Presentations & Public Speaking

Are Irrational Reactions to Unfairness Truly Emotionally-Driven? Dissociated Behavioural and Emotional Responses In the Ultimatum Game Task

Neuropsychological and neuroimaging studies suggest distinct body representations involved in coding one's and others’ body. Other influential theories, however, instead posit a unique model behind coding multisensory information about one's
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institutionand sharing with colleagues.Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third partywebsites are prohibited.In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further informationregarding Elsevier’s archiving and manuscript policies areencouraged to visit:  Author's personal copy Are irrational reactions to unfairness truly emotionally-driven? Dissociatedbehavioural and emotional responses in the Ultimatum Game task Claudia Civai a , Corrado Corradi-Dell’Acqua a, * , Matthias Gamer b , Raffaella I. Rumiati a,c a Cognitive Neuroscience Sector, SISSA, Via Beirut 2-4, 34100 Trieste (TS), Italy b Department of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany c Center for Mind/Brain Sciences (CIM e C), University of Trento, Italy a r t i c l e i n f o  Article history: Received 8 October 2008Revised 21 August 2009Accepted 1 September 2009 Keywords: Economical decision-makingUnfairnessAltruistic punishmentFrustrationSkin conductance responseUtilitarian rejections a b s t r a c t The ‘‘irrational” rejections of unfair offers by people playing the Ultimatum Game (UG), awidely used laboratory model of economical decision-making, have traditionally beenassociated with negative emotions, such as frustration, elicited by unfairness (Sanfey, Ril-ling, Aronson, Nystrom, & Cohen, 2003; van’t Wout, Kahn, Sanfey, & Aleman, 2006). Werecorded skin conductance responses as a measure of emotional activation while partici-pants performed a modified version of the UG, in which they were asked to play bothfor themselves and on behalf of a third-party. Our findings show that even unfair offersarerejectedwhenparticipants’payoffisnotaffected( third-party  condition);however,theyshow an increase in the emotional activation specifically when they are rejecting offersdirectedtowardsthemselves( myself   condition).Theseresultssuggestthattheoriesempha-sizing negative emotions as the critical factor of ‘‘irrational” rejections (Pillutla & Murnin-ghan, 1996) should be re-discussed. Psychological mechanisms other than emotions mightbe better candidates for explaining this behaviour.   2009 Elsevier B.V. All rights reserved. 1. Introduction Inrecentyearsthestudyoftheroleofemotionsindeci-sion-making has become an increasingly prominent issuein cognitive neuroscience. A wealth of studies havehypothesized an emotional pathway in the brain thatseems to operate in many types of decisional processes,including moral judgment (Moll, Zahn, de Oliveira-Souza,Krueger, & Grafman, 2005, for a review) and economicaldecision-making (Pillutla & Murninghan, 1996; Sanfeyet al., 2003), that have traditionally been linked to rationalthinking and choices (Kohlberg, 1969; von Neumann &Morgestern, 1947).TheUltimatumGame(UG),amodelofeconomicaldeci-sion-making employed in the laboratory, has always beenthought of as a classical example of emotionally-drivenbehaviour. In this task, one player (the  proposer  ) makes of-fers to a second player (the  responder  ) of how to split anamount of money given by the experimenter; the  respon-der  , in turn, can either accept or reject the offers. If the  re-sponder   accepts, the money will be divided as the  proposer  has decided, otherwise both players will receive nothing.Classical economical theories posit that, to maximize his/herowngain,the  proposer  shouldalwaysofferthesmallestamountof money, whilstthe responder  , followingtheprin-ciplethat ‘‘fewisbetterthannothing”, shouldaccept everyoffer. However, the behavioural findings clearly show thatthe  proposer   typically divides the money equally, and thatthe  responder   rejects offers which favor the  proposer   toomuch, and those that he/she considers unfair (Bolton &Zwick, 1995). Importantly, this behavioural pattern hasalso been observed in both the single-shot UG, in whichthe two players interact only once, and in the  covered  UG,inwhichthe  proposer   isnotinformedaboutthe responder  ’sreaction (Abbink, Sadrieh, & Zamir, 1999; Zamir, 2001),both of which are paradigms where rejections lose theirrole as negotiating tools. 0010-0277/$ - see front matter   2009 Elsevier B.V. All rights reserved.doi:10.1016/j.cognition.2009.09.001 *  Corresponding author. Tel.: +39 040 3787628; fax: +39 040 3787528. E-mail address: (C. Corradi-Dell’Acqua).Cognition 114 (2010) 89–95 Contents lists available at ScienceDirect Cognition journal homepage:  Author's personal copy Individuals’ ‘‘irrational” choices have been explained interms of altruistic punishment (Fehr & Gachter, 2002).According to this theory, the punishment, even if costlyand yielding no direct benefit for the punisher (as in thecase of single-shot UG), is used to penalize selfish behav-iourof others, as it leads themtocooperateinfutureinter-actions (Fehr & Gachter, 2002). It has been suggested thatirrational rejections might be best explained by negativeemotions, such as frustration, that drive participants topunish rather than making an utilitarian choice (Fehr &Gachter, 2002; Pillutla & Murninghan, 1996). Consistentwith this view, Sanfey et al. (2003) have recently associ-ated the rejection of unfair offers with an increase of boththe neural activity in anterior insula, traditionally corre-lated with feelings of anger and disgust (Calder, Lawrence,& Young, 2001; Phillips et al., 1997), and the skin conduc-tanceresponse(van’tWoutetal.,2006),ameasureofemo-tional activation (Bouscein, 1992).It has also been argued that only self-centered emo-tions, for instance anger and frustration, play some rolein the UG, as individual payoff is heavily involved in thistask (Moll & de Oliveira-Souza, 2007). However, it has alsobeen proved that individuals choose an act of punishmenteventhoughtheirpayoffsarenotdirectlyaffectedbyavio-lation of fairness and cooperation norms (i.e. the third-party punishment). Fehr and Fischbacher (2004), for in-stance, found that participants decided to give up someof their own money to punish the unfair behaviour of one player towards another. Thus, altruistic punishmentalso occurs in conditions in which unfairness should notelicit, at least in principle, any self-centered emotion. Thisraises the questionof whether, in the UGtask, the ‘‘irratio-nal” punishing behaviour and negative emotions are al-ways causally related, or whether they can operateseparately depending on the  myself  / third-party  distinction.In the present study, we investigated the role of emo-tions in the UG by measuring skin conductance responses(SCR) while participants played as  responders  in a modifiedversion of the UG and by collecting emotional ratings afterthey completed the task to measure the valence of thehypotheticalarousal.Participantscarriedoutboththeclas-sicalversionoftheUGandamodifiedversionofthetaskinwhich any putative monetary income was not going intothe participants’ own pocket, but into a third-party’s (seeSection 2). Indeed, in the latter condition the  proposer  ’s of-fer did not directlyaddress the participant’s payoff, unfair-ness should in principle, elicit neither self-centeredemotions (Moll & de Oliveira-Souza, 2007), nor, as conse-quence, SCR increases when the offer is about to be re- jected (van’t Wout et al., 2006). Thus, the accountaccording to which the punishing behaviour and the nega-tive emotions are causally related (Fehr & Gachter, 2002;Pillutla & Murninghan, 1996), also predicts that such emo-tional decrease should be associated with a similar de-crease in the amount of punishing choices (rejections)(van’tWoutetal.,2006).However,basedonpreviousstud-iesofaltruisticpunishment(e.g.Fehr&Fischbacher,2004),we predicted that participants should reject unfair offersaddressing a third-party; if this were indeed the case, weexpected a significant increase in SCR for offers about tobe rejected even in the third-party condition as well. 2. Methods  2.1. Participants Thirty-fourhealthyItalianvolunteers(22females), whoranged in age from 18 to 35years ( M   =23.56, SD=3.90),took part in the experiment. They all were paid for partic-ipating in the study, the scientific goal of which was un-known to them. The study was approved by the localethics committee and conducted in accordance with theDeclaration of Helsinki.  2.2. Task Participants were required to play as  responders  in amodified version of the UG and had either to accept or re- ject the offers the  proposer   made, following the classicalrules explained above. Before starting the game, theywere introduced to a collaborator of the experimenter,who pretended to play as the  proposer  , in order tostrengthen the illusion of playing against a human adver-sary, whereas they were actually playing against a com-puter. They were told that the opponent had been givena number of 10euros bank notes and would have to makeoffers on how to split each of them. Consistent with pre-vious studies (e.g. Polezzi et al., 2008), offers in each trialcould be either 1, 2, 3, 4 or 5euros out of 10. Further-more, participants were informed that, in one condition,they and their opponent would play for themselves (con-sistent with the classical UG), whereas, in another condi-tion, they would play on behalf of those players acting as  proposer   and  responder   in the upcoming testing session(see Fig. 1). In order to make our task compatible to thesingle-shot UG, participants were told that the opponentwould receive feedback only at the end of the experiment,when they have both been informed on how much eachof them had gained, depending on the choices they hadmade; in this way, they knew rationally that they couldnot affect the opponent’s behaviour through theirrejections.To control for the social interactive nature of the UG,participants performed a control task (Free Win [FW]task) in which they either accepted or rejected a vari-able amount of money given by the computer (1, 2, 3,4 or 5euros). As in the case of the UG, they could de-cide for themselves or on behalf of the next participant.If they accepted the offer, they/the third-party wouldreceive that amount, otherwise they/the third-partywould receive nothing. This yielded to a 2  2  5 de-sign, with TASK (UG  vs . FW), TARGET ( myself vs .  third- party ) and GAIN (1, 2, 3, 4 or 5euros) as within-subjectsfactors.Participants were informed that their compensation forparticipating in the experiment would be proportional totheamountofmoneygainedinthe myself  condition.More-over, they knew that a percentage of the money split onbehalf of third partieswouldbegiventonext players; theywere also informed that, following the same principle,their starting stakes were percentages of the money thatprevious players had split on their behalf. Irrespective of  90  C. Civai et al./Cognition 114 (2010) 89–95  Author's personal copy their performance on the task, participants received thesame amount of money as compensation. Although wedid not systematically investigate whether participantshad doubts about the authenticity of the situation, themajorityofthem,wheninformallyinterviewedafterwards,said they believed they had played against a human oppo-nent. Only a few reported having doubts at the end of theexperimental session.  2.3. Apparatus and procedure All theparticipantsweretestedinaquietroomat SISSAusing a PC and a 15-in monitor (Olidata s.p.a.). Presenta-tion  12.0 Software ( was usedtoconstructanddelivertheexperimentalstimuli.Theofferappeared on the screen for 5s, followed by a 6s blankscreen. Participants were required to respond by buttonpress, highlighted on the computer keyboard, as soon asthe question ‘‘do you accept?” appeared on the screen,where it lasted for 2s (see Fig. 2). The inter-trial intervalwas around 11s on average, to allow skin conductance toreturn to its baseline. All 20conditions, each of whichwas repeated four times, were presented in a randomizedorder. The whole experiment (80trials  24s of trial dura-tion) includingashort breakof 1minafter halfof thetrialslasted approximately 33min.  2.4. Skin conductance recordings Skin conductance was recorded during the wholeexperiment using a pair of prewired 8mm Ag/AgCl elec-trodes, attached to the distal phalanx surfaces of the indexand little finger of the non-dominant hand. The electrodepair was excited with a constant voltage of 0.5V and con-ductance was recorded using a DC amplifier with a low-pass filter set at 64Hz and a sample frequency of 256Hz.Values of skin conductance were automatically trans-formed to microsiemens values by the Procomp InfinitySystem (Bio-Medical Instruments, Inc., Warren, MI, USA ) .Before starting the task, 1min of baseline was recorded.We measured the artifact-free amplitude of the skin con-ductance response that began between 1 and 3s after thepresentation of the offer and exceeded a threshold of  Fig. 1.  Illustrationofthetaskasitwaspresentedtoparticipantswhengivingtheinstructions.Therearefourconditions:thefirstandthesecondrefertotheUltimatumGame andthe thirdand the fourth refer to the control task (Free Winsituation). Inthe first and in the thirdconditions participants are askedtodecide for themselves, whereas in the second and in the fourth they are asked to choose on behalf of a third-party (next participant). Fig. 2.  Time line for each single trial of the Ultimatum Game. Each triallasted 24s. First, participants sawthe offer on the screen for 5s, followedby6sofblankscreen.Next,thequestion‘‘doyouaccept/doyouacceptonhis behalf” appeared on the screen for 2s, within which participants hadto answer by buttonpress. An average of 11s inter-trial interval followedthe question. C. Civai et al./Cognition 114 (2010) 89–95  91  Author's personal copy 0.05 l S. Inthecaseof overlappingresponses, theinflectionpoint between the two responses served as the baseline orpeak, depending on the latency criterion. The resultingamplitudes were  z  -transformed within each participantinorder to eliminateindividual differencesin responsivity.  2.5. Emotional ratings To further investigate the emotional reactions in ourstudy, participants rated their feelings in the most crucialconditions (i.e. 1, 3 and 5 euros of gain when playing theUG in both the  myself   and the  third-party  condition) atthe end of the experimental session. Since we were inter-ested in detecting the presence of a perceived emotion,here we included only the extreme conditions (and notthose associated with 2 and 4euros) on the assumptionthat they were those more likely to elicita stronger, andthus the most easily detectable, response. The mid-valueoffer (3euros) was chosen as a baseline, since it could re-flect the emotionally neutral condition. Participants useda 12-point Likert-scale for each condition ranging from  6, correspondingtostrongnegativeemotions,to+6, indi-cating strong positive emotions. 3. Results  3.1. Rejection rates For each subject and condition, the rejection rates werecalculated across all 4 repetitions, and used in a 2 (TASK:UG, FW)  2 (TARGET:  myself  ,  third-party )  5 (GAIN: 1, 2,3, 4, 5euros) RepeatedMeasuresANOVA. Statistical Analy-sis was carried out using SPSS 11.5 Software (SPSS Inc.,Chertsey UK). Results indicated a significant main effectof TASK ( F  (1, 33)=76.24,  p  <.001,  g 2  p  ¼ : 69), with the UGelicitingalarger amount of rejections thanthe FW(seeTa-ble 1 and Fig. 3), as well as a main effect of GAIN ( F  (4,132)=52.7,  p  <.001,  g 2  p  ¼ : 61), with low offers being re- jected more than high offers. This effect is however drivenby the TASK  GAIN interaction, which was found to besignificant as well ( F  (4, 132)=49.89,  p  <.001,  g 2  p  ¼ : 60),suggesting that low offers are rejected significantly moreoften than high offers in the UG but not in the FW. Noneof the remaining effects of the ANOVA were found to bestatistically significant.  3.2. Emotional ratings We analyzed the emotional ratings for the most unfairoffer (1euros out of 10), the fairest offer (5euros out of 10) and the mid-value offer (3euros out of 10), both forthe  myself   and the  third-party  conditions. One-sampletwo-tailed  T  -tests showed that for the mid-value offerthe ratings did not differ significantly from zero (i.e. theneutral emotion), while for both targets, the ratings forthe unfair offer were significantly different from0 towardsthe negative emotion (UG (1:9)  myself  :  t   (33)=  9.79,  p  <.001 UG (1:9)  third-party :  t   (33)=  4.37,  p  <.005), andso were those for the fair offer towards the positive emo-tion (UG (5:5)  myself  :  t   (33)=22.29,  p  <.001; UG (5:5)  Table 1 Rejection rates (RR) (%) and skin conductance response amplitudes (SCR amp) (  z  -transformed  l S) for the four conditions collapsed by gain. UG FWMyself Third-party Myself Third-partyRR (SEM) 35.73 (5.59) 38.09 (4.93) 2.64 (0.68) 8.77 (0.59)SCR amp (SEM) .0887 (.0157)   .0073 (.0164) .0257 (.0146)   .1096 (.0105) Note:  Corresponding standard errors of the mean are printed in brackets. Fig. 3.  Behavioural results. Rejection rates (%) plotted as a function of GAIN in the  myself   (A) and the  third-party  (B) condition.92  C. Civai et al./Cognition 114 (2010) 89–95
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!