourcesofcontrol.JournalofPersonalityandSocialPsychology,99 Kendler,K.S.,Liu,X.,Gardner,C.O.,McCullough,M.E.,Larson,D.,&Prescott,C.A.(2003).Dimensionsofreligiosityandtheirrelationshiptolifetimepsychiatricandsubstanceusedisorders.ArchivesofGeneralPsychiatry,160 Koenig,L.B.,McGue,M.,Krueger,R.F.,&Bouchard,T.J.Jr.,(2005).Geneticandenvironmentalin”uencesonreligiousness:Findingsforretrospectiveandcurrentreligiousnessratings.JournalofPersonality,73,471…488 Lewis,G.J.,Ritchie,S.J.,&Bates,T.C.(2011).Therelationshipbetweenintelligenceandmultipledomainsofreligiousbelief:EvidencefromalargeadultUSsample.Intelligence,39,468…472 Mackintosh,M.A.,Gatz,M.,Wetherell,J.L.,&Pedersen,N.L.(2006).AtwinstudyoflifetimeGeneralizedAnxietyDisorder(GAD)inolderadults:Geneticandenvironmentalin”uencessharedbyneuroticismandGAD.TwinResearchandHumanGenetics,9,30…37 Martin,N.G.,Eaves,L.J.,Heath,A.C.,Jardine,R.,Feingold,L.M.,&Eysenck,H.J.(1986).Transmissionofsocialattitudes.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,83,4364…4368 McCrae,R.R.,&John,O.P.(1992).Anintroductiontothe“ve-factormodelanditsJournalofPersonality,60,175…215 McGue,M.,&Bouchard,T.J.(1984).Adjustmentoftwindatafortheeffectsofageandsex.BehaviorGenetics,14,325…343 G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 Munafò,M.R.,&Flint,J.(2011).DissectingthegeneticarchitectureofhumanTrendsinCognitiveScience,15,395…400 Neale,M.C.(2003).Twinsstudies:Softwareandalgorithms.InD.N.Cooper&N.J.Hoboken(Eds.),Encyclopediaofthehumangenome(pp.88…96).London:Nature Neale,M.,&Cardon,L.(1992).MethodologyforgeneticstudiesoftwinsandfamiliesDordrecht,TheNetherlands:KluwerAcademicPublishers Norenzayan,A.,&Gervais,W.M.(2013).Theoriginsofreligiousdisbelief.TrendsinCognitiveSciences,17,20…25 Norenzayan,A.,&Hansen,I.G.(2006).BeliefinsupernaturalagentsinthefaceofPersonalityandSocialPsychologyBulletin,32,174…187 Norenzayan,A.,&Shariff,A.F.(2008).TheoriginandevolutionofreligiousScience,322,58…62 Peterson,J.B.(1999).Mapsofmeaning:Thearchitectureofbelief.NewYork: Purcell,S.(2008).Statisticalmethodsinbehavioralgenetics.InR.Plomin,J.C.DeFries,G.E.McClearn,&P.McGuf“n(Eds.),Behavioralgenetics(Appendix)WorthPublishers RDevelopmentCoreTeam(2009).R:Alanguageandenvironmentforstatistical.RFoundationforStatisticalComputing.Vienna,Austria. Rutjens,B.T.,vanderPligt,J.,&vanHarreveld,F.(2010).DeusorDarwin:Randomnessandbeliefintheoriesabouttheoriginoflife.JournalofExperimentalSocialPsychology,46,1078…1080 Saroglou,V.(2010).Religiousnessasaculturaladaptationofbasictraits:AFive-FactorModelperspective.PersonalityandSocialPsychologyReview,14,108…125 Swatos,W.H.(1998).Encyclopediaofreligionandsociety.WalnutCreek,CA:AltamiraPress Vail,K.E.,III,Rothschild,Z.K.,Weise,D.,Solomon,S.,Pyszczynski,T.,&Greenberg,J.(2010).Terrormanagementtheoryandreligiosity.PersonalityandSocialPsychologyReview,14,84…94 Vance,T.,Maes,H.H.,&Kendler,K.S.(2010).Geneticandenvironmentalin”uencesonmultipledimensionsofreligiosity:Atwinstudy.JournalofNervousandMentalDisease,198,755…761 Waller,N.G.,Kojetin,B.A.,Bouchard,T.J.,Jr.,Lykken,D.T.,&Tellegen,A.(1990).Geneticandenvironmentalin”uencesonreligiousinterests,attitudes,andvalues:Astudyoftwinsrearedapartandtogether.PsychologicalScience,1 Zinnbauer,B.J.,&Pargament,K.I.(2005).Religiousnessandspirituality.InR.F.Paloutzian&C.L.Park(Eds.),Handbookofthepsychologyofreligionand(pp.21…42).NewYork:Guilford G.J.Lewis,T.C.Bates/JournalofResearchinPersonality47(2013)398…405 405