the role of cultural orientation in bargaining under incomplete information: differences in causal...

www.elsevier.com/locate/obhdp

Organizational Behavior and Human Decision Processes 96 (2005) 72–88

The role of cultural orientation in bargaining underincomplete information: Differences in causal attributionsq

Ana Valenzuelaa,*, Joydeep Srivastavab, Seonsu Leec

a College of Business, San Francisco State University, San Francisco, CA 94132, United Statesb Robert H. Smith School of Business, University of Maryland, College Park, MD 20742, United States

c College of Business and Economics, Wonkwang Univerity, Iksan City, Chollabuk-Do 570-749, Republic of Korea

Received 18 March 2004Available online 8 December 2004

Abstract

This research examines how differences in cultural orientation influence causal attributions and thus the behavioral outcomes inan incomplete information bargaining situation. Using ultimatum bargaining, three experiments demonstrate that acceptance ratesdiffer across Western and East Asian cultures because of the differences in implicit theories of behavior. The results of Experiment 1shows that East Asians are more sensitive to both external constraints and group influences but only when there is informationabout the opponent�s situation to discount personality traits. Experiment 2 shows that reasons for an opponent�s behavior mediatethe influence of cultural orientation on bargaining outcomes when situational constraints are made salient. Experiment 3 shows thatreasons for an opponent�s behavior based on the saliency of a group context mediate the influence of cultural orientation on behav-ior. The paper concludes by discussing the implications of the findings and suggesting directions for future research.� 2004 Elsevier Inc. All rights reserved.

Keywords: Culture; Bargaining; Negotiations; Incomplete information; Inferences; Causal attributions

With the growing trend in the globalization of busi-ness activities, there has been a significant increase inthe frequency of cross-cultural business interactions(Triandis, 1994). As economies become more intercon-nected, it is critical to understand the influence of culture

0749-5978/$ - see front matter � 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.obhdp.2004.09.001

q This research was supported in part by the grant awarded byWonkwang University in 2002 to the third author and by the DeanContribution Fund and the Affirmative Action Research Grantawarded by San Francisco State University to the first author. Theauthors acknowledge the assistance of Lane Auten, Foo Nin Ho, andDiane Huang in collecting the data. The authors also thank PriyaRaghubir, Sharon Shavitt, Susan Taylor, Bob Wyer, the conferenceparticipants at the Association of Marketing Science World Confer-ence, Perth, Australia, and the Association of Consumer ResearchConference, Toronto, Canada for their helpful comments on earlierversions of this paper.

* Corresponding author. Fax: +1 415 338 0596.E-mail addresses: [email protected] (A. Valenzuela), srivasta@

rhsmith.umd.edu (J. Srivastava), [email protected] (S. Lee).

on all aspects of organizational behavior including bar-gaining and negotiations. Many companies depend onthe outcomes of cross-cultural negotiations for theircontinued success and profitability in today�s globaleconomy. Much of the existing negotiations literaturein organizational psychology however consists predom-inantly of experimental simulations investigating theo-retical frameworks developed using evidence fromWestern cultures, predominantly the United States (cf.Graham, Kim, Lin, & Robinson, 1988). According toone estimate, about 90% of social and organizationaltheories have been developed and tested in Western con-texts (Triandis, 1994). As such, relatively little is knownabout the generalizability of existing theoretical frame-works across diverse cultures and the validity of theseframeworks in predicting processes and outcomes innon-Western cultures is not clear. There is therefore apressing need to explicitly examine whether existingframeworks are universal or are laden with assumptions

mailto:[email protected]

mailto:srivasta@

mailto:[email protected]

A. Valenzuela et al. / Organizational Behavior and Human Decision Processes 96 (2005) 72–88 73

derived from Western cultures (Gelfand et al., 2002; Tri-andis, 1995).

The main purpose of this research is to examine therole of cultural orientation on bargaining outcomes.The general premise, following the recent literature incultural psychology, is that many judgments and deci-sions are the result of cognitive processes that are cultur-ally imposed (e.g., Chiu, Morris, Menon, & Hong, 2000;Menon, Morris, Chiu, & Hong, 1999). The rationale isthat exposure to different ecological factors and socialstructures perpetuates different cultural values and ide-als and thus certain judgment ‘‘biases’’ are likely to bemore prevalent in one culture than another (Triandis,1995). Specifically, this research examines how differ-ences in cultural orientation influence behavioral out-comes due to differences in causal attributions andinferences in an incomplete information bargaining situ-ation. Incomplete information bargaining scenarios areparticularly important because most business transac-tions are characterized by information asymmetry anduncertainty about the mutual gains from trade. In theabsence of clearly specified referents, potential bargain-ing outcomes cannot be easily assessed and there is anatural inclination to search for causal explanationsfor others� (opponent�s) behavior (Morris, Larrick, &Su, 1999; Srivastava, 2001). The nature of causal attri-butions may differ based on cultural orientation therebyaffecting bargaining processes and outcomes (Morris &Peng, 1994).

In this research, we use ultimatum bargaining as thesetting to study the role of cultural orientation (e.g., Bu-chan, Croson, & Johnson, 2004). In ultimatum bargain-ing, an agent (proposer) makes an offer to another agent(responder) that divides a specified sum of money be-tween the two agents. The responder can then either ac-cept or reject the offer. If the offer is accepted, the sum ofmoney is divided as proposed and the game ends. If theoffer is rejected, both agents receive nothing and thegame ends. The ultimatum game provides the appropri-ate context as it is not only a model for basic transac-tions but also represents the end state of anycontinuous bargaining. In addition, social psychologistsargue that the tendency to attribute observed behaviorto individual traits is more likely to occur in interper-sonal conflict situations such as ultimatum bargaining.Perhaps one of the main reasons for our choice of theultimatum game as a means of investigating the role ofculturally dependent inferences is that offer acceptancebehavior has been shown to depend on the responder�sevaluations of the proposer�s offer and the reasons lead-ing to making that offer (Henrich et al., 2001; Thaler,1988). Finally, ultimatum bargaining offers a simplestructure that allows isolation of the factors of interest.

Reflecting its importance as a building block for morecomplex bargaining, ultimatum bargaining has beenextensively studied in the literature (see Camerer & Tha-

ler, 1995). The standard economic prediction is thatsince any amount of money is better than no money, arational proposer should demand (and the respondershould accept) the smallest amount over zero. Contraryto the normative prediction, proposers tend to makegenerous offers (averaging about 40% of the surplus)and responders often reject positive offers when theyare below 20% of the total surplus (Camerer & Thaler,1995). These systematic deviations from the normativeprediction have been attributed to considerations offairness (Thaler, 1988).

Ultimatum bargaining has also been examined incross-cultural contexts. For instance, Roth, Prasnikar,Okuno-Fujiwara, and Zamir (1991) in a four-countrystudy found that proposer offers in Jerusalem, Ljusblj-ana, Pittsburgh and Tokyo all lay within the range of20–50%. However, there were notable differences be-tween the distributions of the offers as well as the distri-bution of acceptance rates across different countries.These differences in bargaining behavior across coun-tries were not due to differences in language or curren-cies, but were tentatively attributed to culturaldifferences. The authors proposed that differences inbargainers� conception of what constitutes a fair (or‘‘reasonable’’) offer might be an important explanatoryvariable. Henrich et al. (2001) compared ultimatum bar-gaining outcomes across 15 small scale societies andfound that proposer offers among the Machiguengatribe in Peru were about 26%, while those among theLamelara in Indonesia were greater than 50%. They alsoobserved zero rejection of positive offers among theAchuar (Ecuador), Ache (Paraguay) and Tsimane (Boli-via) tribes, even though some of the offers were at or be-low 30%. They attributed the large variations inbargaining behavior across the different cultural groupsto group-specific conditions, such as social institutionsor cultural norms about fairness. They further suggestthat subjects use situations in their daily experience todefine appropriate behavior and thus differences in thestructures of social interaction, such as the necessity tocooperate or share, and in the mode of livelihood aretransferred to experimental settings making certain soci-eties more or less predisposed to forego material payoffsin order to share with others, or to punish unfairactions.

However, both these studies, and much of the ultima-tum bargaining literature, focus on complete informa-tion situations, where the agents have full informationabout one another�s payoffs and thus inferences aboutthe causes for an opponent�s behavior are not necessary(e.g., Buchan et al., 2004; Camerer & Thaler, 1995).Arguably, incomplete information ultimatum bargain-ing requires a different assessment of what a reasonableoffer constitutes. Because of the uncertainty about thetotal surplus to be split between the agents, an assess-ment of what constitutes a normative fair or ‘‘reason-

74 A. Valenzuela et al. / Organizational Behavior and Human Decision Processes 96 (2005) 72–88

able’’ offer is not possible. Instead bargainers tend to in-fer the reasons underlying the opponent�s bargainingbehavior in order to justify their own decision. Indeed,research shows that bargainers� attributions for oppo-nents� behavior play a critical role in incomplete infor-mation bargaining (Srivastava, Chakravarti, &Rapoport, 2000; Srivastava, 2001). Bargainers� cultur-ally imposed structures for social interaction are ante-cedents of the tendency to attribute behavior topersonal dispositions instead of contextual factors(Nisbett, 2003; Morris & Peng, 1994). The objective ofthis research is to examine whether differences in causalattributions that are culturally driven affect the out-comes in ultimatum bargaining under incompleteinformation.

Broadly, this research adds to the literature on bar-gaining and cross-cultural psychology by examininghow cultural orientation affects bargaining behaviorand outcomes in situations characterized by the absenceof objective referents and standards against which tojudge potential outcomes (White & Neale, 1994). Thecultural psychology literature suggests that causal attri-butions are culturally dependent (e.g., Choi, Nisbett, &Norenzayan, 1999) and thus systematic differences inbargaining outcomes across cultures may be due to dif-ferences in the causal attributions for opponents� behav-ior (Shafir, Simonson, & Tversky, 1993). This researchextends this line of inquiry to a bargaining context andattempts to link causal attributions to bargaining out-comes. Three experiments show that subjects from Wes-tern cultures have a tendency to seek causalexplanations for an opponent�s behavior in terms ofindividual personality traits. In contrast, subjects fromEastern cultures are more likely to recognize that anopponent�s behavior may be dictated by situational fac-tors or by other-oriented traits but only where the exter-nal constraints are made salient.

Conceptual background

Conflict resolution and cultural orientation

Cross-cultural psychologists have proposed explana-tory models of how culture influences negotiations andresearchers in this tradition explain cultural differencesin terms of the stable, general characteristics of negotia-tors such as the degree to which their values are individ-ualistic as opposed to collectivistic (e.g. Triandis, 1994).Individualistic versus collectivistic scores have beenempirically associated with differences in negotiationbehavior between Western (US) versus Chinese subjectssuch as the extent of the fixed pie error (Gelfand & Chri-stakopoulou, 1999) and rewards distribution (Leung,1987). Similarly cultural differences in self concept suchthat the Chinese, believing in a relatively fixed social

world, would be more likely than Americans to focuson collective duties and social conservatism, and Amer-icans, believing in a relatively malleable social world,would be more likely than Chinese to focus on individ-ual rights and decision autonomy, have been put forthto explain differences in negotiation styles between cul-tures (Chen, Mannix, & Okumura, 2003). Researchhas also documented that negotiation behavior is depen-dent on cultural differences in theories of fairness andjustice (Morris, Leung, & Ames, 1999).

However, individualistic-collectivistic value scorescannot explain when culture has a strong influence andwhen a weak influence on a given individual. To addressthis issue, researchers suggest that cultural differences inconflict resolution behavior may reflect individualknowledge structures (Morris & Fu, 2001). Theseknowledge structures, such as implicit theories or mentalmodels, guide judgments and decisions and, ultimately,behavior. Social psychologists argue that negotiationcontexts are particularly likely to give rise to interpreta-tions of the counterpart�s behavior (Morris, Larricket al., 1999). This is even more likely when there is ambi-guity in the negotiation context, such as in incompleteinformation bargaining (Srivastava et al., 2000; Srivast-ava, 2001). The knowledge structures most relevant tothese judgments are the beliefs about the dominant cau-sal explanations for behavior. These beliefs determineindividuals� causal attributions of observed behavior.

Causal attributions and cultural orientation

One of most commonly documented judgment bias insocial psychology, labeled the fundamental attributionerror (also known as correspondence bias or overattri-bution in the literature), is the tendency to attribute ob-served behavior to internal personality characteristicsrather than external situational constraints (Nisbett &Ross, 1980). Does the fundamental attribution errorvary across cultures? Describing the fundamental attri-bution error as a universal human tendency, social psy-chologists have traditionally assumed that principles ofcausal attribution are invariant across cultures (Nisbett& Ross, 1980). According to the universalist view, cul-tural orientation should not affect causal attributions.

In contrast, more recent research recognizes thatattribution patterns reflect implicit theories acquiredfrom induction and socialization and thus vary withthe perceiver�s cultural orientation (e.g., Choi et al.,1999; Menon et al., 1999; Morris & Peng, 1994). Forexample, research suggests that the person-centered the-ory that behavior is caused by stable internal personalitytraits is more entrenched in individualistic cultures be-cause individuals are viewed as autonomous entitiesand are socialized to behave according to personal pref-erences. In contrast, the situation-centered theory thatbehavior is shaped by relationships and situational fac-


tors is more prevalent in collectivistic cultures becauseindividuals are part of a social collective and are social-ized to behave according to situational constraints andgroup norms (Morris & Peng, 1994; Triandis, 1995).Consistent with this idea, several studies have reportedthat causal explanations of behavior focus largely onthe situational context in East Asian cultures whereasWestern cultures prefer explanations based on individ-ual traits (e.g., Choi et al., 1999). Other studies haveadded that cultures differ in implicit theories of individ-uals and groups so that East Asians are more likely thanNorth Americans to focus and explain behavior usingdispositions of collectives (e.g., Menon et al., 1999).According to the dispositional view, culture always af-fects causal attributions in one way or another.

Despite the numerous studies on cultural differencesin the literature, the dispositional view that culturealways matters has been undermined by studies that re-port no cultural differences. Further, a recent meta-anal-ysis of cross-cultural research found no reliable culturaldifferences based on the individualism-collectivismdimension (Oyserman, Kemmelmeier, & Coon, 2002;Takano & Osaka, 1999). These and other related find-ings suggest that the role of cultural orientation maybe contextually driven such that cultural differencesare influential only when people are required to drawupon the implicit theories that differ across cultures (Bri-ley, Morris, & Simonson, 2000; Morris & Fu, 2001). Bri-ley et al. (2000) showed that cultural differences weremanifested only when decision makers were asked toprovide reasons for their decisions. Providing reasonsfor the decisions activates knowledge structures thatare different across cultures (Higgins, 1996; Morris &Fu, 2001). Briley et al. (2000) suggest that possessingan implicit theory does not imply that it will be used uni-versally, but only when it is needed or is made accessible(Choi & Nisbett, 1998; Knowles, Morris, & Chiu, 2001).

Causal attributions, cultural orientation and ultimatum

bargaining

Does the nature of the causal attributions and, there-by, behavioral outcomes vary across cultures? The uni-versalist view of culture predicts no difference. Thedispositional view suggests that Western cultures wouldalways be more likely to seek causal explanations basedon personality traits whereas East Asian cultures wouldalways be more sensitive to situational factors (Morris,Leung, & Iyengar, 2004). In contrast to the dispositionalview, research shows that both East Asians and West-erners exhibit the fundamental attribution error (Morris& Peng, 1994; Krull et al., 1999). Further, Choi andNisbett (1998) compared attributions across Koreaand the United States and found that when situationalconstraints are made salient Korean subjects were moresensitive to the situational constraints than the US sub-

jects (also Krull et al., 1999). Similarly, Chiu et al. (2000)found that individual differences in need for closure(NFC) moderate the effect of culture on inference mak-ing. The authors propose that NFC leads people tointerpret an ambiguous social event by increasing theirreliance on implicit theories received from acculturation.These findings suggest a third and more dynamic view ofculture. Taking the dynamic view of the role of culturalorientation, this research proposes that behavioral out-comes will change due to cultural differences in causalattributions only when there is a need to draw uponthe implicit theories that differ across cultures or whenthese theories are activated (Briley et al., 2000; Choi &Nisbett, 1998; Morris & Fu, 2001).

Three ultimatum bargaining experiments examine thedynamic role of cultural orientation in affecting behav-ior under incomplete information by comparing bar-gaining outcomes between subjects from the US, ahighly individualistic Western culture, and Korea, ahighly collectivistic East Asian culture. In an ultimatumbargaining context, rejection of a positive offer has beenexplained by the intent to punish a competitive and un-fair proposer (Bolton & Zwick, 1995). As such, the morean opponent�s behavior (the offer) is attributed to theproposer�s competitive personality, the higher the prob-ability of rejection (Srivastava, 2001). We propose thatthere will be no difference in the causal attributions foran opponent�s behavior and thereby bargaining out-comes across cultures when there is no informationavailable about an opponent�s situation to discount per-sonality attributions. However, when information aboutan opponent�s situation is accessible, implicit theoriesthat differ across cultures are activated leading to differ-ences in behavioral outcomes across cultures. It is pre-dicted that while Western subjects will continue tofocus on internal personality traits and exhibit similarbehavioral patterns, East Asian subjects will be moresensitive to the external constraints (e.g., the possibilitythat the available surplus could be small), therebyincreasing acceptance rates.

Hypotheses

Individual versus situational attributions

Consider a one-sided incomplete information ultima-tum situation where the responder is uncertain but theproposer knows the total amount available for division.Uncertain about the total amount available for divisionand thus unable to assess the fairness of an offer, a pri-mary concern of the responder is about the relativeshare or proportion of the total amount that a pro-poser�s offer represents (e.g., Camerer & Thaler, 1995;Croson, 1996). In this situation, responders may usethe proposer�s offer to infer the total amount available


and thereby the proportion of the total amount that aspecific offer represents (Croson, 1996; Morris, Larricket al., 1999). Given the uncertainty, there is also a natu-ral inclination to develop a causal explanation for anopponent�s behavior (Blount, 1995) which then deter-mines whether or not an offer is accepted.

Consider the relatively high and low offers of $12.50and $7.50, respectively, when the total amount availablefor division is $25, but the responder only knows thatthe total amount could be any whole number between$10 and $40 with equal probability (uniform distribu-tion). Since $25 represents the mean of the distribution,the likelihood is higher that an offer of $12.50 representsa more fair and favorable division of the total amountrelative to an offer of $7.50. In fact, responders are likelyto attribute the proposer�s unfavorable offer to his/herpersonality traits such as level of competitiveness (oropportunism) rather than to the total amount available(situational constraint), even if the offer of $7.50 were torepresent an equal split of the total amount available(e.g., Morris, Larrick et al., 1999; Srivastava, 2001). Itis proposed that when situational constraints are notmade salient or when there is no additional informationabout the proposer�s situation to discount personalityattributions, bargainers in both cultures are likely toattribute the proposer�s offer to personality traits (Choi& Nisbett, 1998; Krull et al., 1999). As such, while thebargaining outcomes are likely to be different acrossthe relatively high and low offers, there will be no differ-ence in the outcomes (acceptance rates) across US andKorean bargainers.

However, when the situational constraints are madesalient, bargainers� causal attributions are likely to dif-fer across cultures. Making the situational constraintssalient or accessible allows bargainers to draw upontheir implicit theories that differ across cultures. Bar-gainers in Korea are more likely to recognize thepower of situational constraints as they draw uponthe implicit theory that an individual�s behavior isshaped by situational factors. In other words, relativeto US bargainers, Korean bargainers are likely to bemore sensitive to situational constraints in their causalattributions of an opponent�s behavior. On observing arelatively low offer of $7.50, Korean responders aremore likely to reduce their tendency to attribute theoffer to proposer�s personality traits (e.g., competitive-ness and opportunism) and take into account the pos-sibility that the low offer is a reflection of the totalamount available. As a result, when situational con-straints are made salient, bargaining outcomes arelikely to differ across US and Korean responders. Spe-cifically, we predict that the acceptance rates for a rel-atively low offer are likely to be higher for Koreanbargainers than US bargainers, when information thatcan be used to discount personality attributions ismade salient or accessible.

H1: While there will be no difference in the accep-tance rates among bargainers in Korea and US when sit-uational constraints are not salient, the acceptance ratesfor a relatively low offer will be higher among Koreanbargainers relative to US bargainers when situationalconstraints are made salient.

Our argument is that the differences in bargainingoutcomes (H1) are caused by differences in causalattributions across cultures and that these differencesare a function of whether information is available oraccessible to discount personality traits. While EastAsians become more sensitive to situational determi-nants of behavior when the constraints are made sali-ent and tend to correct the extent of their attributionsto personality characteristics, Westerners continue toattribute behavior to personality traits regardless ofthe saliency of the situational constraints (Choi &Nisbett, 1998; Krull et al., 1999). In other words, wepropose a mediation account in which the accessibilityof situational influences increases the strength of bar-gainers� attributions of the low offer to reasons otherthan the proposer�s personality traits, which, in turn,increases acceptance of the low offers, but only forKorean bargainers. Specifically, we test twohypotheses:

H2: When situational constraints are salient, Koreanbargainers will rate situation-based reasons as a moreimportant cause for a relatively low offer than USbargainers.

H3: The acceptance rates of a relatively low offer willbe mediated by individuals� importance ratings of situa-tion-based reasons.

Individual versus group attributions

While a majority of the research on cross-cultural dif-ferences in causal attributions has focused on implicittheories of persons (i.e., personality-based versus situa-tional attributions), recent research argues that whileWestern cultures believe that individuals have stableinternal traits that can cause social outcomes and groupsdo not, East Asian cultures hold that groups have thepower to influence social outcomes (Chiu et al., 2000;Menon et al., 1999). In other words, cultural differencesin causal attributions and subsequent behavior may notonly be based on differential sensitivity to situational/contextual accounts but also on differences in implicittrait beliefs. Although group constraints can be con-ceived of as another situational constraint, in contrastto non-social situational factors that influence behavior,a group is a social entity that possesses dispositionsindependent of individual traits and can cause outcomes(Higgins & Bryant, 1982). Importantly, to the extentthat cultures differ in their implicit theories of groups,there may be differences in behavioral outcomes acrossKoreans and North Americans. From a practical per-


spective, individuals commonly bargain on behalf of agroup.

The implicit theories of the group vary across culturesbecause of differences in conceptions of individual ver-sus collective autonomy. As discussed earlier, individu-alistic cultures consider people to be autonomousentities possessing stable internal traits, and these enti-ties should not be influenced by collectives unless theydesire (Markus & Kitayama, 1991). In other words,the prevalent thinking is that an individual�s action isprimarily dictated by stable personality traits and thatthe responsibility for his/her behavior does not disap-pear even with the possibility of group influence. In fact,a group consists of individuals and these individualsinfluence the group�s outcomes. In contrast, in collectiv-istic cultures, individuals are perceived as part of agroup and the group is believed to be autonomous andpowerful. Entrenched in the notion of collectivism, indi-vidual behavior is perceived to be shaped by relation-ships and conformity to group norms or directives. Asa result, East Asians are more likely than Westernersto attribute an individual�s behavior to group influenceinstead of to the individual�s personality (Menon et al.,1999). While cultural differences in implicit theories ofthe group have been documented in social settings, thisresearch extends the inquiry to economically meaningfuldecision contexts such as bargaining.

In an ultimatum bargaining scenario, consider an of-fer of $7.50 when the total amount available for divisionis $25 and the responder only knows that the totalamount available could be anywhere from $10 to $40.Given the high likelihood that the $7.50 offer is lowand represents an unfavorable division of the totalamount, as discussed previously, we expect bargainersin both Korea and US to attribute the proposer�s offerto his/her personality traits (e.g., level of competitive-ness). There should therefore be no differences in behav-ioral outcomes when the proposer is not makingdecisions in a group context.

In contrast, when the proposer is making the decisionof how much to offer in a group context, the differencesin implicit theories of the group across individualisticand collectivistic cultures should lead to differences inbargaining outcomes. Because collectivistic cultures be-lieve that individual behavior is shaped by group normsand directives, individuals in collectivistic cultures aremore likely to give the benefit of the doubt to the pro-poser and attribute the low offer to the group. On theother hand, individualistic cultures believe that an indi-vidual is accountable for his/her individual behavior andhas the capability to influence group behavior, evenwhen decisions are made in a group context. As a con-sequence, North Americans are more likely to continueattributing the low offer to the proposer�s personalitytraits (e.g., competitiveness or opportunism) whereasKoreans are more likely to be sensitive to the group

influence and the extent to which the individual had togive in to the group�s priorities and, therefore, is notpunishable for it. A relatively low offer of $7.50 is there-fore more likely to be accepted by bargainers in Koreathan in US, when the proposer is believed to be makinghis/her decision in a group context.

H4: While there will be no difference in the accep-tance rates among bargainers in Korea and US whenthe decision is made in an individual context, the accep-tance rates for a relatively low offer will be higher amongKorean bargainers relative to US bargainers when theproposer�s decision is made in a group context.

Consistent with previous research (e.g., Chiu et al.,2000), differences in implicit theories of group acrossthe two cultures underlie the behavioral differences suchas those hypothesized in H4. In Western cultures, theprevalent thinking is that individuals can shape groupbehavior and, thus, are primarily responsible for theirdecisions even in a group context. In East Asian cul-tures, the thinking is that individual behavior is shapedby the group and that conformity to group norms ordirectives is responsible for an individual�s decision-making in a group context (Menon et al., 1999). There-fore, in an incomplete information bargaining situation,Korean responders, when facing a relatively low offer,are more likely to consider the influence of the groupon the individual�s decision relative to US responders.Further, the differences in causal reasoning mediatethe influence of cultural orientation on bargaining out-comes when the decision of how much to offer is madein a group context.

H5: In a group decision context, Korean bargainers�ratings of group-based reasons as a cause for a relativelylow offer will be higher than US bargainers� ratings.

H6: Acceptance rates of a relatively low offer will bemediated by individuals� importance ratings of group-based reasons.

Overview of experimental studies

In sum, we propose a model in which cultural differ-ences in bargaining behavior arise when there are rea-sons for behavior that can be used to discount theeffect of personality traits. In particular, we study differ-ences in bargaining outcomes when situational con-straints are made salient as well as when there is agroup-based decision context. Experiment 1 comparesbargaining outcomes in the form of acceptance ratesfor ultimatum offers across Korean and US responders.We expect to find significant differences between culturesin acceptance rates for relatively low offers when wemake proposers� situational constraints (i.e., totalamount available for division) salient (H1). In addition,we expect to find significant differences in ultimatumbargaining outcomes when an opponent�s behavior can


be attributed to a collective instead of directly to theindividual (H4).

We propose further that cultural differences in ulti-matum bargaining outcomes are due to differences incausal attributions for the proposer�s low offer. Linkingattributions to outcomes, Experiment 2 shows that cau-sal attributions (i.e., relative causal ratings) that differacross cultures when situational constraints are salientmediate the influence of cultural orientation on bargain-ing outcomes (H2 and H3). Similarly, Experiment 3shows that cultural differences in reasons for an oppo-nent�s behavior when the group context is salient medi-ate the influence of cultural orientation on bargainingoutcomes (H5 and H6). It is important to note that weexamine both individual trait beliefs versus contextualbeliefs (or non-social factors) as well as individual versusgroup (or other-oriented) trait beliefs in causal attribu-tions. This research thus not only examines the differen-tial role of cultural orientation on attributions to agentsversus situations but also differences in implicit theoriesof the relative influence of different agents (e.g., individ-ual versus group) in a society (e.g., Menon et al., 1999).

Experiment 1: Differences in behavioral outcomes

Method

Subjects and procedure

Participants were 133 undergraduates at a universityin the United States and 164 undergraduates at a univer-sity in Korea, all of whom participated in the study forpartial course credit. Simultaneously run in both coun-tries, subjects participated in a 2 (culture: Korea, UnitedStates) · 2 (offer size: $12.50 and $7.50) · 3 (control, sit-uational constraints salient, group decision-making con-text) between-subjects design. Participants wereinformed that they would be taking part in a pro-poser–responder game with another individual.

As participants entered the session, they were ran-domly divided into two groups. The study was con-ducted in two adjacent rooms, each group being seatedin a different room. All participants were required toread the instructions, which described the proposer–re-sponder game. Participants read that in the proposer–re-sponder game, two individuals, the proposer and theresponder have to agree on how to divide a givenamount of money between them (say $10). The proposerstarts by making an offer of $X, which is less than orequal to $10, to the responder in any way s/he choosesto do so. The responder can then either accept the offer,in which case s/he will receive $X and the proposer willget to keep the balance, $(10 – X), or the responder canreject the offer, whereupon both receive nothing. Theinstructions emphasized that the proposer can onlymake one offer while the responder can accept or reject

the offer. Subjects were told that they would receive themoney if they accepted the offer.

Because of the one-sided incomplete information sit-uation, participants were told that the amount that willbe given to the proposer to divide was determined as fol-lows. The instructions mentioned that the experimenterhad randomly drawn a ticket from a jar that had ticketsnumbered from 10 to 40. The number on the ticket thatthe experimenter randomly picked was the amount thatthe proposer has to divide. The proposer knows the ex-act amount to be divided but the responder will onlyknow that the amount could be any whole number be-tween 10 and 40 with equal probability. They were alsotold that both parties know that the proposer knows theexact amount to be divided while the responder onlyknows that the amount could assume any number be-tween 10 and 40 with equal probability.

Participants were instructed that they will be ran-domly paired with another student in their same classwho is not in the room and the identities of all individ-uals will remain anonymous before, during, and afterthe study. Participants were assigned a number whichthey were led to believe would be used to match themwith another individual. They were told that studentshad been randomly assigned the role of the proposeror responder. Based on a coin toss, students in theirroom had been assigned the role of the responder. Toenhance the credibility of the cover story, about 10 min-utes into the study, an individual was sent into each ofthe two rooms with a large brown envelope, purportedlycontaining the offers, which proposers in the other roomhad just made. Participants were given 10 minutes toread the instructions and seek clarification, if required.Pilot tests revealed no problem in the understanding ofthe task or the instructions by participants in eithercountry. After ensuring that everyone understood thegame and the task, the numbers assigned in the begin-ning were called out in random order. Participants werethen handed a sheet with their respective offers, purport-edly from another student playing the role of the pro-poser. The instructions emphasized that they couldeither accept or reject the offer. Participants had to ac-cept or reject the offer on the same sheet that containedthe assigned proposer�s offer and then respond to addi-tional measures on a separate booklet.

No communication was allowed through the dura-tion of the study that took about 30 minutes. After com-pletion, the participants were thoroughly debriefed. Thedebriefing revealed that everyone believed the coverstory that they were actually playing the proposer–re-sponder game with another individual who had beenrandomly assigned the role of the proposer.

Cross-cultural designBased on previous research, we used the US to rep-

resent an individualistic culture and Korea to represent


a collectivistic culture (e.g., Hofstede, 1991). The fun-damental premise is that because United States andKorea lie at opposite ends of the individualism-collec-tivism dimension, we assume that substantial variationin the implicit theories of behavior exists across thesecultures (e.g., Choi & Nisbett, 1998; Morris & Peng,1994). University students in the US and Korea,matched in demographic profile, were thus recruitedto participate in this study. The mean age of the partic-ipants in both countries was 21, most of the partici-pants were mainly of middle-class social status (85%in the US, 89% in Korea), and the gender distributionin both countries was similar (44% of the sample weremale in the US and 48% of the sample were male inKorea). Although the initial questionnaire was draftedin English, it was translated in Korea and back-trans-lated to ensure comparability across countries. Experi-menters in both countries were trained to reduceprocedural disparities.

Offer size manipulation

Offer size was manipulated at two levels. In the rela-tively high offer condition, the proposer�s offer was$12.50 while in the relatively low offer condition, the of-fer was $7.50. The offer sizes in Korea were adjusted forcurrency and purchase parity1 ($12.50 became 12,500wons and $7.50 became 7500 wons). The offers were pre-sented on a separate sheet of paper and were handwrit-ten to be consistent with the cover story.

In another study (Srivastava & Valenzuela, 2004), wehad pre-tested that subjects perceived an offer of $7.50as having a higher likelihood of representing an unfa-vorable share of the total amount than an offer of$12.50, given the uncertainty. The pretest, conductedwith 90 undergraduate business students, used proce-dures that were identical to those used in the controlcondition of this study. After ensuring that subjectsunderstood the proposer–responder game, subjects re-ceived their respective offers ($12.50 or $7.50) and wereasked to respond to several dependent measures. First,subjects were asked ‘‘In your estimate, how likely is itthat the offer you received is half of the total amountavailable for division?’’ (1 = Very unlikely; 9 = Verylikely). As expected, perceived likelihood of the offerbeing half of the total amount was higher in the$12.50 condition than in the $7.50 condition (M�s =4.05 and 3.02; F (1,87) = 5.49, p < .02). Second, subjectswere asked ‘‘How likely it is that the offer is less than

1 Offer sizes were pre-tested to make sure that students in bothcountries could afford to acquire similar products with these amounts.In deciding offer sizes, we also considered the psychological value of around number as well as the relative size of the offer given theuncertainty that the total amount could be any value between $10 and$40.

half of the total amount available for division?’’(1 = Very unlikely; 9 = Very likely) and ‘‘How certainare you about your assessment above?’’ (1 = Very cer-tain; 9 = Very uncertain). Subjects� perceived the likeli-hood of the offer to be less than half of the total to besignificantly higher in the $7.50 condition relative tothe $12.50 condition (M�s = 7.34 and 6.40; F (187) =6.25, p < .01). Importantly, subjects were more certainof their assessment that the offer is less than half ofthe total amount in the $7.50 versus the $12.50 condition(M�s = 3.52 and 5.03; F (1,87) = 10.97, p < .001). Thedata thus provide empirical evidence that when subjectsonly know that the total amount available could assumeany value between $10 and $40, they tend to consider$12.50 a relatively high offer and $7.50 a relatively lowoffer (less than half of the total amount).

Situational saliency manipulation

In the salient situational constraints condition, beforereceiving the offers, participants were asked to imaginethemselves as playing the role of a proposer who has$10, $20, and $40 dollars available to share with a re-sponder. They were asked to write down the amountthat they would offer a responder in all three situationsand predict whether a responder would accept these of-fers or not. Participants had to complete this task beforeresponding to the dependent measures. The idea is thatresponders would allocate much more cognitive effort inevaluating possible scenarios in the game and moreimportantly this procedure should raise the salience ofsituational constraints in assessing the proposer�s offer(Choi & Nisbett, 1998; Krull et al., 1999). In the controlcondition, participants did not complete this task andwere asked to respond to the dependent measures afterobtaining their offers.

We pre-tested the efficacy of the situational saliencymanipulation. We expect all subjects to believe that thereis a stronger link between the total amount of moneyavailable and the proposer�s offer when the situationalconstraints are made salient relative to the control con-dition. Specifically, 35 US and 37 Korean subjects inthe situational salient and control condition were askedto respond to the question ‘‘In this game, how likely is itthat the proposer will typically offer less money when thepot of money is lower’’ (1 = Very unlikely, 9 = Verylikely). Overall, subjects believed that the proposer�s of-fer and total pot of money are more likely to be relatedin the situational salient condition than in the controlcondition (M�s = 6.54 and 4.97; F (1,70) = 10.53,p < 0.01). Importantly, the perceived likelihood of thelink between offer and total amount was significantlyhigher in the salient situational constraint condition rel-ative to the control condition for both East Asian(M�s = 6.41 and 4.70; F (1,33) = 4.95, p < 0.03) and USresponders (M�s = 6.65 and 5.43; F (1,35) = 3.59,p < 0.05).

Fig. 1. Experiment 1. Acceptance rates. (A) Situational constraints notsalient (control). (B) Situational constraints salient.


Group decision context

In the group decision-making condition, responderswere informed that the individual assigned the role ofthe proposer had to make decisions within a group con-text. Responders were told that the proposer had towork in groups of six and the proposer had to work withhis/her group to reach a consensus on the amount thats/he could offer the responder. Only the proposer wouldbenefit/suffer from the responder�s decision. The deci-sion context was intended to allow attributions ofbehavior to the group.

A pre-test assessed the efficacy of the group saliencymanipulation. Subjects (36 US and 30 Korean) in bothgroup salient and control conditions were asked ‘‘Didthe proposer have to work with a group to decide howmuch money to offer?’’ Overall, 78% of the subjects con-firmed that the proposer had to work with a group in thegroup salient condition relative to 24% of the subjects inthe control condition (v2 (1) = 18.57, p < .001). Theanalysis conducted for each culture shows that a signif-icantly higher percentage of the subjects confirmed thatthe proposer had to work with a group in the group sali-ent than in the control condition for both Korean (75and 40%; v2 (1) = 3.52, p < .06) and US respondents(81 and 13%; v2 (1) = 16.05, p < .001).

Results

Checks

Both Korean and American students were asked tocomplete Triandis�s (1995) individualism (Cronbach�sa = .68) and collectivism (Cronbach�s a = .72) scales.Following Triandis (1995), the scores were standardizedfor each participant. Consistent with the literature thatindividualism and collectivism are not really bipolar(e.g., Triandis, 1995; Krull et al., 1999), the correlationbetween individualism and collectivism was not signifi-cant (r = .18, ns). As an assessment of participant�s rel-ative individualism versus collectivism, we subtractedthe collectivism score from the individualism score suchthat positive numbers indicate more individualism thancollectivism (Triandis, 1995; Choi et al., 1999). As ex-pected, the mean difference score was significantly high-er for the US participants relative to the Koreanparticipants (M�s = 1.40 and �.30, F (1,296) = 382.01,p < .0001), suggesting that the US participants are moreindividualistic than their Korean counterparts.

Hypothesis 1. Participants� decision to accept or rejectwas analyzed using a log-linear model as a function ofthe independent variables and their interactions. H1suggests that when situational constraints are not salient(control condition), there should be no difference in theacceptance rates across US and Korean bargainers.However, when situational constraints are made salient,Korean bargainers should be more sensitive to the

situational determinants of a proposer�s low offer andtherefore the acceptance rates should be higher forKorean bargainers than US bargainers.

Fig. 1 displays the acceptance rates of both US andKorean responders in the high and low offer conditionswhen the situational constraints are not salient (Fig. 1A)and when they are salient (Fig. 1B). Not surprisingly,acceptance rates were significantly higher in the $12.50offer condition relative to the $7.50 offer condition forboth US (63.3 and 39.1%; v2 (1) = 5.64, p < .02) andKorean bargainers (65.7 and 33.3%; v2 (1) = 5.41,p < .05).

Importantly, in the control condition where the situ-ational constraint was not salient, acceptance ratesacross Korean and American respondents in both the$12.50 (65.7 and 63.3%; v2 (1) = .12, ns) and $7.50(33.3 and 39.1%; v2 (1) = .40, ns) did not varysignificantly.

When the situational constraint was made salient,consistent with H1, acceptance rates for the relativelylow $7.50 offer were significantly higher for Koreanresponders than US responders (55.2 and 40.9%;v2 (1) = 2.75, p < .05). There was, however, no differenceacross the Korean and US responders in the $12.50 con-dition (73.3 and 76%; v2 (1) = .18, ns). Follow-up analy-ses show that for the relative low $7.50 offer, theacceptance rates for Korean responders were signifi-cantly higher when the situational constraints weremade salient relative to when they were not (55.2 and33.3%; v2 (1) = 2.86, p < .05) but the saliency of situa-tional constraints did not affect the acceptance rates ofthe US responders (40.9 and 39.1%; v2 (1) = .02, ns).These data support H1.

Fig. 2. Experiment 1. Acceptance rates. (A) Individual context(control). (B) Group decision context.


Hypothesis 4. Fig. 2 displays the acceptance rates acrossthe two cultures. As described before, there was nodifference in the bargaining outcomes across Korean andUS responders in the control condition, which representsan individual decision making context. Specifically,responders in both cultures rejected more than half ofthe $7.50 offers thereby denying profits to either party.

However, in contrast to the control condition, whenproposers were making the decision in a group context,the acceptance rates for the low $7.50 offer were signif-icantly higher for Korean bargainers relative to US bar-gainers (60 and 25.4%; v2 (1) = 3.80, p < .05). Theacceptance rates for the $7.50 offer were also signifi-cantly higher for Korean bargainers in the group deci-sion context relative to the control, individual decisioncontext (60 and 33.3%; v2 (1) = 3.80, p < .05). Addi-tional analyses reveal that in the group decision context,the acceptance rates were not significantly higher in the$12.50 offer condition relative to the $7.50 offer condi-tion for Korean responders (67.1 and 60%;v2 (1) = .20, ns). On the other hand, for US responders,the acceptance rates were significantly higher in the$12.50 offer condition relative to the $7.50 offer condi-tion (64.5 and 25.5%; v2 (1) = 4.23, p < .05). Consistentwith H4, the data clearly shows that the reactions tothe offers of bargainers in collectivistic cultures are notas extreme as those of bargainers in individualistic cul-tures in a group decision context.

Discussion

Experiment 1 demonstrates that when certain condi-tions arise that allow for the manifestation of cultural

differences, bargaining outcomes change significantlyacross countries. First of all, responders� reactions torelatively low offers change across cultures if situa-tional constraints are made salient. Specifically, we ob-served no differences in acceptance rates in anincomplete bargaining situation in the control condi-tion suggesting that ultimatum bargainers in both USand Korea are prone to making the fundamental attri-bution error. As such, both US and Korean respondersreacted similarly to the low $7.50 offer and more thanhalf of these offers were rejected. The difference be-tween US and Korean responders is however broughtto the fore when the situational constraints are madesalient. While US responders reacted similarly to therelatively low $7.50 offers as in the control condition,Korean responders accepted more than half of theseoffers. Consistent with previous findings (e.g., Choi &Nisbett, 1998), participants from East Asian culturesrelative to Western cultures seem to be able to adjusttheir attributions to account for the possibility of situ-ational constraints that may have dictated a bargainingopponent�s behavior.

Experiment 1 also shows that cultural differences inbehavioral outcomes between bargainers in Korea andthe US emerge when decisions are made in a group con-text. While US responders do not change their accep-tance behavior regardless of whether proposers�decisions are made individually or in a group, Koreanresponders accepted relatively low $7.50 offers signifi-cantly more often when proposers� decisions wereframed in a group decision-making context. These re-sults support the idea that implicit theories of the roleof the group in explaining individuals� behavior varyacross the two cultures (Menon et al., 1999). These the-ories influence decision making when they are activatedas a guide to the interpretation of behavior (Morris &Fu, 2001). Specifically, East Asian cultures seem morelikely to give the benefit of the doubt to an individualwhen constrained by group directives whereas Ameri-cans are not as sensitive to the group context.

Experiment 2: Situational attributions as mediators

Experiment 1 demonstrates the differences in bargain-ing outcomes across Korean and US participants whensituational constraints are made salient. These differ-ences in bargaining outcomes are presumed to be dueto the differential causal inferences across cultures.While East Asians are more sensitive to situationaldeterminants of behavior and tend to correct the extentof causal attribution to personality characteristics whenthe constraints are made salient, Westerners continue toattribute behavior to personality traits regardless of thesaliency of the situational constraints (Choi & Nisbett,1998; Krull et al., 1999). Experiment 2, designed to test


H2 and H3, provides a more direct link between causalattributions and bargaining outcomes.

Method

Subjects and procedureOne hundred and one undergraduates (42 US stu-

dents and 59 Korean students) participated in a 2 (cul-ture: US and Korea) · 2 (situational constraints: notsalient and salient) between-subjects design. Participantswere randomly assigned to the second factor. Offer sizewas always $7.50 (or 7500 wons).2 The experimentalmanipulations and procedures used were identical tothose used in Experiment 1. After ensuring that partici-pants understood the proposer–responder game, partic-ipants received their respective offer and indicated theirdecision to accept or reject. They were then asked toevaluate the underlying reasons for the offer.

Subjects were asked to evaluate the underlying rea-sons for the proposer�s offer: ‘‘The total amount ofmoney that was available to the proposer to split be-tween you and him/her,’’ ‘‘The proposer�s competitive-ness,’’ and filler causes such as ‘‘The proposer did notunderstand the game well’’ and ‘‘The proposer did notput forth enough effort’’ (1 = Not a cause at all;7 = Very important cause). Because of the differencesin use of the scale across US and Korean participants(US participants tended to use the high end of the scaleand Korean participants tended to use the mid-points ofthe scale), as an assessment of how much participantsfocus on the total amount of money available (situa-tional constraint) more than the other reasons forbehavior, we subtracted each participants� mean ratingon the relevant causes from their ratings of total amountof money available as a cause for the offer (Triandis,1995). Positive numbers indicate that participants be-lieved that the total amount available was a more impor-tant cause for the offer than the other causes. Thesechange scores provide a more meaningful measure ofparticipants� causal beliefs as a function of the manipu-lation, since they adjust for participants� response styles(e.g., Gilovich, 1981; Triandis, 1995).

2 We also collected ratings for $12.50 offers. Overall, ratings ofrelative importance of situation-based reasons for a $12.50 offer arehigher than when the offer is $7.50 (1.20 for $12.50 and 0.74 for $7.50,F (1,139) = 2.09, p < .10). They do not differ across conditions (1.02and 0.96 respectively for the situational saliency and control conditionsF (1,68) = 0.77, p > .70) or across cultures (1.10 and 0.95 respectivelyfor Westerners and East Asians, F (1,68) = .20, p > .60). However,high relative ratings of situation-based reasons did not affect accep-tance rates (53.1% compared to 44.7% respectively for high and lowrelative ratings, v2 (1) = 0.49, p > .40).

Results

Hypothesis 3. As expected, the mean differential ratingsfavoring situation-based cause (Fig. 3) was positive andsignificantly higher for Korean students when thesituational constraints were salient relative to whennot salient (M�s = 1.3 and. 74, F (1,57) = 2.51, p < .05).In contrast, the differential ratings did not vary signif-icantly for US students across the salient and not salientconditions (M�s = .70 and. 82, F (1,40) = .66, ns).Importantly, consistent with H2, the ratings in thesituational salient condition were significantly higher forKorean students relative to US students (M�s = 1.3 and.7; F (1,47) = 2.92, p < .05) but not when the situationalconstraint was not salient.

Hypothesis 4. We tested the mediating role of causalattributions on the effect of cultural orientation on bar-gaining outcomes using a series of logistic regressions(see Baron & Kenny, 1986). In the situational constraintsalient condition, consistent with Experiment 1, theacceptance rates were significantly higher for Koreanbargainers than for US bargainers (58.8 and 35.7%,v2 (1) = 3.33, p < .05). Further, the mean differencescore favoring situational-based reasons for the lowoffer was significantly higher for Korean bargainersthan for US bargainers (M�s = 1.30 and. 70;F (1,47) = 2.92, p < .05). When both causal differenceratings and culture are included as explanatory variablesin the logistic regression, causal ratings have a signifi-cant effect on acceptance rates (v2 (1) = 4.97, p < .05),but the effect of culture on acceptance rates is notsignificant (v2(1) = 1.137, p > .20). These results supportH3.

Discussion

Experiment 2 provides additional support for the ideathat Korean bargainers are more likely to be sensitive tosituational constraints when the constraints are madesalient. The results also show that differences in causalattributions when the situational constraints are made

Fig. 3. Experiment 2. Ratings situation-based reasons.

3 In terms of ratings for $12.50 offers, relative importance of group-based reasons for a $12.50 offer are higher than when the offer is $7.50(�0.37 and �0.76; F (1,102) = 2.47, p < .10). They do not differ acrossthe control and group context conditions (�.31 and �0.45; F (1,48) = 0.16, p > .60) or across Korean and US responders (�0.26 and�0.62; F (1,48) = 0.95, p > .30). However, high relative ratings ofgroup-based reasons do not have an effect on acceptance rates (44.4%compared to 34.4% respectively for high and low relative ratings,v2 (1) = 0.50, p > .40).

4 There is no mention of a group decision-making context in thecontrol condition. However, we use the same question in the controlcondition to measure other possible group-based attributions such asbehavior caused by mere group identity (i.e., accepting smaller offersfrom an in-group or a suspected in-group). Any increase in ratingsfrom the control condition to the group condition should be caused bythe increased saliency of the group decision-making context.


salient fully mediate the effect of culture on bargainingoutcomes. The implication is that Korean bargainersare more likely to correct for the fundamental attribu-tion error than their Western counterparts, when infor-mation is accessible to discount the personalityattributions. Such correction is what drives the changein acceptance rates or, more generally, behavioraloutcomes.

Together, the findings of Experiments 1 and 2 areconsistent with the notion that the implicit theories thatdiffer across cultures are activated when situational con-straints are made salient. These theories are then used inattributing an opponent�s behavior to predominantlypersonality traits or situational constraints. While Wes-tern cultures emphasize that behavior is caused by stableinternal personality traits, East Asian cultures are morelikely to draw upon their situation-centered theory thatbehavior is shaped by situational factors. As such, ulti-matum bargainers in East Asian cultures are more likelyto give the benefit of the doubt to opponents and recog-nize that an opponent�s low offer could be due to theamount available for division.

Although our operationalization of situational sal-iency in Experiments 1 and 2 has been used previously(e.g. Choi & Nisbett, 1998), a potential explanation ofour findings is that when asked to put themselves inthe shoes of the proposer, there may be a difference inthe ability to take the perspective of in-group others be-tween individualistic and collectivistic cultures. Specifi-cally, individuals in collectivistic cultures may be moreable to take the perspective of the in-group other andthereby account for the situational constraints better.In addition, people in individualistic cultures may notdepend as much on external rewards as collectivistic cul-tures (Leung et al., 2002), or may be less drawn to thestatus quo (Briley et al., 2000), or less motivated toavoid conflict (Leung, 1987; Leung & Tjosvold, 1998).These differences between individualistic and collectivis-tic cultures may explain our findings. We rule out thesealternative explanations by showing that in Experiment1, the collectivism (or individualism) scores do not medi-ate the differences in behavioral outcomes between thetwo cultures. By doing so, we provide more supportfor a dynamic rather than a dispositional model of cul-tural influence.

Using data from Experiment 1, we tested the mediat-ing role of collectivist values on the effect of cultural ori-entation on bargaining outcomes using a series oflogistic regressions. As discussed in Experiment 1, whenthe situational constraint was salient, the acceptancerates for the relatively low $7.50 offer were significantlyhigher for Korean than for US responders (55.2 and40.9%, v2 (1) = 2.75, p < .05). Further, the mean differ-ence score of these participants relative individualisticversus collectivistic values (higher difference score =more agreement with individualistic values) was

significantly higher for US than for Korean responders(M�s = 1.51 and �.17; F (1,50) = 73.79, p < .0001).However, when both the individualism scores and cul-ture are included as explanatory variables in the logisticregression, the individualism scores do not have a signif-icant effect on acceptance rates (v2 (1) = 0.232, p > .60),which is a requirement for full or partial mediation.Accordingly, we rule out alternative explanations.

Experiment 3: Group-based attributions as mediators

Experiment 1 clearly demonstrated the differences inbehavioral outcomes across Korean and US responderswhen decisions are made in a group context. Our con-tention is that the observed differences are due to cul-tural divergence in assigning causality to eitherindividuals or groups. Experiment 3 provides a test ofH5 and H6 by examining the differences in causal infer-ences that underlie behavioral outcomes across cultureswhen the proposer is making the decision of how muchto offer in a group context.

Method

Subjects and procedure

Ninety-one undergraduate students (33 in US and 58in Korea) participated in a 2 (culture: United States andKorea) · 2 (conditions: group context and individualcontext) between-subjects design. The experimentalmanipulations and procedures used were identical tothose used in Experiment 1 in the group decision con-text, except that the offer was always $7.50 (7500 wons).3

After ensuring that participants understood the pro-poser–responder game, participants received theirrespective offers, accepted or rejected their offer, andwere asked to evaluate the underlying reasons for theproposer�s offer: ‘‘The influence of the group on the pro-poser�s decision,4’’ ‘‘The total amount of money thatwas available to the proposer to split between you andhim/her,’’ ‘‘The proposer�s competitiveness,’’ and filler


causes such as ‘‘The proposer did not understand thegame well,’’ ‘‘The proposer did not put forth enough ef-fort’’ (1 = Not a cause at all; 7 = Very important cause).As in Experiment 2, as an assessment of how much par-ticipants focused on group influence on the proposer�sdecision more than any other reason for behavior, wesubtracted each participants� mean rating on the rele-vant causes from their rating of group as an importantcause for the low offer.

Results

Hypothesis 5. Consistent with H5, the mean differentialrating for group as the underlying cause (see Fig. 4) washigher for Koreans in the group context versus thecontrol condition (M�s = �.17 and �.72;F (1,56) = 3.07, p < .05). However, the mean rating didnot vary significantly for US bargainers across the groupcontext and control conditions (M�s = �1.20 and �.64;F (1,31) = 1.23, ns). Importantly, the scores were signif-icantly higher for Korean bargainers relative to USbargainers in the group context (M�s = �.17 and �1.20;F (1,37) = 5.58, p < .02) but not in the control condition.

Hypothesis 6. We tested the mediating role of causalattributions on the effect of cultural orientation on bar-gaining outcomes using a series of logistic regressions.As in Experiment 1, when the proposer�s decision ismade in a group context, culture has a significant effecton bargaining outcomes such that acceptance rates aresignificantly higher for Korean responders relative toUS responders (60 and 31%; v2 (1) = 5.06, p < .02). Inthe group decision context, the mean difference scorefavoring group as a reason for the low offer are also sig-nificantly higher for Korean bargainers relative toAmerican bargainers (M�s = �.17 and �1.20;F (1,37) = 5.58, p < .02). However, when both causaldifference ratings and culture are included as explana-tory variables in the logistic regression, causal ratingshave a significant effect on acceptance rates(v2 (1) = 5.96, p < .02), and the effect of culture onacceptance rates is not significant (v2 (1) = .51, ns).

Fig. 4. Experiment 3. Ratings group-based reasons.

Discussion

These results show that the differences in causal attri-butions to group influences fully mediate the differencesin behavioral outcomes between Korean and US bar-gainers, when the decision is made in a group context.Said differently, attribution of causality to differentagents appears to explain the difference in ultimatumbargaining outcomes across Korean and US bargainers.However, as outlined earlier, subjects� differential abilityto take the perspective of in-group others may also ac-count for the differences in behavioral outcomes be-tween cultures when a decision is made in a groupcontext. Using data from Experiment 1, we testedwhether collectivism scores mediate the effect of culturalorientation on bargaining outcomes. In Experiment 1, inthe group context, acceptance rates for the relatively low$7.50 offer were significantly higher for Korean respond-ers relative to US responders (60 and 24.4%;v2 (1) = 3.80, p < .05). Further, the mean difference scoreof these participants (individualistic – collectivistic;higher difference score implies more agreement withindividualistic values) was significantly higher for USbargainers relative to Korean bargainers (M�s = 1.25and �.68; F (1,22) = 47.21, p < .0001). When both indi-vidualism score and culture are included as explanatoryvariables in the logistic regression, the individualismscores do not have a significant effect on acceptancerates (v2 (1) = 1.65, p > .20). The alternative explanationthat individuals from collectivistic cultures are betterable to take the perspective of the in-group other istherefore ruled out.

In sum, Experiment 3 provides strong support for thecontention that in a group decision context, Korean bar-gainers are less likely to assign causality to personalitytraits and thus are more likely to accept unfavorable of-fers. In contrast, US bargainers are prone to assigningcausality to personality traits and thus exhibit no differ-ence in acceptance behavior even when other possiblereasons for such behavior are available or accessible.These findings are consistent with the dynamic view ofculture where the role of cultural orientation is contextu-ally driven such that cultural differences are influentialonly when people are required to draw upon the implicittheories that differ across cultures (Briley et al., 2000;Choi & Nisbett, 1998). Accordingly, in our experiments,only Koreans were sensitive to the constraints that bothsituations and groups pose on behavior. However, suchknowledge structures need to be activated or madeaccessible for use in the decision processes.

General discussion

Spurred by the economic and theoretical importanceof cross-cultural business interactions in an organiza-


tional context, the purpose of this research was to exam-ine how differences in cultural orientation influence cau-sal attributions and thereby transactional outcomes inan incomplete information situation. This is particularlyimportant because most business transactions are char-acterized by incomplete information and causal attribu-tions for an opponent�s behavior play a particularlyimportant role in these situations. Although social psy-chologists argue that interpersonal conflicts are particu-larly likely to lead to the fundamental attribution erroror the tendency to interpret and evaluate an opponent�sbehavior in terms of personality traits rather than situa-tional constraints, the question as to when such an attri-butional process generalizes to other cultures remainslargely unanswered. This research thus examines theinfluence of cultural orientation on participants� sensi-tivity to internal personality traits versus external con-straints in their causal attributions, which in turn,affect their decisions and outcomes. Towards this objec-tive, we use ultimatum bargaining as a means of investi-gating these effects. We consider the ultimatumbargaining game as an appropriate context because cul-turally imposed beliefs, which influence player�s percep-tions of what constitutes a reasonable offer, have beenfound to lead to rejection of offered amounts sometimesas large as someone�s monthly salary (e.g., Henrichet al., 2001). However, most of this literature focuseson complete information situations where bargainershave full information about one another�s payoffs.

Three experiments are reported that were conductedsimultaneously in the US and Korea. Experiment 1showed that in an incomplete information setting, therewas no difference in behavioral outcomes across Wes-tern and East Asian cultures as ultimatum bargainersfrom both US and Korea were equally likely to attributea low offer to opponent�s personality traits. However,when situational constraints were made salient, Koreanbargainers were more likely to accept a low offer relativeto US bargainers. Experiment 2 links the causal attribu-tions to the behavior outcomes and, consistent with ear-lier work (e.g., Choi & Nisbett, 1998), our resultssuggest that East Asian cultures are more likely to cor-rect for the fundamental attribution error when situa-tional constraints are made salient. In other words,when there is no accessible information to discount per-sonality trait attributions, both Western and East Asiancultures are prone to the fundamental attribution error.These results also suggest that cultures perhaps differ intheir implicit theories of the influence of external con-straints rather than in their implicit theories of personal-ity dispositions (Chiu et al., 2000; Krull et al., 1999).

Based on the assumption that the group is a naturalunit of agency for East Asians and the individual isthe unit of agency for Western cultures (Choi et al.,1999; Menon et al., 1999), Experiment 1 also supportsthat Korean bargainers are more sensitive to the poten-

tial group influence in affecting an opponent�s behavior.Specifically, while there was no difference in bargainingoutcomes between US and Korean bargainers in anindividual decision context, Korean bargainers weremore likely to accept low offers when the proposer wassupposedly making the decision in a group context. Itis important to note that groups occupy the physicalspace surrounding individual actors and, in some cases,can be considered merely an additional situational con-straint (Menon et al., 1999). However, groups may alsobe more than just a passive non-social situational con-straint. Rather, the group may be perceived as a socialentity that can direct action. As with individual disposi-tions, groups may also be perceived as possessing uniquedispositions (Higgins & Bryant, 1982). Experiment 3unambiguously shows that differences in causal attribu-tions to group dispositions mediate the influence of cul-tural orientation on bargaining outcomes when thegroup context is salient.

Together, the results indicate that there are boundaryconditions to the influence of culture on judgment anddecision-making. In other words, there are certain con-ditions under which cultural differences are manifestedsupporting the more recent, dynamic view of the influ-ence of cultural orientation (Briley et al., 2000; Choi &Nisbett, 1998). In our context, penalization of apparentcompetitive (or unfair) intent seems to be a universalphenomenon as ultimatum bargainers from both USand Korea exhibit the fundamental attribution error.However, when there is accessible information thatcan be used to discount the personality based attribu-tions, cultural difference emerge. In particular, Koreanbargainers are more likely to recognize and acknowledgealternative reasons for observed behavior, which tend tocorrect the initial tendency to seek causal explanationsin terms of personality dispositions.

Implications for bargaining theory and practice

Given that research exploring the psychological pro-cesses that create behavioral mismatches or misunder-standings is still in its infancy, our findings not onlyhave implications for cross-cultural bargaining but alsofor cross-cultural social interactions in general. It is clearthat inferential sources of misunderstanding could occurevery time two individuals from different cultures try toexplain opponents� behavior under incomplete informa-tion. An opponent�s behavior can be attributed to thesituation, to the personality of the individual, or to thecollective forces potentially influencing the individual.In incomplete information situations, unwarranted per-ceptions of unfairness and competitiveness could lead tosub-optimal agreements, disagreements, or impasses.This research examines whether there is preference forexplanations in terms of personality dispositions in Wes-


tern cultures relative to East Asian cultures and impor-tantly the conditions under which cultural differences incausal attributions are manifested. Although the ten-dency for causal attributions to be based on personalitydispositions seems to be universal, we identify the condi-tions under which East Asians are more likely thanWesterns to correct their dispositional tendencies. Theseattributional corrections may enhance the cooperativestance of East Asians resulting in potentially more effi-cient outcomes, fewer disagreements and impasses(e.g., Gelfand et al., 2002; Morris et al., 2004). Improv-ing the level of cooperativeness benefits both parties.Our results clearly suggest that the cultural context inwhich the interpersonal conflict is embedded plays animportant role in determining resolution outcomes,and directing attention to the conditions that facilitateattributional corrections, at least in East Asian cultures,would render judgments and decisions less prone to theoverattribution bias, thereby, improving final outcomes.

Implications for cross-cultural research

There are many mechanisms by which cultural orien-tation may affect bargaining outcomes. As discussedearlier, cultural differences in subjects� rights versus dutyorientation (Hong, Ip, Chiu, Morris, & Menon, 2001) orin the beliefs about fairness and justice (Morris, Larricket al., 1999; Morris, Leung et al., 1999) may affect re-sponses to ultimatum bargaining. Additionally, individ-ualistic societies such as the US may not depend asmuch on external rewards as collectivistic societies(Leung et al., 2002) and, therefore, may be more com-fortable rejecting a relatively small offer. A collectivisticorientation may also produce more agreements to posi-tive, albeit small offers, because of a higher likelihood tocompromise (Briley et al., 2000). Finally, collectivistsocieties may accept relatively small offers because oftheir stronger motivation to avoid conflict (Leung,1987; Leung & Tjosvold, 1998). These mechanismsmay concurrently affect differences in ultimatum bar-gaining outcomes. In this paper, we test the hypothesisthat differences in ultimatum outcomes are due to cul-tural differences in the tendency to attribute ambiguousbehavior to personality traits.

Our results are worth discussing in the context of theprevailing model of social inference. The cross-culturalpsychology literature suggests that attribution styles re-sults from culturally conferred schemas or knowledgestructures (e.g., Gelfand & Christakopoulou, 1999;Morris et al., 2004). Such schemas may take the formof either declarative knowledge or procedural knowl-edge. In other words, differences in attribution stylesmay be due to culturally specific beliefs about the rela-tive importance of possible reasons for behavior (e.g.,Church et al., 2003; Leung et al., 2002) or to differences

in the process by which information is encoded and ad-justed to judge the cause for behavior (e.g., Knowleset al., 2001; Krull et al., 1999). Disentangling the sourcesbehind our reported differences in attributional patternsis offered as a direction for future research.

Our findings are, however, consistent with a two-stage model in which behavior is first interpreted draw-ing an initial inference which is later revised to arrive atthe final judgment (e.g., Krull et al., 1999). Our resultssuggest that both cultures draw inferences initially basedon personality dispositions and fail to revise their initialinferences. However, when information is salient that al-lows discounting of the personality dispositions, EastAsian cultures are better able to revise their initial infer-ences to take into account situational constraints. Inother words, in East Asian cultures, there seems to bea dual system of cultural beliefs entailing both auto-matic reasons, which may be revised, and interpretativereasons, which seem to require cognitive resources or, atleast, a very explicit applicability to the stimuli. In con-trast, Western cultures have a belief system that makes itdifficult to incorporate situational reasons for behaviorboth at the initial (more automatic) level and at theinterpretational level.

Our findings also qualify the automatized situationalcorrection model of person perception (Knowles et al.,2001). This model posits that whereas automatic dispo-sitional inference is universal, situational correction maybe relatively resource-independent for holders of a situ-ation-based theory (i.e., for East Asians). Our results areconsistent with the model that in the initial stage, Kor-ean and US subjects behave similarly. However, eventhough Koreans could be deploying certain extent ofautomatic situational correction, our findings suggestthat cognitive and motivational factors, such as the pres-ence and/or need to explain behavior, may even furtherenhance the situational correction. Future researchshould focus on the conditions under which cultures dif-fer in the level of automatic versus controlled processingof the causal attributions and inferences drawn based onobserved behavior.

Limitations and future research

Although a major attraction of ultimatum bargainingis that it allows isolation of the factors of interest, a lim-itation of our studies is that actual bargaining may in-volve multiple rounds and these rounds provideinformation that is invaluable in decision making. Not-withstanding the simplicity of ultimatum bargaining,this research is perhaps one of few that examinescross-cultural differences in incomplete information bar-gaining. This research thus provides the backdrop forfuture cultural inquiries into interpersonal conflict out-comes from both a practical and theoretical perspective.


As discussed earlier in the paper, cultural differencesin causal attributions may be due to both differences inbeliefs of what is a relevant cause for behavior and differ-ences in the inferential processes. This is an important to-pic for future research to provide a better understandingof the underlying mechanism of social inferences. Futureresearch should address the question of whether culturesdiffer just in the content of their social judgments and/orin the process by which these judgments are made. Fu-ture research should also examine how cultural patternsof attributions for group-based acts depend on the typeof group under consideration. For example, the litera-ture differentiates between intimacy groups, task groups,or temporary loose associations (Yzerbyt, Rogier, &Fiske, 1998). Dispositions to groups should be strongerin the case of groups that are more of an entity (i.e., inti-macy group and task group). We suggest conducting amore systematic study of group dispositions to furtherexamine their prevalence both within Western and EastAsian causal reasoning.

References

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediatordistinction in social psychological research: Conceptual, strategic,and statistical considerations. Journal of Personality and Social

Psychology, 51(6), 1173–1182.Blount, S. (1995). When social outcomes aren�t fair: The effect of

causal attributions on preferences. Organizational Behavior and

Human Decision Processes, 63(August), 131–144.Bolton, G., & Zwick, R. (1995). Anonymity versus punishment in

ultimatum bargaining. Games and Economic Behavior, 10, 95–121.Briley, D., Morris, M. W., & Simonson, I. (2000). Reasons as carriers

of culture: Dynamic versus dispositional models of culturalinfluence on decision making. Journal of Consumer Research,

27(September), 157–178.Buchan, N. R., Croson, R. T. A., & Johnson, E. J. (2004). When do

fair beliefs influence bargaining behavior. Experimental bargainingin Japan and the United States. Journal of Consumer Research,

31(June), 181–190.Camerer, C., & Thaler, R. H. (1995). Anomalies: Ultimatums,

dictators, and manners. Journal of Economic Perspectives, 9(2),209–219.

Chen, Y., Mannix, E. A., & Okumura, T. (2003). The importance ofwho you meet: Effects of self- versus other- concerns amongnegotiators in the United States, the People�s Republic of China,and Japan. Journal of Experimental Social Psychology, 39(Janu-ary), 1–15.

Chiu, C., Morris, M. W., Menon, T., & Hong, Y. (2000). Motivatedcultural cognition: The impact of implicit cultural theories ondispositional attribution varies as a function of need for closure.Journal of Personality and Social Psychology, 78(2), 247–259.

Choi, I., & Nisbett, R. E. (1998). Situational salience and culturaldifferences in the correspondence bias and actor-observer bias.Personality and Social Psychology Bulletin, 24(9), 949–960.

Choi, I., Nisbett, R. E., & Norenzayan, A. (1999). Causal attributionacross cultures: Variation and universality. Psychological Bulletin,125(1), 47–63.

Church, T. A., Ortiz, F. A., Katigbak, M. S., Avdeyeva, T. V.,Emerson, A. M., Vargas Flores, J., & Ibanez Reyes, J. (2003).Measuring individual and cultural differences in implicit trait

theories. Journal of Personality and Social Psychology, 85(2),332–347.

Croson, R. T. A. (1996). Information in ultimatum games: Anexperimental study. Journal of Economic Behavior and Organiza-

tion, 30(August), 197–212.Gelfand, M. J., Higgins, M., Nishii, L. H., Raver, J. L., Dominguez,

A., Murakami, F., Yamaguchi, S., & Toyoma, M. (2002). Cultureand egocentric perceptions of fairness in conflict and negotiation.Journal of Applied Psychology, 87(October), 833–845.

Gelfand, M. J., & Christakopoulou, S. (1999). Culture and negotiatorcognition: Judgment accuracy and negotiation processes in indi-vidualistic and collectivistic cultures. Organizational Behavior and

Human Decision Processes, 79(3), 248–269.Gilovich, T. (1981). Seeing the past in the present: The effect of

associations to familiar events on judgments and decisions. Journalof Personality and Social Psychology, 40(May), 797–808.

Graham, J. L., Kim, D. K., Lin, C., & Robinson, M. (1988). Buyer–seller negotiations around the Pacific Rim: Differences in funda-mental exchange processes. Journal of Consumer Research,

15(June), 48–54.Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., &

McElreath, R. (2001). In Search of Homo Economicus: Behavioralexperiments in 15 small-scale societies. American Economic Review,

91(May), 73–78.Higgins, T. E., & Bryant, S. L. (1982). Consensus information and

fundamental attribution error: The role of development and in-group versus out-group knowledge. Journal of Personality and

Social Psychology, 43(5), 889–900.Higgins, T. E. (1996). Knowledge activation: accessibility, applicabil-

ity, and salience. In Tory Higgins & A. W. Kruglanski (Eds.),Social Psychology: Handbook of Basic Principles (pp. 133–168).New York: Guilford.

Hofstede, G. (1991). Cultures and organizations: Software of the mind.London: McGraw-Hill.

Hong, Y., Ip, G., Chiu, C., Morris, M., & Menon, T. (2001). Culturalidentity and dynamic construction of the self: Collective duties andindividual rights in Chinese and American Cultures. Social

Cognition, 19(3), 251–268.Knowles, E. D., Morris, M. W., & Chiu, C. (2001). Culture and the

process of person perception: Evidence for automaticity amongEast Asians in correcting for situational influences onbehavior. Personality and Social Psychology Bulletin, 27(10),1344–1356.

Krull, D. S., Loy, M. H., Lin, J., Wang, C., Chen, S., & Zhao, X.(1999). The fundamental attribution error: Correpondence bias inindividualistic and collectivistic cultures. Personality and Social

Psychology Bulletin, 25(10), 1208–1219.Leung, K., Bond, M. H., Reimel de Carrasquel, S., Munoz, C.,

Hernandez, M., Murakami, F., Yamaguchi, S., Bierbrauer, G., &Singelis, T. M. (2002). Social axioms: The search for universaldimensions of general beliefs about how the world functions.Journal of Cross-Cultural Psychology, 33, 286–302.

Leung, K., & Tjosvold, D. W. (Eds.). (1998). Conflict management in

Asia Pacific rim. Singapore: Wiley.Leung, K. (1987). Some determinants of reaction to procedural models

for conflict resolution: A cross-national study. Journal of Person-ality and Social Psychology, 53, 898–908.

Markus, H., & Kitayama, S. (1991). Culture and the self: Implicationsfor cognition, emotion and motivation. Psychological Review,

98(April), 224–253.Menon, T., Morris, M. W., Chiu, C., & Hong, Y. (1999). Culture and

the construal of agency: Attribution to individual versus groupdispositions. Journal of Personality and Social Psychology, 76(5),701–717.

Morris, M. W., & Peng, K. (1994). Culture and cause: American andChinese attributions for social and physical events. Journal of

Personality and Social Psychology, 67, 949–971.


Morris, M. W., Larrick, R. P., & Su, S. K. (1999). Misperceivingnegotiation counterparts: When situationally determined bargain-ing behaviors are attributed to personality traits. Journal of

Personality and Social Psychology, 77(1), 52–67.Morris, M. W., Leung, K., & Ames, D. (1999). View from the inside

and outside: Integrating emic and etic insights about culture andjustice judgment. Academy of Management Review, 24(4),781–796.

Morris, M. W., & Fu, H. (2001). How does culture influence conflictresolution? A dynamic constructivist analysis. Social Cognition,

19(July), 324–349.Morris, M. W., Leung, K., & Iyengar, S. S. (2004). Person

perception in the heat of conflict: Negative trait attributions affectprocedural preferences and account for situational andcultural differences. Asian Journal of Social Psychology, 7, 127–147.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and

shortcomings for social judgment. Englewood Cliffs, NJ: Prentice-Hall.

Nisbett, R. (2003). The geography of thought—How Asians and

Westerners think differently—and why. NY: Simon & SchusterAdult Publishing Group.

Oyserman, D., Kemmelmeier, M., & Coon, H. M. (2002). Culturalpsychology, a new look: Reply to Bond (2002), Fiske (2002),Kitayama (2002), and Miller (2002). Psychological Bulletin, 128(1),110–117.

Roth, A. E., Prasnikar, V., Okuno-Fujiwara, M., & Zamir, S. (1991).Bargaining and market behavior in Jerusalem, Ljubljana, Pitts-burgh and Tokio: An experimental study. American Economic

Review, 81(5), 1068–1095.

Shafir, E., Simonson, I., & Tversky, A. (1993). Reason-based choice.Cognition, 49, 11–36.

Srivastava, J. (2001). The role of inferences in sequential bargainingwith one-sided incomplete information: Some experimental evi-dence. Organizational Behavior and Human Decision Processes,

85(1), 166–187.Srivastava, J., Chakravarti, D., & Rapoport, A. (2000). Price and

margin negotiations in marketing channels: An experimental studyof sequential bargaining under one-sided uncertainty and oppor-tunity cost of delay. Marketing Science, 19, 163–184.

Srivastava, J., & Valenzuela, A. (2004). Role of social identity andinformation in facilitating exchange: Effects on fairness percep-tions, attribution, and behavior. Working paper, University ofMaryland, College Park, MD.

Takano, Y., & Osaka, E. (1999). An unsupported common view:Comparing Japan and the U.S. on individualism/collectivism.Asian Journal of Social Psychology, 2(December), 311–341.

Triandis, H. C. (1994). Culture and social behavior. New York: Simon& Schuster.

Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO:Westview Press.

Thaler, R. H. (1988). The ultimatum game. Journal of Economic

Perspectives, 2(Autumn), 195–206.White, S. B., & Neale, M. A. (1994). The role of negotiator aspirations

and settlement expectancies in bargaining outcomes.Organizational

Behavior and Human Decision Processes, 57(February), 303–317.Yzerbyt, V. Y., Rogier, A., & Fiske, S. T. (1998). Group entitativity

and social attribution: On translating situational constraints intostereotypes. Personality and Social Psychology Bulletin, 24(10),1090–1104.

the role of cultural orientation in bargaining under incomplete information: differences in causal...

Documents