Altruism predicts mating success in humans

20 Pages·2017·0.31 MB·English

416
BritishJournalofPsychology(2017),108,416–435
©2016TheBritishPsychologicalSociety
www.wileyonlinelibrary.com
Altruism predicts mating success in humans
Steven Arnocky1*, Tina Pich(cid:1)e1, Graham Albert1,
Danielle Ouellette1 and Pat Barclay2
1DepartmentofPsychology,NipissingUniversity,NorthBay,Ontario,Canada
2DepartmentofPsychology,UniversityofGuelph,Canada
Inorderfornon-kinaltruismtoevolve,altruistsmustreceivefitnessbenefitsfortheir
actions that outweigh thecosts.Several researchers havesuggested that altruism isa
costlysignalofdesirablequalities,suchthatitcouldhaveevolvedbysexualselection.In
two studies, we show that altruism is broadly linked with mating success. In Study 1,
participantswhoscoredhigheronaself-reportaltruismmeasurereportedtheywere
moredesirabletotheoppositesex,aswellasreportedhavingmoresexpartners,more
casualsexpartners,andhavingsexmoreoftenwithinrelationships.Sexmoderatedsome
oftheserelationships,suchthataltruismmatteredmoreformen’snumberoflifetimeand
casual sex partners. In Study 2, participants who were willing to donate potential
monetarywinnings(inamodifieddictatordilemma)reportedhavingmorelifetimesex
partners,morecasualsexpartners,andmoresexpartnersoverthepastyear.Menwho
werewillingtodonatealsoreportedhavingmorelifetimedatingpartners.Furthermore,
thesepatternspersisted,evenwhencontrollingfornarcissism,BigFivepersonalitytraits,
andsociallydesirableresponding.Theseresultssuggestthataltruistshavehighermating
successthannon-altruistsandsupportthehypothesisthataltruismisasexuallyselected
costlysignalofdifficult-to-observequalities.
Altruistic behaviours1 have been observed across diverse vertebrate species of birds
(Krams, Krama, Igaune, & M€and, 2008), fish (Bshary, 2002),amphibians(To(cid:1)th,Hoy, &
Hettyey,2011),reptiles(Sinervoet al.,2006),andmammals(Wilkinson,1988),including
modern humans (Trivers, 1971). Early research on altruistic behaviour among non-kin
centred upon Trivers’ (1971) theory of reciprocity – the notion that individuals may
behave selflessly to receive some benefit in the future. However, researchers have
recently suggested that reciprocity alone cannot sufficiently account for all altruistic
behaviours among unrelated individuals (Fehr & Fischbacher, 2003; Sober & Wilson,
1998;reviewedbyBarclay&VanVugt,2015).Ithasbeenfurtherarguedthatreciprocity
mightonlyevolveundercircumscribedenvironmentalconditions(Roberts,1998).This
haspromptedresearcherstoconsideradditionalfactors,whichmayhavecontributedto
theevolutionofaltruisticbehaviour.
Onehypothesis,whichhasgainedrecentempiricalsupportfromstudiesonhuman
behaviour,isthepossibilitythataltruismfunctionsasacostlysignalofqualitiesthatare
*CorrespondenceshouldbeaddressedtoStevenArnocky,DepartmentofPsychology,NipissingUniversity,100CollegeDrive,
NorthBay,ON,CanadaP1B8L7(email:[email protected]).
1By‘altruism’,wemeanactsthatdecreasesomefitness-proxyintheactor(atleasttemporarily)andfunctiontoconferbenefits
onarecipient,regardlessoftheunderlyingpsychologicalmotivations.Thosewhodefine‘altruism’differentlyshouldsubstitute
theirpreferredtermforsuchphenomena.
DOI:10.1111/bjop.12208
Altruismandmatingsuccess 417
otherwise difficult to observe (Gintis, Smith, & Bowles, 2001; Zahavi, 1975). For
example,huntingandsharinglargegame–especiallywithoutexpectationofreciprocity
–demonstrateone’sskillsasagoodhunterandthusone’svalueasanallyormate(Jaeggi
& Gurven, 2013; Smith & Bliege Bird, 2000). In particular, it can be valuable to signal
desirablequalitiestopotentialmates.Insupportofthis,evidencesuggeststhataltruists
maybeviewedbyothersasmoreattractivematingpartnersrelativetothosewhoareless
altruistic (Barclay, 2010; Farrelly, Lazarus, & Roberts, 2007; Moore et al., 2013; Oda,
Okuda,Takeda,&Hiraishi,2014;Phillips,Barnard,Ferguson,&Reader,2008).Yetsuch
studies on preferences for altruistic mates fall short of determining whether altruism
does indeed confer a mating advantage to the altruist. Given that mate preferences are
notalwaysgoodindicatorsofactualmatechoice(Todd,Penke,Fasolo,&Lenton,2007),
thegoalofthisstudywastoexplorewhetherself-reportedaltruisticbehaviourpredicted
individuals’ mating behaviour in terms of their self-reported mating success, their
lifetime number of sex partners, their lifetime number of casual sex partners, and the
frequency with which they copulate with their current romantic partner and whether
these relationships vary as a functionof sex.
Costlysignalling
Acostlysignalisanymorphologicalorbehaviouraltraitwhichadvertisesanorganism’s
qualityatsomeexpense(Zahavi,1975;Zahavi&Zahavi,2007).Providedthatthesignal
is sufficiently costly for those who do not actually possess the quality (Getty, 2006;
Sz(cid:1)amado(cid:1),2011),bearingtheburdenofthatfeaturecanserveasanhonestreflectionof
thesignaller’squalityorcooperativeintent(Barclay&Reeve,2012;Gintiset al.,2001).
By attending to such signals, observers gain useful information about the qualities of
the signaller(s).
Someresearchershaveconceptualizedaltruismwithinthisframeworkasacourtship
display which can advertise one’s mate-value characteristics including genetic endow-
ment,health,resourcecontrol,orvigour(Miller,2000,2007;Smith&BliegeBird,2000;
Tessman,1995).Totheextentthataltruismisdesiredbypotentialmatingpartners,itis
widelypresumedthatanassociatedincreaseinmatingsuccesswouldservetorecoupthe
initialcostofthealtruisticactsordisposition(Hardy&VanVugt,2006).Insupportofthe
costly signalling hypothesis, McAndrew and Perilloux (2012) showed that individuals
indeedviewothers’altruisticbehaviourasbeingcostly,challenging,andimportant.The
potentialutilityofaltruismasacostlysignalofphenotypicqualityorcooperativeintent
mayleadindividualstoattempttooutcompeterivalswithinthisdomain–aphenomena
termed ‘competitive altruism’ (Barclay, 2011, 2013; Roberts, 1998). Barclay and Willer
(2007) found that when observers could choose whom to associate with, participants
competedtobemoregenerousthanothers.Giventheparticularimportanceofsuccessful
intrasexual competition towards augmenting male reproductive potential (Bateman,
1948),suchcompetitivealtruismmaybemoreprevalentamongmenrelativetowomen.
Multiple studies have shown that men are more generous when observed by women,
whereasthereverseisoftennottrue(Bo€hm&Regner,2013;Tognetti,Berticat,Raymond,
&Faurie, 2012;Van Vugt&Iredale,2013).McAndrewandPerilloux(2012)found that
groups containing multiple males evidenced more competition to become an altruist
relativetogroupscontainingmultiplefemales(Bo€hm&Regner,2013;Hardy&VanVugt,
2006;Iredale,VanVugt,&Dunbar,2008).Similarly,RaihaniandSmith(2015)foundthat
men make more online charitable donations to attractive women than to unattractive
womenandwillcompetewithothermenoverdonationstowomen.
418 StevenArnockyet al.
Dowepreferaltruisticmates?
Empiricaltestsofaltruismasastrategyforaccessingmatingopportunitieshavehitherto
focussedonindividuals’penchantsforaltruisticmates.Forexample,Phillipset al.(2008)
assessedmen’sandwomen’sstatedpreferencesforaltruistictraitsinalong-termmating
partner(boyfriend,girlfriend,husband,orwife).Resultsshowedthatwomenexhibiteda
strongerpreferenceforaltruistictraitsinaprospectiveromanticpartnerrelativetomen,
controllingforrelationshiplengthandsocialdesirabilityscores.Farrelly(2011)andOda
et al.(2014)havealsodemonstratedthatwomenpreferaltruisticbehavioursinpotential
long-termmates;apreferencethatwasnotaffectedbywomen’sovulatorycycle.Barclay
(2010)askedyoungmenandwomentoreadfictitiousdatingadvertisementsofopposite
sexindividualsforshort-andlong-termrelationships,whichvariedintheexpressionof
altruistic tendencies. Participants rated the ads on their interest in pursuing various
relationshipswiththatindividual,aswellasonthetarget’sattractivenessandpersonality
characteristics. Barclay (2010) found that both men and women rated individuals with
altruistic traits as desirable for a long-term relationship, whereas women but not men
ratedaltruisticopposite-sexindividualsasdesirableforsingledates(thiscouldbeshort-
term mating, but it is important to note that desirability for one-night stands was
unaffected).Takentogether, the results ofpreviousresearch suggestthat altruismmay
benefitbothmen’sandwomen’sdesirability,especiallyasalong-termmate.Itiscurrently
ambiguouswhetheraltruismaffectsone’sdesirabilityasashort-termpartner,withsome
studies saying no (Barclay, 2010; Oda et al., 2014), and other studies suggesting that
altruistic men are more desirable short-term partners than non-altruistic men (Farrelly,
Clemson,&Guthrie,2016;Mooreet al.,2013).
Interestingly, both altruistic personality and individuals’ preferences for altruistic
matesappeartobeatleastpartlyheritable,asevidencedbygreaterconcordanceofthese
factorsamongmonozygotictwinscomparedtodyzygotictwins,withgenesaccounting
for much of the variance in the relationship (Phillips, Ferguson, & Rijsdijk, 2010).
Accordingly,Phillipset al.(2010)predictedthatinancestralpopulations,altruistswould
havematedfrequentlywithindividualswhoexhibitedapreferenceforhavinganaltruistic
partner. Studies of individuals’ stated preferences for altruistic mates are nevertheless
limited in the conclusions that can be drawn from them regarding the evolution of
altruismasamatingstrategy.Giventhatmatepreferencesareoneamongahostoffactors,
including the availability of mates and resources and the presence of competitors that
determineactualmatingdecisionsandbehaviours(Arnocky,Ribout,Mirza,&Knack,
2014; Arnocky, Sunderani, & Vaillancourt, 2013; Todd et al., 2007), it remains unclear
whetherornotaltruisticbehaviourisindeedcorrelatedwithenhancedmatingsuccess.
Thisisanimportantdistinction,giventhatitisactualmatingoutcomeswhichultimately
contributetotheevolutionofparticularphenotypes(Zohar&Guttman,1989).Therefore,
to the extent that mating with an altruist confers some reproductive advantage, or has
done so over the course of human evolutionary history, we would expect altruists to
exhibitincreasedmatingsuccessrelativetotheirlessaltruisticcounterparts.
Indirect evidence (typically from hunter-gatherer populations) lends circumstantial
support to this hypothesis. It has been found that successful male hunters (who often
share the meat they have obtained) receive greater reproductive access to females
(Hawkes,1991;Hawkes,O’Connell,Hill,&Charnov,1985;Hill&Hurtado,1996;Hill&
Kaplan,1988).Thisevidenceseemsconsistentwithsexualselectionfavouringaltruists
(Gintiset al.,2001;Gurven&vonRueden,2006).Tognetti,Berticat,Faurie,andRaymond
(2014) found that rural Senegalese marry assortatively on cooperativeness, such that
cooperative men tend to marry cooperative women. However, such assortment alone
Altruismandmatingsuccess 419
could reflect either cooperators being highly desirable and choosing each other or
cooperatorsbeingundesirableandmarryingeachotherasa‘lastresort’(e.g.,cooperation
as compensation for low mate value, Barclay, 2013, 2016). It is therefore essential to
expandupontheseinitialfindingsbyexplicitlymeasuringaltruisminrelationtomating
success,whichcanbedoneamongyoungadultsinacontemporaryindustrializedculture.
Acrosstwostudies,wetestedthehypothesisofaltruismasacostlysignalbyexamining
whether self-reported altruism (Study 1) and observable altruism in terms of monetary
donation(Study2)predictedvariousmatingsuccessindicesinsamplesofundergraduate
menandwomen.Rhodes,Simmons,andPeters(2005)suggestmatingsuccessintheway
of sexual behaviour is a sufficient index of reproductive success in ancestral times;
however, modern-day contraception attenuates this link between mating success
translatingintoreproductivesuccess.
STUDY1
We hypothesized that self-reported altruism would predict greater self-reported
desirability as a mate, having had more lifetime sex partners, more lifetime casual sex
partners,andamongthosecurrentlyinacommittedromanticrelationship,morefrequent
copulation with one’s partner within the past 30 days. Moreover, given that some
researchhassuggestedthataltruismmaybemoreimportantasacostlysignalamongmen
(especiallyforshort-termmatings,seeBarclay,2010;Bo€hm&Regner,2013;Farrellyet al.,
2016;Mooreet al.,2013;Raihani&Smith,2015;VanVugt&Iredale,2013),weexplored
interactioneffectsofsexasapotentialmoderatortotheserelationships.Thesehypotheses
weretestedinconsiderationofapotentialconfoundingaspectofpersonality.Specifically,
itispossiblethatanyrelationshipsbetweenaltruismandmatingsuccessmaybeanartefact
ofBigFivepersonalitydimensions,whichhavebeenshowncross-culturallytocorrelate
with altruism (Johnson et al., 1989). For example, extraversion has been consistently
linkedtoincreasedsexualbehaviouracrossmany(butnotall)worldregions(Schenk&
Pfrang, 1986; Schmitt, 2004). Participant age was also controlled for given previous
associationswithnumberofsexpartnersinuniversitysamples(Arnockyet al.,2013).
Method
Participants
Our sample consisted of 192 unmarried women and 105 unmarried men living in
NorthernOntario,Canada,betweentheagesof16and33(M = 21.21,SD = 3.48).The
age
ethniccompositionofthesamplewasasfollows:Caucasian(90%),Aboriginal(4%),Black
(3%),Asian(2%),andSouthAsian(1%).Participantswererecruitedincommonareasofa
smalluniversityandcollege.Asapartofalargerstudyonhumanmatingbehaviour,each
participant completed a paper-and-pencil-based questionnaire containing self-report
measures of altruistic behaviour, personality, mating success, and sexual history.
Participantswerecompensatedwithanopportunitytowinoneofthree$100draws.
Materialsandprocedure
Altruism
The 20-item Self-report Altruism Scale was used to measure dispositional altruism
(Rushton, Chrisjohn, & Fekken, 1981). This is a widely used measure, which was
420 StevenArnockyet al.
developedinthesameProvinceofCanadainwhichthecurrentresearchtookplace.The
measure consists of items that assess the frequency of altruistic acts towards others,
wherebydirectreciprocitywouldbeunlikelyorunexpected.Exampleitemsincludethe
following:‘Ihavegivenmoneytoacharity’,‘Ihavehelpedpushastranger’scaroutofthe
snow’,‘Ihavegivendirectionstoastranger’,‘Ihavedonatedblood’,‘Ihaveofferedtohelp
ahandicappedorelderlystrangeracrossastreet’,and‘IhavehelpedaclassmatewhoIdid
notknowthatwellwithanassignmentwhenmyknowledgewasgreaterthanhisorhers’.
Responseoptionsrangingalonga5-pointLikert-typescale(0 = Never,4 = VeryOften).
The measure showed good internal consistency in both men (a = .86) and women
(a = .87).Anexploratoryfactoranalysiswasconductedtodeterminethefitoftheitemsto
one fixed factor: all items loaded well on a single factor, and there were no major
differencesintherelationshipbetweenaltruismandmatingsuccessindicesatthesingle-
itemlevel(seeSupportinginformation).
Self-reportedmatingsuccess
Self-reporteddesirabilitytotheoppositesexwasassessedusingtheself-perceivedmating
success scale (SPMS; Landolt, Lalumi(cid:3)ere, & Quinsey, 1995). The measure consists of 8
items with response options ranging along a 7-point Likert-type scale (1 = disagree,
7 = agree).Theitemsareasfollows:‘MembersoftheoppositesexthatIlike,tendtolike
meback’,‘Membersoftheoppositesexnoticeme’,‘Ireceivemanycomplimentsfrom
membersoftheoppositesex’,‘Membersoftheoppositesexarenotveryattractedtome’
(reverse scored), ‘I receive sexual invitations from members of the opposite sex’,
‘Membersoftheoppositesexareattractedtome’,‘IcanhaveasmanysexualpartnersasI
choose’, and ‘I do not receive many compliments from members of the opposite sex’
(reversescored).Themeasureshowedgoodinternalconsistency(a = .85).
Matingsuccessindices(sexualhistoryvariables)
Inassessingparticipants’pastsexualbehaviours,respondentsself-reportedon(1)their
numberofconsensuallifetimesexpartnersand(2)theirnumberofconsensualcasualsex
partnerswherecasualsexwasdefinedas‘asexualpartnerwho[participants]wereNOT
inanexclusive,committedrelationshipwith’(Arnockyet al.,2013).Finally,participants
whoself-reportedasbeingcurrentlyinaromanticrelationship(foratleast1 month)and
whoreportedhavinghadsexwiththeircurrentpartneratleastonceindicatedthenumber
oftimestheyhadengagedinconsensualsexualintercourseoverthepast30 dayswiththat
partner,asameasureofsexualaccesswithintheromanticdyad.Ofthetotalsample,111
participants(37.4%)metthecriteriaforcurrentlybeinginaromanticrelationshipwitha
personwhomtheyhadpreviouslycopulatedwith(47men,64women).
Personalitydimensions
BigFivepersonalitydimensionsweremeasuredusingtheTen-ItemPersonalityInventory
(TIPI; Gosling, Rentfrow, & Swann, 2003). Participants read the following set of
instructions:‘Hereareanumberofpersonalitytraitsthatmayormaynotapplytoyou.
Pleasewriteanumbernexttoeachstatementtoindicatetheextenttowhichyouagreeor
disagreewiththatstatement.Youshouldratetheextenttowhichthepairoftraitsapplies
toyou,evenifonecharacteristicappliesmorestronglythantheother’.Participantsthen
reported on the following ten items using a 7-point Likert-type scale (1 = disagree
Altruismandmatingsuccess 421
strongly, 7 = agree strongly). Example items include the following: ‘Extraverted,
enthusiastic’, and ‘Reserved, quiet’ (reverse scored) comprising the extraversion
subscale. The TIPI has shown to converge with self, observer, and peer ratings of
personality, as well as good construct validity and test–retest reliability (Gosling et al.,
2003; Jonason, Teicher, &Schmitt, 2011).In this study,the subscalesall showed good
intercorrelation,r = .22,p < .001–.52,p < .001.Alloftheanalysesofaltruismproduce
qualitativelysimilarresultswhetherageandpersonalityareincludedorexcludedinthe
regressionmodels(Table 1).
Statisticalanalysis
MultipleordinaryleastsquaresregressionanalyseswereconductedusingthePROCESS
macro for SPSS (Hayes, 2013). Variables were mean centred (for continuous variables:
altruism,age,extraversion,lifetimesexpartners,lifetimecasualsexpartners,andmating
success) and dummy coded (for dichotomous variable sex, M = (cid:1)1, F = 1). Altruistic
behaviour,sex(asthemoderatingvariable)andtheirinteractionwerecalculatedwithage
andBigFivepersonalitytraitsenteredascovariatesforeachofthefourmatingsuccess
outcomes:self-reportedmatingsuccess,lifetimenumberofsexpartners,lifetimenumber
of casual sex partners, and (for romantically involved participants) number of partner
copulations in the past 30 days. Following the rationale outlined by Hayes (n.d.), we
present unstandardized regression coefficients and standardized t-scores for each
relationship(correlationsbetweenaltruismandmatingsuccessindicesarepresentedin
theSupporting information).Wepresenttheeffects ofaltruismbothwithandwithout
covariates(ageandpersonalityfactors).
ResultsandDiscussion
Wefirstexaminedtherelationshipbetweenaltruism,sex,andparticipants’self-reported
mating success, controlling for age and TIPI scores (Table 1). Results showed that
altruistic behaviour significantly predicted self-reported mating success, b = 0.52,
p = .0002.Sexwasnotrelatedtoself-reportedmatingsuccess,b = (cid:1)0.04,p = .61.The
altruism 9 sex interaction was non-significant, suggesting that altruism predicts self-
reportedmatingsuccess regardless ofsex,b = 0.19,p = .16.Themodelaccountedfor
19%(R2 = .19)ofexplainedvarianceinindividuals’self-reportedmatingsuccess.
Wenextexaminedtherelationshipbetweenaltruism,sex,andlifetimenumberofsex
partners (Table 1). Results revealed that altruistic behaviour significantly predicted
lifetimenumberofsexpartners,b = 2.60,p = .007(Figure 1).Sexalsopredictedlifetime
number of sex partners, b = (cid:1)1.50, p = .02, suggesting that men report having more
lifetime sex partners than women. Moreover, the altruism 9 sex interaction was also
significant,b = (cid:1)3.01,p = .002,suggestingthataltruismmattersmoreformen’saccess
tosexpartnersrelativetowomen.Simpleslopesanalysisindeedindicatedthataltruism
predictedlifetimenumberofsexpartnersformen,b = 6.34,b = .40,p = .0001,butnot
for women, b = 0.76, b = .02, p = .57. The model accounted for 26% (R2 = .26) of
explainedvarianceinindividuals’totalnumberoflifetimesexpartners.
Inexaminingindividuals’lifetimenumberofcasualsexpartners,itwassimilarlyfound
that altruistic behaviour significantly predicted total number of casual sex partners,
b = 2.16,p = .009(Table 1).Sexpredicted numberofcasualsexpartners,b = (cid:1)1.38,
p = .009,suggestingthatmenreportmorecasualsexpartnersthandidwomen.There
422 StevenArnockyet al.
dices, ncy SE 1.690.921.660.230.600.770.650.850.77 1.370.740.12
n e
i u
s q
atingsucces pulationfre b .27.02.21.03.08(cid:1).19.15.03(cid:1).03 .30(cid:1).02.15
m o
c
orof pair b †8312050741†36822217 26**1824
dict In- 3.0.3.0.0.(cid:1)1.0.0.(cid:1)0. 4.(cid:1)0.2.
e
r
p
a
s 334334756 651
a s E 858134344 748
m er S 0.0.0.0.0.0.0.0.0. 0.0.0.
s n
estingaltrui ualsexpart b .16(cid:1).18(cid:1).22.21.21(cid:1).15(cid:1).07.01.01 .20(cid:1).19.23
t s
s a
el c
d e ** * *
formo Lifetim b 2.16**(cid:1)1.38**(cid:1)3.01**0.45**1.04**(cid:1)1.05*(cid:1)0.360.090.04 2.66**(cid:1)1.46**(cid:1)3.05**
s
r
o
r
r
e 728581334 558
d E 969135455 959
dar ers S 0.0.0.0.0.0.0.0.0. 0.0.0.
n n
andstaures) expart b .16(cid:1).17(cid:1).19.30.21(cid:1).13(cid:1).09.04(cid:1).01 .21(cid:1).18(cid:1).20
ncoefficientsonalitymeas Lifetimes b 2.60**(cid:1)1.50*(cid:1)3.01**0.75***1.30**(cid:1)1.10*(cid:1)0.570.35(cid:1)0.07 3.27***(cid:1)1.61**(cid:1)3.18**
sioers
sp
Table1.Study1:Unstandardizedandstandardizedregrebothwithandwithoutcontrolvariables(ageandBigFive Self-reportmatingsuccess bbSE ModelwithcontrolvariablesAltruism0.52***.250.13=(cid:1)=(cid:1)(cid:1)Sex(M1,F1)0.04.040.099AltruismSex0.19.090.14Age0.09.030.02Extraversion0.28***.340.05Conscientiousness0.06.050.07Emotionalstability0.03.030.06(cid:1)(cid:1)Openness0.07.070.07(cid:1)(cid:1)Agreeableness0.05.050.07ModelwithoutcontrolvariablesAltruism0.55***.260.13(cid:1)(cid:1)Sex0.06.050.079AltruismSex0.18.080.13 †<*<**<***<Note.p.10;p.05;p.01;p.005.
Altruismandmatingsuccess 423
Figure 1. Self-reportedaltruisminStudy1andreportedmatingsuccessofmales(dashedlines)and
females(solidlines),asmeasuredby(a)scoresonascaleofself-reportedmatingsuccess,(b)numberof
lifetimesexpartners,(c)numberofcasualsexpartners,and(d)numberofin-paircopulationsinthelast
month.Participantsexmoderatestherelationshipbetweenaltruismandlifetimenumberofsexpartners
andnumberofpast-yearsexpartners,butnotself-reportedmatingsuccessorlifetimecasualsex.Note.
Fordisplaypurposes,lowaltruism = 1 SDbelowthemeanandhighaltruism = 1 SDabovethemean.
[Colourfigurecanbeviewedatwileyonlinelibrary.com]
was a significant altruism 9 sex interaction, b = (cid:1)3.01, p = .0004, suggesting that
altruismrelated morestronglytonumberofcasualsexpartnersamongmenrelativeto
women (Figure 1). Simple slopes analysis indeed indicated that altruism predicted
lifetimenumberofsexpartnersformen,b = 5.88,b = .45,p = .0001,butnotforwomen,
b = (cid:1)0.14, b = (cid:1).01, p = .89. The model accounted for 23% (R2 = .23) of explained
varianceinindividuals’numberofcasualsexpartners.
We next examined the relationship between altruism and number of times
participantshadsexwiththeir partnerswithinthepast 30 daysamong asubset ofour
sample who had a dating partner at the time of data collection (n = 111 participants,
Table 1). Results showed that altruistic behaviour was modestly related to number of
copulations in the past month, b = 3.83, p = .08. Sex did not predict number of
copulations, b = 0.12, p = .90. The altruism 9 sex interaction was not significant,
b = 3.05, p = .13. The model accounted for 13% (R2 = .13) of explained variance in
individuals’in-paircopulations.Withoutthecontrolvariablesinthemodel,theeffectof
altruismincreasedinsignificance,b = 4.26,p = .01.
STUDY2
Study1showedalinkbetweenaltruismandmatingsuccessvariables.However,itwas
limitedintheuseofaself-reportmeasureofaltruism.ForStudy2,weutilizedamodified
424 StevenArnockyet al.
dictator dilemma as an observable measure of altruism, in which participants were
enteredintoadrawfor$100andwereaskedtoindicatewhethertheywanted tokeep
theirwinnings,ortohavetheresearchersdonateaspecificportionoftheirwinningson
theirbehalftoaregisteredcharityoftheirchoosing.
Study 2 also extended upon Study 1 by considering the potential role of social
desirability bias in responding to altruistic and sexual history items. Research on the
reporting ofsexual historieshas shown thatsocial desirabilitybias mightinfluencesex
differencesinreportingnumberofsexpartners,wherebymentendtoover-report,and
women to under-report, their number of past sexual partners (Fenton, Johnson,
McManus, & Erens, 2001; Jonason, 2007). Research has also linked social desirability
biastocharitablegiving(Lee&Sargeant,2011).Similarly,itisalsopossiblethatthosehigh
innarcissism(grandiose,inflatedviewoftheself)mightengageinsomealtruisticactivities
out of self-interest (Brunell, Tumblin, & Buelow, 2014) and narcissism also relates to
differencesinmatingbehaviour,suchasinmoreshort-termmating(Jonason,Li,Webster,
&Schmitt,2009).IftheresultsfromStudy1weredrivenbysociallydesirableresponding
or by narcissists over-reporting their altruism, then any altruism-mating link should
disappearwhencontrollingfornarcissismandsociallydesirableresponding.
Thus,ourgoalwastodeterminewhetherobservablealtruismintheformofmonetary
donationwouldpredictthesamematingsuccessvariablesasinStudy1,alongwithtwo
additional mating success variables: total number of lifetime dating partners and total
number of sex partners within the past year, while controlling for Big Five personality
factors(assessedusingamorecomprehensivemeasure),narcissism,andsociallydesirable
responding.
Method
Participants
Our sample consisted of 335 undergraduate women and 189 undergraduate men from
northernOntario,Canada,betweentheagesof18and47(M = 20.53,SD = 3.35).The
age
ethnic composition of the sample was largely Caucasian (94%). Participants were
recruitedfromclassroomsatasmalluniversityandcollege.Participantcompletedapaper-
and-pencil-based questionnaire containing self-report measures of personality, social
desirabilityresponding,narcissism,matingsuccess,andsexualhistory.Participantswere
compensatedwithanopportunitytowina$100drawandweregiventheopportunityto
indicatewhethertheywouldliketobeawardedthewinningsortohavetheresearchers
donateit(oraportion)toaregisteredcharityoftheirchoosing(ourbehaviouralmeasure
ofaltruism).
Altruism(monetarydonationtocharity)
Participants completed a self-report survey after which they were told they would be
entered into a draw for a chance to win $100 CAD. At the end of their questionnaire,
participantsreadthefollowingstatement:‘Forcompletingthissurvey,yournamewillbe
enteredintoa$100draw.Ifyouhappentowinthedraw,youmay(1)keepthemoney,(2)
allowustodonatethemoneytoaregisteredcharityonyourbehalf,or(3)acombinationof
thetwo.Ifyouwouldliketodonatetoaregisteredcharity,pleaseanswerthefollowing:I
wouldliketheresearchertodonate$___________ofmywinningstothefollowingcharity
on my behalf (any money not donated will be given to you the participant and can be
Altruismandmatingsuccess 425
picked up at the Evolution Lab A-222H upon notification of winning the draw)’.
Participants opting to donate had the option of World Wildlife Foundation Canada,
CanadianCancerSociety,ALSSocietyofCanada,CanadianMentalHealthAssociation,or
any other registered Canadian charity (as indicated by the participant in the space
provided). Following previous research (Johannesson & Persson, 2000; Nettle et al.,
2013),wewereinterestedinwhetherornotaparticipantmadeadonation(ratherthan
the total amount of money donated), and we therefore operationalized donations as a
dichotomous variable (0 = no, 1 = yes).2 In the present sample, and similar to
proportions of giving in some other laboratory-based dictator games (Nettle et al.,
2013; Rigdon, Ishii, Watabe, & Kitayama, 2009), 65.3% of participants, male = 57.7%,
female = 69.6%, v2(1, N = 524) = 7.52, p = .006, donated some of their potential
winnings,suggestingthatconsistentwithpreviousliterature,womenweremoregiving
thanmen(Eckel&Grossman,1998).
Self-reportedmatingsuccessandmatingsuccessindices(sexualhistoryvariables)
Self-reported mating success was again measured using self-perceived mating success
scale(SPMS;Landoltet al.,1995).AsinStudy1,respondentsagainself-reportedon(1)
theirnumberofconsensuallifetimesexpartners,(2)theirnumberofconsensualcasual
sexpartners,andforthoseinaromanticrelationship(foratleastonemonth)andwho
reportedhavinghadsexwiththeircurrentpartneratleastonce,and(3)thenumberof
timestheyhadengagedinconsensualsexualintercourseoverthepast30 dayswiththat
partner.Ofthetotalsample,250participants(47.7%)metthecriteriaforcurrentlybeing
inaromanticrelationshipwithapersonwhomtheyhadpreviouslycopulatedwith(101
men, 149 women). In Study 2, participants additionally reported their total number of
lifetimedatingpartners,aswellastheirtotalnumberofsexpartnerswithinthepastyear.
Personalitydimensions
The 44-item Big Five Inventory (BFI; John & Srivastava, 1999) was used to measure
extraversion versus introversion (a = .86), agreeableness versus antagonism (a = .75),
conscientiousness versus lack of direction (a = .60), neuroticism versus emotional
stability(a = .83),andopennessversusclosednesstoexperience(a = .73).Participants
respondtoeachitemusinga5-pointLikert-typescale(1 = disagreestrongly,5 = agree
strongly).
Narcissism
The16-itemNarcissisticPersonalityInventory(NPI)wasusedtomeasuretraitnarcissism
(Ames,Rose,&Anderson,2006).TheNPI-16hasshowngoodfacevalidity,discriminant,
and predictive validity. Each item consists of two response items, and the respondent
places an‘X’besidethestatement thatmostcloselydescribestheirfeelingsandbeliefs
aboutthemselves.Anexampleitemisasfollows:‘Ireallyliketobethecenterofattention
ORItmakesmeuncomfortabletobethecenterofattention’.Themeasureshowedgood
internalconsistency(a = .72).
2Thevariable‘amountofmoneydonated’wasalsocollectedbutwasnotnormallydistributed,Shapiro–WilkW(524)=.81,
p=.0001,andwasthusexcludedfromfurtheranalyses.