Cross-cultural consensus in personality judgments

12
INTERPERSONAL RELATIONS AND GROUP PROCESSES Cross-Cultural Consensus in Personality Judgments Linda Albright Westfield State College Thomas E. Malloy Rhode Island College Qi Dong Beijing Normal University David A. Kenny University of Connecticut Xiaoyi Fang Beijing Normal University Lynn Winquist Rhode Island College DaYu University of Connecticut Building on recent research demonstrating consensus and accuracy in interpersonal perception based on minimal information, the present studies examined American and Chinese participants' within- and cross-cultural judgments. In Study I, the authors used the zero-acquaintance paradigm in the People's Republic of China and found consensus on all personality dimensions. In Study 2, Chinese arid American participants judged each other on the basis of photographs, and consensus was found among Americans' judgments of Chinese and Chinese participants' judgments of Americans. Further, by correlating target effects based on within-culture zero-acquaintance judgments and cross-cultural photographic judgments, the authors found agreement in the judgments of individuals by members of their own culture and the other culture for both Chinese and Americans. Linda Albright, Department of Psychology, Westfield State College; Thomas E. Malloy and Lynn Winquist, Department of Psychology, Rhode Island College; Qi Dong and Xiaoyi Fang, Institute of Develop- mental Psychology, Beijing Normal University, Beijing, China; David A. Kenny, Department of Psychology, University of Connecticut; Da Yu, Department of Management, University of Connecticut. Lynn Winquist is now at the Department of Psychology, University of Connecticut. This research was conducted while Qi Dong was a fellow at the Center for Advanced Studies in the Behavioral Sciences at Stanford University and a visiting scholar at the University of California at Berke- ley and while Xianyi Fang was a visiting scholar at Rhode Island College and the University of Maryland Medical School. This research was supported by grants from the University of Connect- icut Research Foundation, National Science Foundation Grant DBS- 9307949, National Institute of Mental Health Grant RO1-MH51969, and by Beijing Normal University research funds. We thank Bing Ru Zheng for her assistance with the translation and preparation of the stimulus materials and Lin Chuan Chu for assistance with the back-translation. Also, we thank our research assistants, Xin Tao, Chen \fong, Suzy Barcelos, and Linda Pelopida. We thank Yueh- Ting Lee for his comments. Correspondence concerning this article should be addressed to Linda Albright, Department of Psychology, Westfield State College, Westfield, Massachusetts 01086. Electronic mail may be sent via the Internet to [email protected]. Research on social perception reflects the fact that a person is both a social and a visual stimulus. Because groups develop social codes and belief systems that link the external aspects of people to internal features, communication of information that is typically studied under the rubric of impression formation (e.g., trait attribution and organization) can occur prior to social interaction. This social construction of the interpersonal world can function to allow skilled social actors to manage particular impressions, but it can also limit the impressions people can attempt to manage. Numerous studies conducted in Western culture have demon- strated consensus or agreement in strangers' judgments of oth- ers' personality characteristics based on minimal information (Albright, Kenny, & Malloy, 1988; Ambady & Rosenthal, 1993; Berry, 1990; Borkenau & Liebler, 1992, 1993; Funder & Colvin, 1988; Gifford, 1994; Kenny, Homer, Kashy, & Chu, 1992; Lev- esque & Kenny, 1993; Norman & Goldberg, 1966; Paunonen, 1989; Watson, 1989). The face-to-face context in which unac- quainted people judge each other's personality characteristics in the absence of social interaction or exposure to verbal behav- ior is referred to as zero acquaintance (Albright et al., 1988). Despite differences among the studies in terms of measures, designs, and analytic strategies, consensus in judgments of Ex- traversion and Conscientiousness has been found consistently. Journal uf Personality and Social Psychology, 1997. Vol. 72, No. 3. 558-569 Copyright 1997 by the American Psychological Association. Inc. 0022-3514/97/S3.00 558

Transcript of Cross-cultural consensus in personality judgments

INTERPERSONAL RELATIONS AND GROUP PROCESSES

Cross-Cultural Consensus in Personality Judgments

Linda AlbrightWestfield State College

Thomas E. MalloyRhode Island College

Qi DongBeijing Normal University

David A. KennyUniversity of Connecticut

Xiaoyi FangBeijing Normal University

Lynn WinquistRhode Island College

DaYuUniversity of Connecticut

Building on recent research demonstrating consensus and accuracy in interpersonal perception basedon minimal information, the present studies examined American and Chinese participants' within-and cross-cultural judgments. In Study I, the authors used the zero-acquaintance paradigm in thePeople's Republic of China and found consensus on all personality dimensions. In Study 2, Chinesearid American participants judged each other on the basis of photographs, and consensus was foundamong Americans' judgments of Chinese and Chinese participants' judgments of Americans. Further,by correlating target effects based on within-culture zero-acquaintance judgments and cross-culturalphotographic judgments, the authors found agreement in the judgments of individuals by membersof their own culture and the other culture for both Chinese and Americans.

Linda Albright, Department of Psychology, Westfield State College;Thomas E. Malloy and Lynn Winquist, Department of Psychology,Rhode Island College; Qi Dong and Xiaoyi Fang, Institute of Develop-mental Psychology, Beijing Normal University, Beijing, China; David A.Kenny, Department of Psychology, University of Connecticut; Da Yu,Department of Management, University of Connecticut.

Lynn Winquist is now at the Department of Psychology, Universityof Connecticut.

This research was conducted while Qi Dong was a fellow at theCenter for Advanced Studies in the Behavioral Sciences at StanfordUniversity and a visiting scholar at the University of California at Berke-ley and while Xianyi Fang was a visiting scholar at Rhode Island Collegeand the University of Maryland Medical School.

This research was supported by grants from the University of Connect-icut Research Foundation, National Science Foundation Grant DBS-9307949, National Institute of Mental Health Grant RO1-MH51969, andby Beijing Normal University research funds.

We thank Bing Ru Zheng for her assistance with the translation andpreparation of the stimulus materials and Lin Chuan Chu for assistancewith the back-translation. Also, we thank our research assistants, XinTao, Chen \fong, Suzy Barcelos, and Linda Pelopida. We thank Yueh-Ting Lee for his comments.

Correspondence concerning this article should be addressed to LindaAlbright, Department of Psychology, Westfield State College, Westfield,Massachusetts 01086. Electronic mail may be sent via the Internet [email protected].

Research on social perception reflects the fact that a personis both a social and a visual stimulus. Because groups developsocial codes and belief systems that link the external aspects ofpeople to internal features, communication of information thatis typically studied under the rubric of impression formation(e.g., trait attribution and organization) can occur prior to socialinteraction. This social construction of the interpersonal worldcan function to allow skilled social actors to manage particularimpressions, but it can also limit the impressions people canattempt to manage.

Numerous studies conducted in Western culture have demon-strated consensus or agreement in strangers' judgments of oth-ers' personality characteristics based on minimal information(Albright, Kenny, & Malloy, 1988; Ambady & Rosenthal, 1993;Berry, 1990; Borkenau & Liebler, 1992, 1993; Funder & Colvin,1988; Gifford, 1994; Kenny, Homer, Kashy, & Chu, 1992; Lev-esque & Kenny, 1993; Norman & Goldberg, 1966; Paunonen,1989; Watson, 1989). The face-to-face context in which unac-quainted people judge each other's personality characteristicsin the absence of social interaction or exposure to verbal behav-ior is referred to as zero acquaintance (Albright et al., 1988).Despite differences among the studies in terms of measures,designs, and analytic strategies, consensus in judgments of Ex-traversion and Conscientiousness has been found consistently.

Journal uf Personality and Social Psychology, 1997. Vol. 72, No. 3. 558-569Copyright 1997 by the American Psychological Association. Inc. 0022-3514/97/S3.00

558

CONSENSUS ACROSS CULTURE 559

Moreover, in most studies the consensual judgments correlatedwith the targets' self-ratings (Albright et al., 1988; Borkenau &Liebler, 1993; Kenny et al., 1992; Norman & Goldberg, 1966)and predicted behavior in an experimental context (Levesque &Kenny, 1993).

Studies of the external cues that could serve as the basis ofjudgments have found that perceptions of both static, appearancecues and dynamic, nonverbal cues are strongly associated withjudgments of particular personality attributes (Albright et al.,1988; Borkenau & Liebler, 1992; Gifford, 1994; Kenny et al.,1992). These findings indicate that in the absence of other infor-mation, people similarly use external cues as a basis for inferringpersonality traits, a social process that renders consensus inpersonality judgments. Presumably, people learn stereotypes thatlink external cues to personality dispositions and then applythem in the zero-acquaintance context. Such stereotypes couldbe adaptive for both groups and individuals in that they reflectand, through socialization, facilitate a shared construction of thesocial world (Vygotsky, 1978), which then allows individualsto predict others' inferences under conditions of minimal orcontrolled information.

Although the zero-acquaintance phenomenon has been clearlyestablished, all of the studies have been conducted in two West-ern cultures—the United States and Germany (Borkenau &Liebler, 1992, 1993);1 no studies have been conducted in anyEastern culture. Given the current awareness of cultural differ-ences and the lack of cross-cultural replication of a numberof important social psychological phenomena (Smith & Bond,1993), we have no basis for knowing whether consensus at zeroacquaintance is limited to Western culture or whether it occurscross-culturally. In addition, there are no studies that have exam-ined this phenomenon in judgments across cultures. That is, wedo not know whether consensus and self-other agreement atzero acquaintance generalizes to contexts in which perceiversand targets are from different cultures.

Assuming the cross-cultural generality of consensus in per-sonality judgments at zero acquaintance (within-culture judg-ments in a face-to-face context), there are two possibilitiesregarding the use of external cues: The pattern of associationsbetween cues and traits may be similar or different across cul-tures. Sociocultural theory would imply that cultural differencesin belief and meaning systems would result in different externalcues for personality judgments in different cultures. Alterna-tively, the patterns of trait-cue associations may not differ be-tween cultures, a result that would support the hypothesis ofuniversality.

In regard to cross-cultural judgment, assuming the existenceof consensus in personality judgments across cultures (judg-ments of members of a different culture), there are three theoret-ical perspectives regarding the use of external cues. Postmodernepistemologies emphasize the worldview embeddedness of per-ception and cognition and suggest that understanding "others"(e.g., historical eras, paradigms, cultures, and social classes) isconstrained by various "centrisms" (Campbell, 1996). Thatis, misinterpretation occurs because in-group observers do notconsider, or perhaps even realize, that others may have a differentworldview and mistake their own worldview as reality itself(Campbell, 1996). In intercultural perception, if people applytheir own cultural beliefs and if their stereotypes are untrue or

meanings systems different, then misperception and misunder-standing may result. From a postmodern epistemological per-spective, then, indigenous belief systems serve as the basis ofintercultural perception, a process that would produce commondifferentiation among members of the same culture when judg-ing those of a different culture.

The empirical results of research on out-group perception, ofwhich intercultural perception is a special case, however, suggesta different pattern of results. This research has indicated thatpeople perceive members of an out-group as relatively homoge-neous (e.g., Quattrone & Jones, 1980). On the basis of thisresearch, when making personality judgments of individualmembers of another culture, people should simply use theirstereotypes about the out-group, a process that would yield alack of differentiation of the targets (i.e., all targets would bejudged by the stereotypes of that group).

The two perspectives discussed above rest on the assumptionof different belief and meaning systems across cultures. If thisassumption is untrue, then there should be one universal patternof associations between trait judgments and external cues in allcontexts of judgment. Given the different theoretical perspec-tives regarding the role of culture in the use of external cueswhen making social judgments under conditions of minimalinformation, we review the empirical results of research on theeffects of appearance and nonverbal behavior in judgments ofpersonality characteristics in Eastern and Western culture.

Culture and Meaning Systems

Western Culture

The effect of physical attractiveness on judgments of person-ality and social characteristics has been examined in numerousstudies. In most studies, participants have been shown photo-graphs of individuals who vary in attractiveness while otherrelevant target characteristics, such as gender or ethnicity, havebeen controlled. Although an entirely ubiquitous attractivenesseffect has not been found, attractive individuals have tendedto be judged more positively (Dion, 1986; Eagly, Ashmore,Makhijani, & Longo, 1991).

The physical attractiveness bias generalizes to face-to-facesituations, as demonstrated by research conducted within thezero-acquaintance paradigm. These studies, which have typi-cally used the Big Five trait taxonomy, have also shown a morelimited effect of attractiveness. Specifically, the perceived per-sonality correlates of attractiveness were limited to traits indicat-ing Extraversion (Albright et al., 1988; Borkenau & Liebler,1992; Kenny et al., 1992).

Research on the perceived personality correlates of baby-facedness has also demonstrated an effect of physical appear-ance under conditions of limited information. These studies have

1 The Borkenau and Liebler (1992, 1993) studies are not technicallyzero-acquaintance studies in that perceivers were exposed to the verbalbehavior of the targets (i.e., watching them recite a news report). How-ever, because this behavior was scripted and identical for each target, itcould not provide idiosyncratic information about the target. In fact,consensus in personality judgments occurred even when participantswere only exposed to a silent videotape of the target.

560 ALBRIGHT ET AL.

shown that individuals with so-called babyfaces are perceivedto be less powerful and dominant but more naive, honest, andwarm(Berry &McArthur, 1986). BorkenauandLiebler (1992)found that people who were perceived to be baby-faced wererated as less conscientious.

Nonverbal behavior has also been associated with stable inter-nal characteristics. Reis et al. (1990) found that people whosmiled were seen as more sincere, sociable, and competent butless independent and masculine than were those who did notsmile. Consistent with these findings, Otta, Lira, Delevati, Cesar,andPires (1994) found that, in a sample of Brazilian undergrad-uates, people were perceived more favorably when they smiled.

In a comprehensive program of research (Gifford, Ng, &Wilkinson, 1985; Gifford, 1991, 1994) investigating the rela-tionships among personality, nonverbal behavior, and perceptionof personality based on nonverbal behavior, Gifford (1994)demonstrated that personality is encoded in nonverbal behavior,that personality judgments are linked to variation in nonverbalbehavior, and that perceivers' judgments based on nonverbalbehavior correlate with self-ratings of personality. Using traitsfrom Leary's (1957) circumplex model, Gifford found a numberof strong correlations between judgments of interpersonal traitsand various nonverbal behaviors (mainly smiling, gesturing,nodding, and eye contact). Moreover, moderate to strong levelsof self-other agreement were found on the traits gregarious -extraverted and aloof-introverted.

Gifford's (1994) results are consistent with findings thatjudgments of sociability are strongly correlated with smilingand body movement (Borkenau & Liebler, 1992; Kenny et al.,1992) and with eye contact (Borkenau & Liebler, 1992). Theyare also consistent with the research cited above that foundcorrelations between consensual judgments of Extraversion atzero acquaintance and the self-ratings of Extraversion (Albrightet al., 1988; Kenny et al., 1992). Indeed, a meta-analysis ofstudies measuring the accuracy of predictions of objective out-comes based on nonverbal behavior showed that such predic-tions are fairly accurate even when based on observations under0.5 min in length (Ambady & Rosenthal, 1992).

Eastern Culture

Although a historical precedent for the belief in the link be-tween appearance and character is evident in the ancient philo-sophical text, The Yellow Emperor's Classic of Internal Medi-cine (cf. Lam & Berrios, 1992), empirical investigation of thishypothesis in Eastern culture is limited. Dion, Pak, and Dion(1990) sampled Chinese students attending school in Canadaand classified them as being high or low in involvement inthe Chinese community. They found that those who were moreinvolved in the Chinese community were less affected by physi-cal attractiveness when judging the characteristics of a targetindividual. Consistent with this finding, Hui and Yam (1987)examined the effect of physical attractiveness and Englishlanguage proficiency on perceptions of Chinese targets andfound no effect of attractiveness on judgments of personalitycharacteristics.

Empirical research on the perceived personality correlates ofnonverbal behavior in Eastern culture is also limited. Consis-lency in the findings of Argyle, Henderson, Bond, Iizuka, and

Contarello (1986) and M. H. Bond (1993) indicates that thereare strong norms against the expression of emotion in Chineseculture. Lau (1982), however, found that Chinese rated smilingtargets as more interpersonally attractive and intelligent thannonsmiling targets. In addition, Matsumoto and Kudoh (1993)found that Japanese and Americans perceived smiling targets asmore sociable, but only Americans perceived smiling targets asmore intelligent.

Cultural Moderation of Belief and Meaning Systems

Empirical research suggests that the formation and use ofstereotypes linking external and internal characteristics is a cul-tural process, whereby group-level beliefs become incorporatedinto individual group members' belief systems and are thenapplied in everyday life. The research reviewed above suggeststhat some stereotypes are cross-cultural, whereas others aregroup specific. Conclusions about cultural differences cannotbe drawn directly from this research, however, because judg-ments were made strictly within culture. In the next section, wereview research in which people make judgments of membersof a different culture.

Cross-Cultural Interpersonal Perception

Studies in which participants judge the personality attributesof individual members of a different culture (i.e., cross-culturalinterpersonal perception) are relatively rare. Studies of facialexpression of emotion, detection of deception, baby-facedness,and perception of physical attractiveness constitute the majorresearch on interpersonal perception in which perceivers andtargets are from different cultures. Although focused on differentphenomena, these studies suggest universality rather than cul-tural moderation.

For example, cross-cultural studies have typically shown thatfacial expressions of emotion are perceived consensually withinand between cultures (Ekman et al., 1987; Gordon, Zukas, &Chan, 1982; Keating et al., 1981; Krauss, Curran, & Ferlinger,1983). Given the robustness and reliability of this phenomenon,Darwin's (1872) hypothesis that facial displays of emotion areuniversal in meaning and biological in origin has become widelyaccepted as valid. Research on lie detection also has showncultural similarity in that people from different cultures usedsimilar external cues to detect deception (C. F. Bond, Omar,Mahmoud, & Bonser, 1990).

Cross-cultural research on the perception of physical attrac-tiveness has also shown evidence of universality. Cunningham,Roberts, Barbee, Druen, and Wu (1995 ) had Asians, Hispanics,and White Americans rate the attractiveness of photographs ofAsian, Hispanic, Black, and White women. They found highcorrelations among the ratings of these groups, although Asianswere less positively influenced by sexual maturity and expres-sive features. Consistent with these results, consensus in judg-ments of attractiveness has been found between Cruzans andAmericans rating Whites (Maret & Harling, 1985); Chinese,Indians, and English rating Greeks (Thakerar & Iwawaki,1979); Whites, Blacks, and Chinese rating Whites and Chinese(Bernstein, Tsai-Ding, & McClelland, 1982); and Americans

CONSENSUS ACROSS CULTURE 561

and White South Africans rating Whites (Morse, Gruzen, &Reis, 1976).

In a series of studies investigating perceived relationshipsbetween external cues and personality traits, Zebrowitz and hercolleagues have demonstrated cross-cultural agreement in baby-facedness-trait associations (McArthur & Berry, 1987;Zebrowitz & Montepare, 1992) and vocal quaJity-trait relation-ships (Montepare & Zebrowitz-McArthur, 1987), but cross-cultural differences in perceived relationships between speechrate and power (Peng, Zebrowitz, & Lee, 1993). Using point-light displays as stimuli, Montepare and Zebrowitz (1993)found cross-cultural agreement in the perception of the target'sage, gender, happiness, sexiness, and physical strength.

Overview of the Present Research

Given the consistent finding of within-culture consensus inpersonality judgments at zero acquaintance in Western culture,the major purposes of Study 1 were to demonstrate the cross-cultural general izability of this phenomenon and to investigateChinese belief and meaning systems in terms of trait—cue asso-ciations. Study 2 was designed to demonstrate consensus inpersonality judgments among members of one culture whenjudging members of another culture and agreement in judgmentsbetween culture.

Conducting cross-cultural research presents a number ofmethodological issues. One issue involves the translation of theinstructions and stimulus materials from the language in whichthey were originally conceived to the host language (Brislin,1980). Central to problems of translation is the emic-etic dis-tinction, a concept, which has received much attention in cross-cultural research, that refers roughly to the difference betweenelements that are culture specific versus culture general, respec-tively (Triandis, 1994). When using an etic framework, whichis necessary for cross-cultural comparisons, emic phenomenamay be overlooked or misrepresented. However, given that thepresent research involved a replication, designing the studiesrequired a balance between keeping aspects constant and beingsensitive to cultural emics.

Because the procedures were fairly simple (primarily entail-ing having participants judge each other and themselves), thenature of the rating dimensions was our main concern withrespect to the emic-etic issue. In this regard, we selected fivetraits from Yang and Bond's (1990) emic taxonomy of personal-ity dimensions, which they labeled the Chinese Big Five, andfive traits from the Western Big Five. Questioning whether theWestern Big Five adequately represents the full range of person-ality dimensions Chinese use to structure interpersonal varia-tion, Yang and Bond (1990) analyzed the factor structure oftrait ratings of 150 emic personality dimensions and 20 traitsfrom the Western Big Five. Their analyses revealed five factors,which they labeled Social Orientation, Competence, Expressive-ness, Self-Control, and Optimism.

Study 1

Study 1 entailed a cross-cultural replication of the zero-ac-quaintance paradigm in the People's Republic of China. Partici-pants, who were students of Beijing Normal University, were

randomly assigned to groups of 5 with the constraint that nogroup member had any previous acquaintance with or knowl-edge of any other group member. Once each group of partici-pants arrived at the laboratory, they were requested to rate them-selves and each other on 10 personality dimensions and 4 exter-nal dimensions, while remaining silent. Our attempts to besensitive to the special methodological problems of cross-cul-tural research consisted of the translation procedure and theselection of the personality dimensions.

Method

Participants. Participants were 80 female and male students fromBeijing Normal University who were randomly selected from five depart-ments within the university. Each group contained 1 student from eachdepartment. Most of the participants were first-year students and camefrom different provinces throughout China. They were recruited on avoluntary basis and were paid $5 each for their participation. To assurezero acquaintance, at recruitment, participants were asked whether theyknew any persons in the other four departments. If not, he or she wasselected as a participant. As an additional check, after the experimentparticipants were asked again whether they knew any other member ofthe group prior to the study.

Measures and translation. Participants rated each other and them-selves on 14 dimensions. Five of these were traits representing each ofthe factors of the Western Big Five: sociable, good-natured, responsible,calm, and intelligent; the other 5 dimensions were traits representing eachfactor of Yang and Bond's (1990) Chinese Big Five: honest, independent,active, self-controlled, and optimistic. Selection of each particular traitfrom the Chinese dimensions was based on two criteria, namely factorloading and ease of translation (see Brislin, 1970). The final 4 dimen-sions were salient external cues: physical attractiveness, neatness ofdress, smiling, and eye contact. All ratings were made on 7-point scalesanchored by the trait term on one end and its opposite on the other end.The scale was reversed for some items to discourage response set.

To produce a semantically equivalent version of the rating instrumentin Chinese, the back-translation method (Brislin, 1980) was used. Whenpreparing the stimulus materials for translation, we followed the relevantguidelines proposed by Brislin (1980). In this regard, we used simplesentences, phrased in the active voice, that contained familiar terms andpresented the directions and variables redundantly to facilitate compre-hension. Brislin also advocated the use of a multimethod approach,which, in the present context, entailed the formation of five factors fromthe 10 personality trait judgments.

Procedure. Groups of 5 participants were scheduled for each experi-mental session of the study. As participants arrived at the laboratory,they were greeted by the experimenter, asked to sit as quietly as possible,and instructed not to talk. Prior to their arrival, packets of rating formsand identification tags were placed face down on desks that were ar-ranged approximately 6 ft (1.8 m) apart in a circle. Each rating packetwas labeled with a group number and letter (A, B, C, D, or E) ; eachidentification tag was marked with the corresponding letter. Thus, per-sons were identified to each other only by a letter, not by names, andonce seated, they were instructed to pin the identification tag on theirleft shoulders. They were then told that they would be requested to makejudgments about themselves and the other members of the group on theforms in front of them. The experimenter then instructed the participantsto take a moment to look at each other. After approximately 10 s, theywere told to turn the forms over, complete them in silence, and turnthem face down again when they were finished.

Once all forms were completed, the experimenter collected them andinformed participants that the first phase of the study was over. Thesecond phase, which was optional, entailed taking photographs of each

562 ALBRIGHT ET AL.

participant and provided the stimuli for Study 2. Participants were thenthanked and paid for their participation.

Design and analysis. The round-robin design was used in this re-search. The round-robin is a reciprocal design (Kenny & Albright, 1987)in which each member of the group rales each other member of thegroup. Thus, each person rates multiple targets and is rated by multipleperceivers. A social relations analysis of the data was accomplished byusing the computer program SOREMO (Kenny, 1995b). This analysispartitions the variance in interpersonal perception data into three compo-nents, which are called perceiver, target, and relationship. A thoroughdescription of the social relations model can be found elsewhere (Kenny,1994; Kenny & La Voie, 1984), so we only briefly discuss the relevantaspects of the model and analysis here. Because consensus was theprimary interest in this research, the target effect is the most relevantcomponent.

The target effect reflects die way the target tends to be seen by theperceivers and, to the extent that the targets are seen similarly by differentperceivers (but discriminated from each other), there will be targetvariance. The tendency for the target to be seen similarly by differentperceivers is called consensus. Target variance on the trait of honesty,for example, would indicate that some targets were consensually seenas honest, whereas other targets were consensually seen as dishonest.

Results

A social relations analysis was performed by using the com-puter program SOREMO (Kenny, 1995b). In a social relationsanalysis, both the absolute variance and the proportion of totalvariance due to each component are estimated. Although Yangand Bond (1990) found the Chinese factors to be somewhatindependent of the Big Five, we found that target effects on theChinese factors did correlate with target effects on the BigFive factors. Therefore, to reduce the number of estimates, thefollowing data reduction strategy was used. The 10 trait judg-ments were used to form five factors, which were based on theBig Five taxonomy. Using Goldberg's (1990) classification, weformed the following factors: Extraversion (sociable and ac-tive), Agreeableness (good-natured, honest, and optimistic),Conscientiousness (responsible), Emotional Stability (calm, in-dependent, and self-controlled), and Culture (intelligent). Tocompute target variances for each factor, we averaged the targetvariances across the multiple indicators of each factor. To com-pute correlations between target effects, we averaged the targeteffects across the multiple indicators of each factor and thencomputed correlations between the averaged target effects.

To estimate the relationship of trait judgments and cues, wefirst conducted a canonical correlation analysis treating the tar-get effects on trait judgments as criteria and observable cues aspredictors while controlling the group effect for each judge.Individual-within-group was the unit for this analysis. If themultivariate test showed a statistically significant associationbetween criteria and predictors, univariate analyses were thenconducted. The target variances and all correlations were testedby t tests, and statistically based inferences were made at the.05 significance level. The unit of analysis was group for boththe target variances and the univariate correlations between tar-get effects. Correlations involving variance components that arenear zero should not be considered as they are meaninglessand sometimes anomalous. For example, if the researcher wasinterested in the correlation between the target effects on honestand smiling, but the target variance on honest was very low

Table 1Proportion of Target Variance in the ChineseZero-Acquaintance Judgments

Variable Proportion of target variance

TraitExtraversionAgreeablenessConscientiousnessEmotional StabilityCulture

External cueAttractivenessNeatly dressedSmilingEye contact

16*10*8*

10*10*

30*8*

25*15*

Note. Entries are percentages of total variance. Unit of analysis isgroup ( # = 15).*p < .05.

(e.g., .01), this correlation would be uninterpretable and mayeven be out of range.

Consensus. Table 1 displays the percentage of the total vari-ance that was due to the target for the personality and the exter-nal cue judgments. As discussed earlier, the proportion of vari-ance due to the target is the estimate of consensus. As can beseen, the proportions of variances that were due to the targetare statistically reliable for all variables. Thus, there is clearevidence of consensus at zero acquaintance among the Chinese.It is interesting that the magnitudes of the target variances ofthe traits were fairly homogeneous, ranging from 8% to 16%.We note that although the magnitudes of the target varianceswere more homogenous than in the American samples, an im-portant pattern replicates: The highest target variance occurredon the Extraversion dimension. As noted earlier, the pattern oftarget variances is quite consistent across American samples inthat consensus tended to occur on the dimensions Extraversionand Conscientiousness, with Extraversion showing the greatestconsensus.

Table 1 also shows the percentage of variance that was dueto the target for the external cues. As can be seen, the proportionof variance due to the target was higher overall for the externalcues than for the traits. It is interesting that the external cue thatseemed the most observable and objective, neatness of dress,had the lowest level of consensus. This counterintuitive resultmay be a matter of differential interpretation of the Chinesetranslation of the English term, or it may be due to homogeneityin dress.

Correlations between external cues and traits. The multi-variate analysis showed that the set of predictor cues was sig-nificantly related to the set of target effects on trait judgments,Hotelling's T2 = 3.34, p < .05 with F(20, 218) - 9.12. Table2 shows the univariate correlations between the target effectson the traits with the target effects on the external cues. Onecan readily see from this table that static cues (attractive andneatly dressed) were generally uncorrelated with the target ef-fects on the traits. The clear exceptions to this pattern are thecorrelation of .95 (p < .05) between Culture (intelligent) andattractive and the correlation of .51 (p < .05) between Culture

CONSENSUS ACROSS CULTURE 563

Table 2Correlations Between Target Effects on External Variablesand Traits for the Chinese Zero-Acquaintance Judgments

Trait

ExtraversionAgreeablenessConscientiousnessEmotional StabilityCulture

Attractiveness

.44

.37

.02- .04

.95*

Note, Unit of analysis is group (df =* p < .05.

External cue

Neatlydressed

.21

.30

.11- .05

.51*

15).

Smiling

.76*

.66*

.34- .85*

.17

Eyecontact

.66*

.36- .16- .35- .03

and neatly dressed. Thus, those who were seen as attractive andneatly dressed were rated as more intelligent.

In contrast to the appearance variables, there were a numberof statistically reliable and strong correlations between the non-verbal behavior cues and the traits. Smiling correlated stronglywith Extraversion, Agreeableness, and Emotional Stability, andeye contact correlated strongly with Extraversion. Note, how-ever, that smiling was negatively correlated with Emotional Sta-bility. Indeed, analyses at the single indicator level showed thatsmiling was associated with a lack of calmness and self-control,and eye contact was associated with a lack of perceivedcalmness.

Discussion

The results of Study 1 confirmed the hypothesis that consen-sus in personality judgments at zero acquaintance is a cross-cultural phenomenon. Consensus in judgments occurred on allpersonality factors and on all external cues. Consistent withtypical results in Western studies using the zero-acquaintanceparadigm (e.g., Albright et al., 1988), consensus was highest onjudgments of Extraversion, accounting for 16% of the variance.Although the pattern replicates, we note that the level of consen-sus was typically higher (approximately 25%) in the Westernstudies.

The correlations between the target effects on the traits andon the external cues suggest that these cues played a role in theconsensus in personality judgments. Correlations between thenonverbal cues and the traits were more numerous than betweenthe appearance cues and traits; however, a strong correlationbetween attractiveness and Culture (intelligence) was found.Further, although most correlations between traits and cues werepositive, the correlation between smiling and Emotional Stabil-ity was strong and negative ( — .85). This result, however, isconsistent with the Chinese cultural norm to suppress the ex-pression of emotion, particularly with strangers, so as not toimpose one's feelings on others.

Study 2

Given the demonstration of consensus in personality judg-ments at zero acquaintance among the Chinese, there were three

primary purposes of Study 2. One was to address the questionof whether consensus in strangers' judgments of personalitygeneralizes to contexts in which perceivers and targets are fromdifferent cultures. Another purpose was to determine whetherpeople would apply the appearance stereotypes and meaningsystems for nonverbal behavior from their own culture to judgemembers of another culture. That is, will perceivers impose theworldviews of their own culture when judging members of othercultures? Or, would the trait-cue associations more stronglyreflect those of the other culture? The third purpose was todetermine the extent to which judgments of strangers are moder-ated by culture. That is, would people from different culturessimilarly judge the personality characteristics of a stranger?

Given these three purposes, we formulated and tested thefollowing hypotheses. First, we predicted that there would beconsensus in both the Americans' ratings of the Chinese (Hy-pothesis la) and the Chinese ratings of the Americans (Hypothe-sis lb) . Second, because theory suggests that cross-culturalperception is constrained by differential worldviews (Campbell,1996), we predicted that for each culture the pattern of correla-tions between the traits and the external cues for the cross-cultural judgments would more closely resemble the patternsobserved in the within-culture judgments. More specifically, onthe basis of the results of Study 1, we predicted that the covaria-tion between perceived attractiveness and Culture would behigher than the correlation between attractiveness and sociability(Extraversion) for the Chinese (Hypothesis 2a) but that thereverse pattern would hold for the Americans (Hypothesis 2b).We also predicted that the covariations between perceived smil-ing and the traits calm and self-control (Emotional Stability)would be negative for the Chinese (Hypothesis 2c) but neutralor positive for the Americans (Hypothesis 2d). Finally, becauseStudy 1 suggested that there is cross-cultural agreement on theperceived external correlates of Extraversion, Hypothesis 3 wasthat there would be cross-cultural consensus in judgments ofExtraversion, That is, we predicted agreement between Ameri-cans' and Chinese judgments of Extraversion for the Americantargets and between Americans' and Chinese judgments of Ex-traversion for the Chinese targets.

Method

Participants. The American participants were 77 female and malestudents from Rhode Island College who received course credit for theirparticipation in the study, and the Chinese participants were 72 maleand female students from Beijing Normal University, all of whom hadparticipated in Study 1. Eight of the participants from Study 1 did notparticipate in Study 2.

Measures and translation. We used the same instrument that wasused in Study 1, except that the instructions were modified. That is,participants were requested to rate the individuals in tiie photographs infront of them (as opposed to the actual people) on 14 dimensions.Because the instrument was virtually identical to that used in Study 1,most of the translation had been accomplished. R>r this study, then, onlythe instructions required translation. Back-translation of this portion ofthe instrument indicated that the American and Chinese participantsreceived the same instructions.

Procedure. Because we needed zero-acquaintance data for theAmerican participants, they were scheduled for the study in groups of5, with die constraint that no individual had any prior acquaintance withany other member of the group. These individuals then participated in

564 ALBRIGHT ET AL.

a zero-acquaintance study that replicated the procedure that was usedin Study 1. These data were collected so that perceiver and target effectsbased on Americans' ratings of Americans could be estimated and usedfor the analyses of cross-cultural consensus. After the zero-acquaintancedata were collected, each participant in the group was photographed.No instructions were given to the participant prior to the photographing.

After the photographs were taken, the American participants ratedthe Chinese targets. Because it would have required too many ratings,participants did not rate every participant from the other culture; rathereach group was assigned randomly a group from the other culture torate. Thus, each of the 16 American groups was paired randomly with1 of the 16 Chinese groups, and each Chinese group rated the Americangroup that rated them.

Located in different cubicles, participants were seated with the ratingform and photographs in front of them and were requested to rate eachindividual on the 14 dimensions. These photographs were arranged in arandom order for each judge, and the individual in each photo had alabel with their participant code (e.g., 3D) pinned on their left shoulder.Once these ratings were completed, participants were debriefed andthanked for their participation. We used the same procedure to obtain theChinese ratings of the American participants. The Chinese participants,however, made these judgments at a second experimental session.

Design and analysis. The block round-robin design (Kenny, 1994)was used in this research. A block round-robin design is one that containsa block and two round-robin data structures. The round-robin is a recip-rocal design (Kenny & Albright, 1987) in which each member of thegroup rates each other member of the group and perhaps provides aself-rating. The block is a design in which subgroups of individualsrate each individual in the other subgroup. In the present research, thesubgroups were the Chinese and the Americans. Each group of Chineserated themselves, each other (round-robin), and each member of a ran-domly assigned American group (half block), and each American grouprated themselves, each other (round-robin), and each member of a Chi-nese group (half block). Table 3 illustrates the data structure of thepresent study.

A social relations analysis of the data was accomplished by using thecomputer programs SOREMO (Kenny, 1995b) and BLOCKO (Kenny,

Table 3Block Round-Robin Design

Table 4

Relative Variances of Americans' Ratingsof Chinese Photographs

Perceiver

ABCDE

FGHIJ

A

S

wwww

YYYYY

B

W

swww

YYYYY

Americans

C

.

wwsww

YYYYY

D

Target

E

Ajnericans

WW

wsw

WWWW

sChinese

YYYYY

YYYYY

F

XXXXX

szzzz

G

XXXXX

zszzz

Chinese

H

XXXXX

zzszz

I

XXXXX

zzzsz

J

XXXXX

zzzzs

Variable Proportion of target variance

TraitExtraversionAgreeablenessConscientiou snessEmotional StabilityCulture

External cueAttractivenessNeatly dressedSmilingEye contact

23*26*15*8*

18*

16*20*43*

7*

Note. Entries are percentages of variance. Unit of analysis is group(df= 15).*p < .05.

1995a). Because of the nature of the data structure in this study, wewere able to compute two target effects for each participant. One targeteffect was computed on the basis of ratings of a participant made bymembers of the participant's culture; the other target effect was basedon ratings of a participant made by members of the other culture. We thencomputed ordinary correlations between the two target effects (cross-cultural target-target correlations).

Results

Again, we formed five factors representing the Big Five withthe 10 personality traits and conducted initial multivariate analy-ses to test for overall significance of cue-trait relationships.The perceiver and target variances and all correlations weretested by t tests against the null hypothesis (i.e., that the vari-ances were zero). The unit of analysis was group for boththe variance components and the within-culture correlations ofvariance components (e.g., trait-cue correlations) and individ-ual-within-group for the cross-cultural correlations between in-dividual-level measures (i.e., target effects).

Americans' ratings of Chinese: Consensus. Table 4 displaysthe consensus estimates for the Americans' ratings of the Chi-nese. As can be seen, these estimates were statistically reliableon all the personality factors and ranged from 8% to 26% ofthe variance. Consensus was highest on the factors Extraversionand Agreeableness and occurred at levels close to those observedamong highly acquainted persons (Kenny, Albright, Malloy, &Kashy, 1994). The external cues also showed statistically reli-able and substantial levels of consensus, ranging from 7% to43% of the variance.2 Overall, the American perceivers agreedto a remarkable extent on the personality traits and externalfeatures of the Chinese targets, thereby supporting Hypothe-sis la.

Americans' ratings of Chinese: Relationships between exter-nal cues and traits. The multivariate analysis showed that theset of predictor cues was significantly related to the set of target

Note. Ss represent self-ratings; Ws represent Americans' ratings ofAmericans; Xs represent Americans' ratings of Chinese; Ys representChinese ratings of Americans; Zs represent Chinese ratings of Chinese.

2 The low level of consensus on the variable eye contact is assumedto be due to the ambiguity of rating this cue on the basis of a photograph.

CONSENSUS ACROSS CULTURE 565

Table 5Correlations Between Target Effects on External Cues andTraits: Americans' Ratings of Chinese

Table 7Correlations Between Target Effects on External Variablesand Traits: Chinese Ratings of Americans

Trait

ExtraversionAgreeablenessConscientiousnessEmotional StabilityCulture

Attractiveness

.64*

.62*

.55*

.32

.64*

Note. Unit of analysis is group (df —

External

Neatlydressed

.04

.29

.90*

.57*

.71*

15).

cue

Smiling

.83*

.86*

.08

.23

.25

Eyecontact

.50*

.44

.64*

.56*

.47

Trait

ExtraversionAgreeablenessConscientiousnessEmotional StabilityCulture

Note. Unit of analysis* p < .05.

Attractiveness

.63*

.65*

.50*

.25

.96*

is group (df =

External cue

Neatly dressed -

.38

.37

.29

.09

.68*

15).

Smiling

1.00*.82*.29

- .36.60*

• p < .05.

effects on trait judgments, Hotelling's T1 = 12.15, p < .05 withF(15, 158) = 42.65. Table 5 shows the univariate correlationsbetween the target effects on the traits and the target effects onthe external cues on the basis of the Americans' ratings of theChinese. Target effects on attractiveness were correlated at astatistically significant level with the target effects on all person-ality factors, except Emotional Stability. Target effects on neatlydressed were significantly correlated with the target effects onConscientiousness, Emotional Stability, and Culture. Target ef-fects on smiling were correlated with target effects on Extraver-sion and Agreeableness. Target effects on eye contact were cor-related with target effects on Extraversion, Conscientiousness 4

and Emotional Stability. Consistent with Hypothesis 2b, the cor-relation between the target effects on attractiveness and sociable(r — .74) was higher, r(15) = 1.76, p — .05, one-tailed, thanthe correlation between the target effects on attractiveness andintelligent (r = .64). Consistent with Hypothesis 2d, there wereno statistically reliable correlations between smiling and calmor self-control (Emotional Stability).

Chinese ratings of Americans: Consensus. Table 6 showsthe proportions of variance that were due to the target, whichis the estimate of consensus.3 These estimates are statisticallysignificant on four of the five personality factors and range from

Table 6Relative Variances of Chinese Ratingsof American Photographs

Variable Proportion of target variance

TraitExtraversionAgreeablenessConscientiousnessEmotional Stability

• CultureExternal cue

AttractivenessNeatly dressedSmiling

44*29*

414*25*

33*31*59*

14% to 44% of the variance, thereby supporting Hypothesis lb.The greatest amount of consensus occurred on judgments ofExtraversion (44% of the variance), whereas judgments of Con-scientiousness showed no reliable target variance. Judgmentson the external cues attractiveness, neatly dressed, and smilingwere also reliably consensual, with partner effect estimates rang-ing from 31 % to 59% of the variance. Thus, there was a moder-ate degree of consensus on the static appearance cues attrac-tiveness and neatly dressed and substantial consensus on thedegree of smiling.

Chinese ratings of Americans: Relationships between exter-nal cues and traits. The multivariate analysis showed that theset of predictor cues was significantly related to the set of targeteffects on trait judgments, Hotelling's T2 = 6.70, p < .05, withF(20, 218) = 18.25. Table 7 shows the univariate correlationsbetween the target effects on the traits and the external cues.Consistent with the analogous correlations from the Americans'ratings of the Chinese, target effects on attractiveness were cor-related with target effects on all personality factors except Emo-tional Stability, The target effect on Culture was correlated withthe target effects on all external cues, but most highly correlatedwith attractiveness. Inconsistent with Hypothesis 2a, however,the correlation between the target effects on attractiveness andCulture (r = .96) was not reliably higher, t(15) = 0.54, thanthe correlation between the target effects on attractiveness andsociable (r = .63). The target effect on neatly dressed correlatedonly with Culture at .68. The target effect on smiling correlatedsignificantly with target effects on Extraversion, Agreeableness,and Culture. Consistent with the Chinese ratings of Chinese(Study 1) but inconsistent with the Americans' ratings of Chi-nese, the target effects on the traits calm and self-control (Emo-tional Stability) were negatively correlated with the target effecton smiling ( — .59 and —.38, respectively), thereby supportingHypothesis 2c. Thus, targets who smiled were judged as beingless emotionally stable (i.e., less calm and self-controlled).

Cross-cultural consensus in judgments. Table 8 shows thecorrelations between the target effects that were based on the

Note. Entries are percentages of variance. Unit of analysis is group< # = 15).* p < .05.

5 The Chinese participants did not rate the degree of eye contact whenjudging the American targets, although the variable appeared on therating instrument. Presumably, rating the degree of eye contact on thebasis of a photograph did not make sense to them.

566 ALBRIGHT ET AL.

Table 8Cross-Cultural Correlations Between Target Effects

Variable

TraitExtraversionAgreeablenessConscientiousnessEmotional StabilityCulture

External cueAttractivenessNeatly dressedSmilingEye contact

American targets

.I9f,27t

- .05.27t

- .11

.461

.34f

.41t—

Chinese targets

.29t

.35t- .08- .08

•22t

.16

.26t

.37t

.13

(Vore. Entries are between-subjects correlations between target effectsbased on ratings made by American participants and target effects basedon ratings made by Chinese participants. Degrees of freedom for Ameri-can participants is 60; for Chinese participants, 63. Dash indicates datawere not available,t p < -05, one tailed.

ratings made by the American perceivers and the target effectsthat were based on the ratings made by the Chinese perceiversfor both the American and the Chinese targets. These estimatesare between-subjects correlations in that they represent the rela-tionship between two variables (target effect estimates) whosemeasures originate from two different groups of judges. Notealso that the two target effect estimates for each participantwere based on ratings made in two different contexts of judg-ment: face-to-face versus photographs. As can be seen in Table8, for the American participants, these cross-cultural target-target correlations were statistically significant for three of thefive personality factors and for all external cues for which mea-sures were available from both cultures, and ranged from .19on Extraversion4 to .46 on attractiveness. Cross-cultural consen-sus was strong on judgments of external cues and moderate onjudgments of personality traits.

Table 8 also shows the cross-cultural consensus estimates forthe Chinese targets. These estimates are correlations betweenthe Chinese target effects based on Chinese ratings, which weremade in a face-to-face context, and the Chinese target effectsbased on the Americans' ratings, which were based on a photo-graph. As can be seen, these cross-cultural target-target correla-tions were statistically significant for three of the five personalitytraits and two of the four external cues, and ranged from .22 to.37. Intercultural consensus for the Chinese targets occurredon Extraversion, Agreeableness, Culture, neatly dressed, andsmiling. Note the absence of intercultural consensus on judg-ments of attractiveness, a result that stands in contrast to theresults for the American targets in which intercultural consensusof attractiveness was strongest.

The cross-cultural consensus estimates for both the Chineseand Americans support Hypothesis 3, which predicted cross-cultural consensus in judgments of Extraversion. Indeed, cross-cultural consensus also occurred on Agreeableness for both theAmerican and Chinese targets, a result that suggests that cross-cultural consensus in personality judgments generalizes to traitsbeyond Extraversion.

General Discussion

Cross-Cultural Replication of Consensus at ZeroAcquaintance

Study 1 was primarily designed to test the hypothesis thatconsensus in personality judgments at zero acquaintance is aphenomenon that generalizes to an Eastern culture. Through theuse of measures from both the American and Chinese Big Fivefactors, statistically reliable consensus occurred on all personal-ity traits among the Chinese. Consensus levels were highest onExtraversion, a finding that is consistent with the Americanstudies.

This replication is important because it suggests thai consen-sus at zero acquaintance is a general psychological phenomenon.The fact that this phenomenon occurs cross-culturally indicatesthat it may serve important social psychological functions. Ingroup formation and task allocation within groups, for example,consensus is essential. Very early in the group process, peoplemust determine who has the intellect, the good heart, or theleadership potential so that efficient task allocation and coordi-nated interpersonal behavior can occur. Social codes and sharedmeaning systems that accurately reflect the correspondence be-tween external features, such as appearance and nonverbal be-havior, and internal characteristics would clearly simplify andfacilitate effective interpersonal processes.

Cross-Cultural Consensus in Personality Judgments

In Study 2, consensus in personality judgments across cul-tures was demonstrated in two ways. First, the Americans agreedamong themselves to a substantial degree about the personalityattributes of Chinese individuals, and the Chinese agreed amongthemselves to a substantial degree about the personality attri-butes of Americans. Thus, judgments made by members ofone culture about members of the other culture were highlyconsensual.

Second, Chinese and Americans' consensual judgments ofAmericans were correlated on three of five personality factors,and Chinese and Americans' consensual judgments of Chinesewere correlated on three of five personality factors. Thus, peoplefrom different cultures, who were unacquainted with each otherand with the target, similarly judged the target on several dimen-sions. This finding is even more remarkable considering notonly that the perceivers were from different cultures but alsothat the context of judgment was different (noninteractive, face-to-face vs. photographs). Given that these judgments can bebased on only physical appearance and nonverbal behavior, theseresults suggest that there is both universality and cultural moder-ation in construals of the social world.

What is the basis of the cross-cultural consensus in personal-ity judgments? That is, what processes underlie the finding that

4 This correlation is likely attenuated because of the low target vari-ance on trait sociability in the Americans' ratings of Americans (4% ofthe variance). We believe that this atypical result is due to the scalereversal on this variable (i.e., 1 represented sociable; 7 representedreserved). Because some participants used the scale appropriately andothers did nol, consensus was attenuated.

CONSENSUS ACROSS CULTURE 567

Americans' consensual judgments of each other agree withChinese consensual judgments of Americans and vice versa?Two personality dimensions showed cross-cultural consen-sus for both the Chinese and Americans: Extraversion andAgreeableness. Considering the commonalities and differencesin the patterns of association between the external cues andthe traits, we propose that the primary basis of cross-culturalagreement is smiling. In all contexts of judgment—Chinesejudging Chinese, Chinese judging Americans, Americans judg-ing Americans, and Americans judging Chinese—smiling wasassociated strongly and positively with judgments of Extraver-sion and Agreeableness,

Cultural Similarities and Differences in CovariationBetween External Cues and Personality

Although in both cultures there was a more general attrac-tiveness-based halo effect on the cross-cultural judgments, in allcontexts, attractiveness was associated with positive personalitytraits. However, whereas attractiveness was most strongly as-sociated with intelligence among the Chinese (when judgingthemselves or Americans), attractiveness was most strongly as-sociated with Extraversion among the Americans (when judgingthemselves or Chinese).

Similarly, in both cultures, smiling was positively associatedwith socially oriented dimensions (Extraversion and Agreeable-ness) whether making within-culture or cross-cultural judg-ments. However, among the Chinese, smiling was associatedwith a lack of self-control and calmness (Emotional Stability),whether making within- or cross-cultural judgments, whereasno analogous associations were made by Americans, whethermaking within- or cross-cultural judgments.

Finally, only one relationship between neatly dressed andpersonality was common across cultures: neatly dressed andCulture (intelligence). For the Chinese, Culture was the onlytrait associated with neatly dressed, whether making within-or cross-cultural judgments. However, among the Americans'judgments of Chinese, neatly dressed was associated with Cul-ture, Conscientiousness, and Emotional Stability. These associa-tions have not been found typically in Western zero-acquaintancestudies (e.g., Americans' within-culture judgments) in whichneatness of dress has usually been strongly associated only withConscientiousness. However, note that in the Americans' judg-ments of Chinese, the strongest correlation with neatly dressedoccurs on Conscientiousness.

Overall, the data concerning the covariation between externalcues and personality trait judgments indicate that there are bothcultural similarities and cultural differences in these belief andmeaning systems. When there were cultural differences, the pat-terns of covariation in the cross-cultural judgments more closelyreflected the pattern of covariation in the within-culture judg-ments; this fact indicates that people do impose their worldviewswhen making intercultural judgments. Clearly, the tendency toimpose one's (different) worldview could produce interculturalmiscommunication and misunderstanding. These data also indi-cate that "they don't [italics added] all look alike" (Zebrawitz,Montepare, & Lee, 1993).

Limitations of This Research

Although these data have important implications for theoryand research on consensus within and across cultures, there areimportant limitations that should be explicit. First, because onlytwo cultures were included in this research, the generalizabilityis limited strictly to Americans and Chinese. Also, because wewere not able to bring the Chinese and American participantsto a single location, judgments across cultures were made byusing photos, whereas within-culture judgments were made ina face-to-face context. We did not compare consensus withinand between culture for this reason. However, it is noteworthythat in spite of the different contexts for the within- and between-culture judgments, the use of cues was similar within eachculture. However, this procedural limitation should be consid-ered in future research. Although we chose traits that were basedon the American Big Five and a Chinese factor structure, thisprincipled approach to trait selection did not yield an equalnumber of traits for each of the five factors.

Conclusion and Implications

This research has demonstrated within-culture consensus inpersonality judgments of members of a very different cultureand between-culture agreement in judgments between membersof two cultures. Further, similarities and differences in the co-variation between perceptions of observable characteristics(e.g., appearance and nonverbal behavior) and personality traitshave been observed between Americans and Chinese. Thesefindings are consistent with both sociocultural (Vygotsky, 1962,1978; Wertsch, 1985) and ecological (Baron & Misovich,1993a, 1993b; Berry & Finch Wero, 1993; McArthur & Baron,1983) theories. That is, many individual-level assumptions thatare applied in everyday life originate from and reflect culturalbeliefs about the social world that have origin in the inherentand real structure of the environment. For example, a smile isgenerally perceived as a nonthreatening social display, becausethe ecologically based belief that smiling signifies benevolentintent has been transmitted and reinforced through socioculturalprocesses. However, to the extent that sociocultural-historicalexperience differs among people, attunement to the stimuluswill differ, in which case a smile may further afford other dispo-sitional properties or states.

The finding of cross-cultural consensus in judgments of Extra-version, in particular, suggests that this trait is afforded throughthe individual's appearance or facial expression and that at-tunement to this information is not moderated by culture. Be-cause the ability to immediately detect information specifyingsocial orientation serves an adaptive function for the individual,direct perception of such dispositional information, as opposedto inference, is necessary. However, the nature of the informationavailable to perceivers in the zero-acquaintance situation or injudging photographic stimuli is characterized by what Baronand Misovich (1993b) called "first-order invariants" (p. 545),or individual-level, as opposed to relational, information.Whereas some traits can be perceived by observing the individ-ual in isolation, others require information derived from socialinteraction. In the present study, the fact that only first-order, orindividual-level, information was available to perceivers may

568 ALBRIGHT ET AL.

explain why there was so much consensus but so little self-other agreement, a result that may indicate a lack of accuracy.

Finally, although research on interpersonal perception withunacquainted individuals provides the necessary control to iso-late and examine a number of basic psychological processes, itis not representative of most of the contexts in which individualsperceive, judge, and form impressions of each other. Represent-ing one interval in time just prior to actual acquaintance, zeroacquaintance is a temporary and relatively stimulus-deprivedcontext, which likely necessitates the use of inferential strategiesthat may or may not facilitate accuracy (see Brunswik, 1956).

References

Albright, L., Kenny, D. A., & Malloy, T. E. (1988). Consensus in person-ality judgments at zero acquaintance. Journal of Personality and So-cial Psychology; 55, 387-395.

Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavioras predictors of interpersonal consequences: A meta-analysis. Psycho-logical Bulletin, 111, 256-274.

Ambady, N., & Rosenthal, R. (1993). Half a minute: Predicting teacherevaluations from thin slices of nonverbal behavior and physical attrac-tiveness. Journal of Personality and Social Psychology, 64, 431-441.

Argyle, M., Henderson, M., Bond, M. H., Iizuka, Y., & Contarello, A.(1986). Cross-cultural variations in relationship rules. InternationalJournal of Psychology, 27, 287—315.

Baron, R. M., & Misovich, S. J. (1993a). Dispositional knowing froman ecological perspective. Personality and Social Psychology Bulletin,19, 541-552.

Baron, R. M., & Misovich, S. J. (1993b). An integration of Gibsonianand Vygotskian perspectives on changing attitudes in group contexts.British Journal of Social Psychology, 32, 53-70.

Bernstein, I. H., Tsai-Ding, L., & McClelland, P. (1982). Cross- vs.within racial judgments of attractiveness. Perception & Psychophys-ics, 32, 495-503.

Berry, D. S. (1990). Taking people at face value: Evidence for the kernelof truth hypothesis. Social Cognition, 8, 343-361.

Berry, D. S., & Finch Wero, J. L. (1993). Accuracy in face perception:A view from ecological psychology. Journal of Personality, 61, 497-520.

Berry, D. S., & McArthur, L. Z. (1986). Perceiving character in faces:The impact of age-related craniofacial changes on social perception.Psychological Bulletin, 100, 3-18.

Bond, C. R, Omar, A., Mahmoud, A., & Bonser, R. N. (1990). Liedetection across cultures. Journal of Nonverbal Behavior, 14, 189-204.

Bond, M. H. (1993). Emotions and their expression in Chinese culture.Journal of Nonverbal Behavior, 17, 245-262.

Borkenau, P., & Liebler, A. ( 1992). Trait inferences: Sources of validityat zero acquaintance. Journal of Personality and Social Psychology,62, 645-657.

Borkenau, P., & Liebler, A. (1993). Convergence of stranger ratingsof personality and intelligence with self-ratings, partner ratings, andmeasured intelligence. Journal of Personality and Social Psychology,65, 546-553.

Brislin, R. W. (1970). Back-translation for cross-cultural research.Journal of Cross-Cultural Psychology, 1, 185-216.

Brislin, R. W. (1980). Translation and content analysis of oral and writ-ten material. In H. C. THandis & J. W. Berry (Eds.), Handbook ofcross-cultural psychology: Vol. 2. Methodology (pp. 389-444). Bos-ton: Allyn & Bacon.

Brunswik, E. (1956). Perception and the representative design of psy-chological experiments. Berkeley: University of California Press.

Campbell, D. T. (1996). Can we overcome world-view incomensurabil-ity/relativity in trying to understand the "other"? In R. Jessor, A.Colby, & R. Shweder (Eds.), Ethnography and human development:Context and meaning in social inquiry (pp. 153-172). Chicago: Uni-versity of Chicago Press.

Cunningham, M. R., Roberts, A. R., Barbee, A. P., Druen, P. B., & Wu,C. H. (1995). "Their ideas of beauty are, on the whole, the same asours": Consistency and variability in the cross-cultural perceptionof female physical attractiveness. Journal of Personality and SocialPsychology, 68, 261-279.

Darwin, C. (1872). The expression of the emotions in man and animals.London: John Murray.

Dion, K. K. (1986). Stereotyping based on physical attractiveness: Is-sues and conceptual perspectives. In C. P. Herman, M. P. Zanna, &E. T. Higgins (Eds.), Physical appearance, stigma, and social behav-ior: The Ontario Symposium (Vol. 3, pp. 7-21) . Hillsdale, NJ:Erlbaum.

Dion, K. K., Pak, A. W., & Dion, K. L. (1990). Stereotyping physicalattractiveness: A sociocultural perspective. Journal of Cross-CulturalPsychology, 21, 158-179.

Eagly, A. H., Ashmore, R. D., Makhijani, M. G., & Longo, L. C. (1991).What is beautiful is good, but. . .: A meta-analytic review of researchon the physical attractiveness stereotype. Psychological Bulletin, 110,109-128.

Ekman, P., Friesen, W. V., O'Sullivan, M., Chan, A., Diacoyanni-Tar-latzis, L, Heider, K., Krause, R., LeCompte, W. A., Pitcaim, T, Ricci-Bitti, P., Scherer, K., Tomita, M., & Tzavaras, A. (1987). Universaland cultural differences in the judgments of facial expressions ofemotion. Journal of Personality and Social Psychology, 53, 712-717.

Funder, D. C , & Colvin, C. R. (1988). Friends and strangers: Acquain-tanceship, agreement, and the accuracy of personality judgment. Jour-nal of Personality and Social Psychology, 55, 149-158.

Gifford, R. (1991). Mapping nonverbal behavior on the interpersonalcircle. Journal of Personality and Social Psychology, 61, 279-288.

Gifford, R. (1994). A lens-mapping framework for understanding theencoding and decoding of interpersonal dispositions in nonverbal be-havior. Journal of Personality and Social Psychology, 66, 398-412.

Gifford, R., Ng, C. E, & Wilkinson, M. (1985). Nonverbal cues in theemployment interview: Links between applicant qualities and inter-viewer judgments. Journal of Applied Psychology, 70, 729-736.

Goldberg, L. R. (1990). An alternative "description of personality":The Big-Five factor structure. Journal of Personality and Social Psy-chology, 59, 1216-1229.

Gordon, I. E., Zukas, M., & Chan, J. (1982). Responses to schematicfaces: A cross-cultural study. Perceptual and Motor Skills, 54, 201 -202.

Hui, C. H., & Yam, Y. M. (1987). Effects of language proficiency andphysical attractiveness on person perception. British Journal of SocialPsychology, 26, 257-261.

Keating, C. F., Mazur, A., Segall, M. H., Cysneiros, P. G., DiVale, W. T,Kilbride, J.E., Komin, S., Leahy, P., Thurman, B., & Wirsing, R.(1981). Culture and the perception of social dominance from facialexpression. Journal of Personality and Social Psychology, 40, 615-626.

Kenny, D. A. (1994). Interpersonal perception: A social relations analy-sis. New "York: Guilford Press.

Kenny, D. A. (1995a). BLOCKO Version V. Unpublished manual, Uni-versity of Connecticut.

Kenny, D. A. (1995b). SOREMO Version V.I. Unpublished manual, Uni-versity of Connecticut.

Kenny, D. A., & Albright, L. (1987). Accuracy in interpersonal percep-tion; A social relations analysis. Psychological Bulletin, 102, 390-402.

CONSENSUS ACROSS CULTURE 569

Kenny, D. A., Albright, L., Malloy, T. E., & Kashy, D. A. (1994). Con-sensus in interpersonal perception: Acquaintance and the Big Five.Psychological Bulletin, 116, 245-258.

Kenny, D. A., Homer, C , Kashy, D. A., & Chu, L. C. (1992). Consensusat zero acquaintance: Replication, behavioral cues, and stability. Jour-nal of Personality and Social Psychology, 62, 88-97.

Kenny, D. A., & La Voie, L. (1984). The social relations model. In L.Berkowitz (Ed.), Advances in experimental social psychology (Vol.18, pp. 142-182). Orlando, FL: Academic Press.

Krauss, R. M., Curran, N. M., & Ferlinger, N. (1983). Expressive con-ventions and the cross-cultural perception of emotion. Basic and Ap-plied Social Psychology, 4, 295-305.

Lam, C. W., & Berrios, G. E. (1992). Psychological concepts and psy-chiatric symptomology in some ancient Chinese medical texts. Historyof Psychiatry, 3, 117-128.

Lau, S. (1982). The effect of smiling on person perception. Journal ofSocial Psychology, 111, 63-67.

Leary, T. (1957). Interpersonal diagnosis of personality—A functionaltheory and methodology for personality evaluation. New \brk: RonaldPress.

Levesque, M. J., & Kenny, D. A. (1993). Accuracy of behavioral predic-tions at zero acquaintance; A social relations analysis. Journal ofPersonality and Social Psychology, 65, 1178—1187.

Maret, S. M., & Harling, C.A. (1985). Cross-cultural perceptions ofphysical attractiveness: Ratings of photographs of Whites by Cruzansand Americans- Perceptual and Motor Skills, 60, 163-166.

Matsumoto, D., & Kudoh, T. (1993). American-Japanese cultural dif-ferences in attributions of personality based on smiles. Journal ofNonverbal Behavior, 17, 231-243.

McArthur, L. Z., & Baron, R. M. (1983). Toward an ecological theoryof social perception. Psychological Review, 90, 215-238.

McArthur, L. Z., & Berry, D. S. (1987). Cross-cultural agreement inperceptions of babyfaced adults. Journal of Cross-Cultural Psychol-ogy, 18, 165-192.

Montepare, J. M., & Zebrowitz, L. A. (1993). A cross-cultural compari-son of impressions created by age-related variations in gait. Journalof Nonverbal Behavior, 17, 55-68.

Montepare, J. M., & Zebrowitz-McArthur, L. (1987). Perception ofadults with childlike voices in two cultures. Journal of ExperimentalSocial Psychology, 23, 331-349.

Morse, S. T, Gruzen, J., & Reis, H. (1976). The "eye of the beholder":A neglected variable in the study of physical attractiveness. Journalof Personality, 44, 209-225.

Norman, W. T., & Goldberg, L. R. (1966). Raters, ratees, and ran-domness in personality structure. Journal of Personality and SocialPsychology, 4, 681-691.

Otta, E., Lira, B. B. P., Delevati, N. M., Cesar, O. P., & Pires, C. S. G.

(1994). The effect of smiling and head tilting on person perception.Journal of Psychology, 128, 323-331.

Paunonen, S. V. (1989). Consensus in personality judgments: Moderat-ing effects of target-rater acquaintanceship and behavior observabil-ity. Journal of Personality and Social Psychology, 56, 823-833.

Peng, Y, Zebrowitz, L. A., & Lee, H. K. (1993). The impact of culturalbackground and cross-cultural experience on impressions of Americanand Korean male speakers. Journal of Cross-Cultural Psychology, 24,203-220.

Quattrone, G. A., & Jones, E. E. (1980). The perception of variabilitywithin in-groups and out-groups: Implications for the law of smallnumbers. Journal of Personality and Social Psychology, 38, 141-152.

Reis, H. T, Wilson, I. M., Monestere, C , Berstein, S., Clark, K., Seidl,E., Franco, M., Gioioso, E., Freeman, L., & Radonane, K. (1990).What is smiling is beautiful and good. European Journal of SocialPsychology, 20, 259-267.

Smith, P. B., & Bond, M. H. (1993). Social psychology across cultures:Analysis and perspectives. Boston: Allyn & Bacon,

Thakerar, J. N., & Iwawaki, S. (1979). Cross-cultural comparisons ininterpersonal attraction of females toward males. Journal of SocialPsychology, 108, 121-122.

Triandis, H. C. (1994). Culture and social behavior. New %rk:McGraw-Hill.

Vygotsky, L. S. (1962). Thought and language (E. Haufman & G. Vahar,Trans.). Cambridge, MA: MIT Press.

Vygotsky, L. S. (1978). Mind in society: The development of higherpsychological processes. Cambridge, England: Cambridge UniversityPress.

Watson, D. (1989). Strangers' ratings of the five robust personalityfactors: Evidence of a surprising convergence with self-report. Jour-nal of Personality and Social Psychology, 57, 120-128.

Wertsch, J. V. (1985). Vygotsky and the social formation of mind. Cam-bridge, MA: Harvard University Press.

Yang, K. S., & Bond, M. H. (1990). Exploring implicit personality theo-ries with indigenous or imported constructs: The Chinese case. Jour-nal of Personality and Social Psychology, 58, 1087-1095.

Zebrowitz, L. A., & Montepare, J. M. (1992). Impressions of babyfacedindividuals across the life span. Developmental Psychology, 28, 1143-1152.

Zebrowitz, L. A., Montepare, J. M., & Lee, H. K. (1993). They don'tall look alike: Individuated impressions of other racial groups. Journalof Personality and Social Psychology, 65, 85-101.

Received April 11, 1996Revision received October 18, 1996

Accepted November 1, 1996 •