Teamwork in the Work Place

10.1177/0022022102239156JOURNAL OF CROSS-CULTURAL PSYCHOLOGYHwang et al. / INDIVIDUALISM-COLLECTIVISM

THE RELATIONSHIP BETWEEN INDIVIDUALISM-COLLECTIVISM, FACE, AND

FEEDBACK AND LEARNING PROCESSES IN HONG KONG, SINGAPORE, AND THE UNITED STATES

ALVIN HWANG Pace University

ANNE MARIE FRANCESCO Hong Kong Baptist University

ERIC KESSLER Pace University

This article focuses on the relationships among face (mianzi), individualism-collectivism, feedback pro- cesses, and learning outcomes in Hong Kong, Singapore, and the United States. An expected effect between individualism and desire to gain mianzi (Mianzigain) was generally confirmed; however, the mirror effect between collectivism and fear of losing mianzi (Mianziloss) was not. As, expected, there was a consistent negative effect between Mianziloss and student question-asking in class (InAsk), but the positive effect from Mianzigain to InAsk was only found in the U.S. sample. Selective effects of feedback forms on learning were highly sensitive to cultural contexts. In the United States, asking questions outside of class was posi- tively related to grades. However, InAsk had a negative effect. In contrast, InAsk was positively related to grades in Hong Kong. For Singaporeans, only checking with students outside of class had an effect on per- formance, but it was negative.

The globalization of business has placed new demands on management thinking and edu- cation, making a basic understanding of cultural differences and their management implica- tions a prerequisite for the global manager or educator. With this in mind, we examined the impact of cultural differences on learning processes, specifically, how differences in individualism-collectivism (IC) affected face concerns and consequent feedback-seeking behaviors and learning performance across three countries: Hong Kong, Singapore, and the United States.

IC

IC has become a key concept in explaining the impact of cultural differences on various organizational behaviors and international management practices (Bond, 1996; Earley & Erez, 1997; Hofstede, 1980; Triandis, 1995; Wagner, 1995). IC has four important attributes (Triandis, 1995): (a) Definition of self emphasizes independence and personal aspects for individualists versus interdependence and group aspects for collectivists; (b) personal goals are more important for individualists, whereas group goals take precedence for collectivists;

72

AUTHORS’ NOTE: Partial funding for this study was provided by a Direct Allocation Grant from the Hong Kong University of Sci- ence and Technology to the second author. The authors would like to thank Karen Phalet and three anonymous reviewers for their helpful comments in developing this article.

JOURNAL OF CROSS-CULTURAL PSYCHOLOGY, Vol. 34 No. 1, January 2003 72-91 DOI: 10.1177/0022022102239156 © 2003 Western Washington University

(c) for individualists, social behavior is determined by attitudes, personal rights, and con- tracts, whereas for collectivists such behaviors are rooted in norms, obligations, and duties; (d) individualists view relationships as rational exchanges, whereas collectivists emphasize the communality of the relationship, even when this represents a disadvantage.

Efforts by researchers to decompose the underlying components of IC have led to many empirical studies in recent years. Researchers have developed different conceptualizations and measures of IC that reflect the complex nature of the concept (cf. Hui, 1988; Singelis, Triandis, Bhawuk, & Gelfand, 1995; Triandis, 1995; Wagner, 1995). Although much has been done, a great deal of controversy still remains. Many IC researchers reject the idea that IC is unidimensional, because categorizing individuals or societies as merely individualistic or collectivistic does not do justice to their full range of behaviors (cf. Kag*ItçIbas7I, 1987; Kag*ItçIbas7I & Berry, 1989; Maznevski, Nason, & DiStefano, 1993; Schwartz, 1990; Sinha & Tripathi, 1994). Thus, although they are still being refined, most conceptualizations and measures of IC reflect this multidimensionalism (cf. Maznevski et al., 1993; Singelis et al., 1995; Wagner, 1995).

Triandis (1995) took the view that individualism and collectivism should be further dis- tinguished by considering a horizontal/vertical (HV) aspect that has been likened by some (cf. Earley, 1997; Earley & Gibson, 1998; Wagner, 1999) as conceptually similar to Hofstede’s (1980) power distance dimension. Triandis (1995) made the HV distinction on the basis of a person’s acceptance of inequality. Those who accept inequality would be clas- sified as vertical, whereas those who do not and who therefore expect equality, would be con- sidered horizontal. Thus, horizontal collectivists (H-C) see themselves as interdependent members of an in-group in which everyone is presumed to be equal, whereas vertical collectivists (V-C), although also seeing themselves as interdependent members of an in- group, accept that some group members have higher status than others. Horizontal individu- alists (H-I) have an independent construal of self and view others as relatively equal in status, whereas vertical individualists (V-I) see themselves as independent and different from others such that inequality would be expected (Singelis et al., 1995).

Another view of IC was presented by Wagner (1995). Based on an exploratory factor analysis of the works of others (Erez & Earley, 1987; Triandis, Bontempo, Villareal, Asai, & Lucca, 1988; Wagner & Moch, 1986), Wagner revealed five IC factors across 20 measure- ment items (see Table 1). These were IC1 (Stand Alone, focusing on individual independ- ence and self-reliance), IC2 (Win Above All, reflecting an all-consuming inclination to win in competitive situations), IC3 (Group Preference, showing a preference to work with others in groups), IC4 (Sacrifice in Group, recognizing that individuals have to make personal sac- rifices in group situations), and IC5 (Individual Thinking, reflecting a need for individual beliefs to be accommodated in group situations). Wagner classified Stand Alone, Win Above All, and Individual Thinking as individualistic factors, and Group Preference and Sacrifice in Group as collectivistic.

If we try to integrate the frameworks presented by Wagner (1995) and Triandis (1995), it seems that Stand Alone reflects horizontal individualism, Win Above All vertical individual- ism, Group Preference horizontal collectivism, and Individual Thinking horizontal individu- alism. The Sacrifice factor does not seem to clearly reflect either individualism or collectiv- ism, and Wagner’s factors do not seem to include the concept of vertical collectivism. It is difficult to synthesize the two models, but we believe this represents a reasonable correspon- dence based on their definitions and factor items.

What, then, are some specific attitudinal and behavioral effects of an individualistic or collectivistic orientation? Persons with a more individualistic orientation tend to engage in

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 73

social loafing behaviors when placed in group situations. However, such individualists would dramatically reduce social loafing behaviors when they perceive that others could trace personal accountability for individual effort in these situations (Earley, 1989). In addi- tion, individualists tend to perform better when they are working alone than in groups, whereas collectivists tend to work better with in-groups than out-groups (Earley, 1993).

The willingness of individualists to reduce social loafing and thus act in a way that is socially acceptable to others in a group situation could possibly be explained by their basic human need for acceptance by others. Individualists may also act in a socially acceptable way due to fear of ostracism or termination from group membership. This basic need for acceptance by others, regardless of IC orientation, has been shown to be present in people across different cultures (Brown & Levinson, 1987; Goffman, 1967; Holtgraves & Yang, 1992; Lerner, 1996). Therefore, an effort should be made to understand how social accep- tance needs influence both individualist and collectivist performance in group situations.

FACE: MIANZI

The need for social acceptance or recognition in group situations is seen in the research on “face,” the image that people strive to maintain before others in pursuit of recognition and inclusion (Hallahan, Lee, & Herzog, 1997). Goffman’s (1967) classic work on face con- sidered the concept in relation to how individuals actively manage verbal interactions with each other so as to take into consideration personal needs of others and self. According to Goffman, all parties to an interaction have an interest in maintaining each other’s face. This is done by articulating threatening requests or responses in a way that would mitigate the risk of rejection. Thus, one might couch ideas in vague forms to determine possible objections before firming them up in ways that would specifically address others’ needs. Or one could use various means of soliciting input from others to take into consideration their personal needs even before articulating any clear ideas.

Use of verbal qualifiers and “prerequest tests” to seek acceptance have been found in cul- tures speaking English, Tamil, Korean, and Javanese languages—all showing a consistent need to manage face (Brown & Levinson, 1987; Lerner, 1996). The presence of such prerequest tests across a range of cultures suggests that face could be an important pancultural construct to explain the desire for social acceptance.

More targeted research on face in the East, specifically among Chinese communities, has shown that face is not a single construct but is instead delineated along two facets, lian and mianzi (Hu, 1944). Whereas lian was described by Hu (1944) as “the confidence of society in the integrity of ego’s moral character” (p. 45), mianzi represented prestige and honor that accrues to a person as a result of successes and possibly ostentatious behaviors before others (Bond & Hwang, 1986).

One could lose lian through immoral behaviors or degrading actions and thus be ostra- cized by others and cut off from all group benefits. The simplest way to avoid losing lian is to adhere to generally accepted norms of behaviors in one’s immediate social setting and to keep in mind broader community expectations. Because of this, it is not surprising that Chi- nese communities tend to emphasize conformance to group expectations, because a person could lose social acceptance and possible material benefits by not doing so.

However, not losing lian does not equate with gaining mianzi, because mianzi can only be attained when others recognize and admire one’s success (Bond, 1996; Gabrenya & Hwang, 1996). To gain mianzi, one has to make an effort to obtain the “admiration” of others by mak- ing one’s successes or abilities known to them, for example, through ostentatious display of

74 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

wealth during social functions or the purchase and display of expensive cars (Gabrenya & Hwang, 1996). Research among Hong Kong Chinese businessmen has also argued for the important role that mianzi plays in sealing business contracts (Redding & Ng, 1982). Thus, face is not only an important psychological need but also has potential business consequences.

RESEARCH HYPOTHESES

IC AND MIANZI

The prominence of face in Asian cultures points to its important social role. It is clear that face requires a collectivistic foundation, because face concerns arise from others’ expecta- tions of self and results in conformance behaviors to gain their recognition. Yet, researchers have found face concerns to be similarly present in both individualistic and collectivistic societies in the East and West (Brown & Levinson, 1987; Holtgraves & Yang, 1992).

For example, in comparing American and Chinese societies, Hsu (1981) suggested that “the Chinese concepts of face and of propriety, and the American sensitivity to prestige and superiority are all familiar expressions of the same need” (p. 110). Earley (1997) argued that with respect to face, cultural dimensions such as IC could provide a context that shaped a pre- disposition to respond to the environment in specific ways. Thus, the concept of face might be manifested in different ways in different societies. In this sense, Ting-Toomey (1988) has suggested that people in more individualistically oriented societies would be more con- cerned with maintaining their own face compared to those in collectivistic cultures where there might be more concern with mutual or other face. Morisaki and Gudykunst (1994) and Kim (1993) have also drawn attention to the influence of IC on face.

A recent study on face in Hong Kong and U.S. journalistic sports writings revealed differ- ent face concerns of writers in these two societies (Hallahan et. al., 1997). U.S. sportswriters focused on face-gaining situations, making strong internal attributions of success. However, their counterparts in Hong Kong tended to explain away face-losing situations by attributing them to external causes. Though not tested, the researchers implied that the different focus between Hong Kong and U.S. sportswriters was due to underlying IC differences. Perhaps more individualistically oriented U.S. sportswriters made internal attributions of success so that the noted athletes could gain recognition before others, whereas the more collectivistic Hong Kong writers were concerned with externalizing failures of athletes in order to help them avoid embarrassment to their in-group.

If this were representative of the larger cultural settings, then people with an individualis- tic orientation would be more driven toward gaining face before others, whereas those with a more collectivistic orientation would be driven toward not losing face. Based on the research that was done by Bond and Hwang (1986), Hallahan et al. (1997), and Redding and Ng (1982), we felt that mianzi (related to gaining or losing prestige and honor) rather than lian (arising from shame or doubt about one’s moral character) was the more appropriate face construct for us to address, as our concern is with how prestige and honor might influence student learning behaviors in different cultural environments. Thus, we propose these two hypotheses regarding the relationships between mianzi and IC:

Hypothesis 1: Individualism is positively related to desire to gain mianzi. Hypothesis 2: Collectivism is positively related to fear of losing mianzi.

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 75

FACE AND FEEDBACK-SEEKING BEHAVIOR

Research has shown that a common expression of face-seeking behaviors is captured in verbal interactions between individuals (Brown & Levinson, 1987; Goffman, 1967). For example, to reduce risk of losing face due to a rejected invitation, a person may make a “preinvitation” to test another person’s likely response before deciding whether to actually extend the invitation (Lerner, 1996). Similarly, research on feedback-seeking behaviors has shown that desire for esteem and image before others (a face concern) creates a significant barrier to communication of negative feedback in the workplace (Ashford 1986; Morrison & Bies, 1991). For example, individuals seeking feedback have been reported to time their feedback-seeking behaviors to obtain favorable responses from their superiors (Larson, 1989; Morrison & Bies, 1991). At the same time, such individuals would avoid feedback information when performing poorly—precisely the times they need it the most (Tesser & Rosen, 1975). Individuals were also known to seek out favorable feedback information from those with reward power, thus managing the enhancement of their image and demonstrating the political nature of the process (Leary & Kowalski, 1990). This challenges the traditional conception that feedback requirements are purely goal driven (Ashford, 1986; Fedor, Rensvold & Adams, 1992), suggesting instead that an unspoken need for managing face before others is also very much a consideration.

Within the learning environment, evidence also indicates that students and trainees have similar face needs. For example, student feedback-seeking behaviors such as asking ques- tions in class have been shown to be negatively influenced by fear of embarrassment (Fassinger, 1995); many students are known to be hesitant or even fearful of asking questions in class (Fassinger, 1995; Ng & Ang, 1997). These anxieties may come from a general fear of embarrassment by students or, as some have found, cultural values of not wanting to lose face in front of others by asking questions that reveal their lack of knowledge (Hwang, 1987; Ng & Ang, 1997).

This concern for face creates a problem for educators, because questioning behaviors have been shown to be positively related to learning outcomes (Carner, 1963; Dillion, 1986; Gall, 1970) such as critical thinking (Smith, 1977), intellectual development and subject interest (Newcastle, 1970), and increased student learning motivation (Chickering, Gamson, & Barsi, 1987). Also, questions may reveal a need for procedural information or clarification and may be used to signal a lack of comprehension or a need for assistance. However, despite the important functions of questioning, fear of losing face may be holding some students back. In contrast, if one were to consider the argument that “ostentatious displays” may be used to gain face (Gabrenya & Hwang, 1996), students who desire to gain face might instead increase feedback-seeking interactions so that they could display what they know before oth- ers. Thus, face concerns may have opposite effects depending on whether one is concerned with gaining or losing face.

In addition to in-class question-asking behavior (InAsk), there are other feedback- seeking processes available to students, such as students asking questions of the professor outside of the classroom (OutAsk) and informal discussions among students without the presence of the professor (OutCheck). Both formal and informal feedback-seeking behav- iors have grown in usage within the learning environment, with some even arguing for greater emphasis on informal channels because of the expanded scope and flexibility that discussions have in meeting a variety of students’ needs for clarification and self-help (Top- ping, 1996). Thus, students do use both formal and informal channels and should be encour- aged to continue to do so in the learning environment.

76 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

Based on research in the learning environment, it would seem that the influence of face on these formal and informal feedback-seeking processes should be examined. Therefore, we suggest the following hypotheses:

Hypothesis 3: There is a negative relationship between fear of losing mianzi and feedback-seeking activities (InAsk, OutAsk, and OutCheck).

Hypothesis 4: There is a positive relationship between desire to gain mianzi and feedback-seeking activities (InAsk, OutAsk, and OutCheck).

Hypothesis 5: There is a positive relationship between the three types of feedback seeking activities (InAsk, OutAsk, and OutCheck) and achievement of learning outcomes (grade performance).

To summarize, this study will first test for the impact of the five IC dimensions on the two mianzi factors: Desire to Gain Mianzi and Fear of Losing Mianzi. The two mianzi factors will in turn be tested for their influence on three feedback-seeking behaviors: InAsk, OutAsk, and OutCheck. As feedback-seeking behaviors have been shown to be important precedents of learning performance (Fassinger, 1995; Topping, 1996), the three feedback- seeking behaviors will also be tested for their impact on student grade performance out- comes.

METHOD

PARTICIPANTS

The sample for the study came from the United States (n = 253), Hong Kong (n = 266), and Singapore (n = 131). All participants were undergraduate business students ranging in age from 18 to 44 (M = 20.8, SD = 2.87). Close to 63% of the sample was female. Nearly all the Hong Kong respondents were born or had at least 10 years’ residency in Hong Kong (97%) with another 2% from China, whereas over 98% of the Singapore respondents were born or had at least 10 years’ residency in Singapore. For the U.S. sample, nearly 92% of the respondents were born or had at least 10 years’ residency in the United States.

MEASURES

Scales from Wagner’s (1995) five IC factors were used to measure individualism- collectivism. The reliability of the measures across the international sample revealed accept- able Cronbach alpha coefficients: Stand Alone = .70 (five items), Win Above All = .73 (five items), Group Preference = .83 (three items), Sacrifice in Group = .79 (four items), and Indi- vidual Thinking = .74 (three items). The face items to measure the Desire to Gain Mianzi, Fear of Losing Mianzi, and the three feedback-seeking measures were adopted from a study of face conducted in Singapore (Ng & Ang, 1997). The Face Gain measure in that study was equivalent to Desire to Gain Mianzi, whereas the Face Loss measure was equivalent to Fear of Losing Mianzi. Cronbach alpha coefficients were .73 (three items) and .82 (three items), respectively. Grade performance was measured by asking students to list their classes and grades for the semester before the study. Grades from the three different grading systems were normalized by country and then included as the performance factor for analysis. Because students were enrolled in different numbers of courses, the four highest normalized grade scores of each student were used as indicators of performance. All the measures except

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 77

grade performance were anchored on a 7-point scale ranging from strongly agree to strongly disagree.

All questionnaires were written in English, the medium of instruction in all the partici- pants’ universities. Also, all participants had either completed the U.S. Scholastic Aptitude Tests or the local equivalent of the British General Certificate of Education (Advanced Level) in English before entry into university. Thus, translation was not necessary for this study.

ANALYSES

In addition to tests for variable reliability, each of the four different group measures, IC, Face, Feedback, and Grade Performance, was factor analyzed to determine internal factor structure stability through a maximum likelihood factor extraction and varimax rotation (described in Table 1 and below). Apart from testing for reliability of variables, a one-way analysis of variance with post hoc comparison of means was used to validate cultural differ- ences across the three countries.

Since IC orientation of people can vary even within a single culture (Triandis et al., 1988), our first step was to run exploratory factor analysis to determine the replicability of Wagner’s (1995) factor structure. In preparation for this analysis, all the data were normalized accord- ing to the Leung and Bond (1989) procedure before a pancultural analysis at the individual level was carried out. This double normalization ensured that country level as well as indi- vidual level biases were removed before the data could be treated as an integrative sample for exploratory factor analysis.

The next major step was to test for significant hypothesized effects in each individual sample and then, based on these significant effects, simultaneously test for significant invari- ant effects across all countries. Although the overall sample of 650 cases was adequate to test hypothesized effects among the 11 constructs based upon 42 response items (response-to- item ratio of 15.4), the number of cases for each individual country did not permit finer grained analyses. The response-to-item ratios were 5.9 for Hong Kong, 4.5 for the United States, and 2.8 for Singapore, not adequate for the recommended range of 5 to 10 times the number of respondents for every item in the desired structural models (Bentler & Chou, 1987).

Consequently, hypothesized relationships were tested using LISREL path models (Jöreskog & Sörbom, 1989) after aggregating item indicators into their respective variables. A two-stage modeling process was undertaken: The hypotheses were first tested for signifi- cance of causal effects and overall model fit within each country, and when these were satis- fied, a cross-sample comparison through an invariance across-the-sample test was applied to the whole data set. During the second-stage invariance analysis, only significant hypothe- sized causal effects (significant gamma and beta relationships) from the first stage that were the focus of this study were made invariant across samples. Correlational relationships among the phi and psi matrices were left free in each sample.

A few fit indices were used to determine individual sample model and across-sample invariance model fit. The traditional chi-square test to estimate the degree of fit between the observed and fitted covariance matrices was used in all the tested models. However, as it is sensitive to sample sizes above 200 and under 100 (Schumacker & Lomax, 1996), other fit indices including the Goodness of Fit Index (GFI), Comparative Fit Index (CFI), and the root mean square error of approximation (RMSEA) were also used to determine model fit.

78 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

Among them, Marsh and Hocevar (1988) indicated that the GFI was one of the more consis- tent estimators under a variety of situations.

RESULTS

An exploratory factor analysis of the IC scales after performing the Leung-Bond normal- ization procedure revealed a five-factor structure that was similar to that in Wagner’s (1995) study (see Table 1).

The total variance explained by these five factors was 54.8%. These five factors were IC1 Stand Alone (Stdalone), reflecting a desire to stand apart from the crowd, IC2 Win Above All (Winall), reflecting an all-consuming inclination to win in competitive situations, IC3 Group Preference (Grpref), reflecting a preference to work with others in groups, IC4 Sacrifice (Sacrif), recognizing that individuals have to make personal sacrifices in group situations, and IC5 Individual Thinking (Indivthnk), reflecting a need for individual beliefs to be accommodated in group situations. A similar factor structure was also seen in the raw data before the Leung-Bond normalization procedure. As further assurance of the stability of the factor structure, country-specific factor analyses produced similar results. Thus, there was considerable assurance on the cross-cultural appropriateness of the IC scale.

The mianzi factors (see Table 1) were also stable in reflecting the two mianzi dimensions, Desire to Gain Mianzi (Mianzigain) and Fear of Losing Mianzi (Mianziloss). The amount of variance explained by the two mianzi factors was 55%. Similarly, the three feedback factors, OutCheck, InAsk, and OutAsk, were stable across the entire sample and accounted for 62% of the item variance in the factor structure variance.

Although we could not perform confirmatory factor analysis on country-specific data due to sample size limitations, the pancultural factor analysis (Leung & Bond, 1989) results indi- cated factor stability for all the constructs and their suitability for individual country analy- sis. Nevertheless, we were conscious that factor structure stability does not equate with simi- larity in factor item values between countries (Byrne & Campbell, 1999).

The next major step in our analysis was to form factor variables to simultaneously test the five hypotheses across the countries. Descriptive statistics for the variables of interest are shown in Table 2. The variables had acceptable Cronbach alpha reliabilities, with most vari- ables in the .80s and .90s and none lower than .72.

A cursory examination of the country means for the IC measures indicates that the pro- files of Hong Kong and the United States were closer together than either was to Singapore’s. Using Wagner’s (1995) interpretation of the five factors, Hong Kong would therefore be the most individualistic and Singapore the most collectivistic sample.

Although there were no hypotheses on cross-cultural IC differences, we were interested in possible differences that have been shown in other studies (cf. Hofstede, 1980). Therefore, a one-way ANOVA test for country differences on each of the five IC dimensions was carried out. This was then followed by Bonferroni post hoc tests to determine specific differences between any two countries. The ANOVA and post hoc tests were done because item score data can vary across cultures despite similar factor structures in the factor analysis stage of a study (Byrne & Campbell, 1999). However, caution is advised in interpreting significant dif- ferences among mean values across the three samples because they could reflect both response tendencies and cultural differences due to lack of score equivalence in the data set. The one-way ANOVA and Bonferroni post hoc test results are shown in Table 3.

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 79

80 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

TABLE 1

Leung-Bond Normalizeda Factor Loadings of Individualism-Collectivism and Other Dimensions

Individualism-Collectivism Factor Loadings

Individualism-Collectivism Factor (r2 = 0.548) 1 2 3 4 5

Stand Alone (α = .70) I1. Only those who depend on themselves get ahead in life. –.11 .00 –.07 .53 –.04 I2. To be superior a person must stand alone. –.05 .08 –.11 .54 –.02 I3. If you want something done right, you’ve got to do it yourself. –.06 .01 –.10 .58 –.03 I4. What happens to me is my own doing. .02 –.01 –.05 .55 –.05 I5. In the long run, the only person you can count on is yourself. .02 .10 –.13 .60 –.03

Win Above All (α = .73) I6. Winning is everything. –.08 .68 –.11 .21 –.03 I7. I feel that winning is important in both work and games. –.04 .58 –.10 .01 –.10 I8. Success is the most important thing in life. –.06 .53 –.01 .08 .01 I9. It annoys me when other people perform better than I do. –.11 .49 –.12 –.08 –.04

I10. Doing your best isn’t enough; it is important to win. –.02 .70 –.07 .01 –.02 Group Preference (α = .83)

I11. I prefer to work with others in a group rather than working alone. –.09 –.12 .81 –.16 .04 I12. Given the choice, I would rather do a job where I can work alone rather

than doing a job where I have to work with others in a group. (reverse scored) –.06 –.19 .63 –.20 –.09

I13. Working with a group is better than working alone. .01 –.14 .80 –.22 .03 Sacrifice in Group (α = .79)

I14. People should be made aware that if they are going to be part of a group, then they are sometimes going to have to do things they don’t want to do. .59 –.04 –.10 –.10 –.06

I15. People who belong to a group should realize that they’re not always going to get what they personally want. .69 –.06 –.07 –.06 –.07

I16. People in a group should realize that they sometimes are going to have to make sacrifices for the sake of the group as a whole. .87 –.15 .04 .00 –.07

I17. People in a group should be willing to make sacrifices for the sake of the group’s well-being. .63 –.08 .03 –.02 –.10

Individual Thinking (α =.74) I18. A group is more productive when its members do what they want to do

rather than what the group wants them to do. –.08 –.10 –.02 –.03 .73 I19. A group is most efficient when its members do what they think is best

rather than doing what the group wants them to do. –.07 –.03 –.03 –.05 .74 I20. A group is more productive when its members follow their own interests

and concerns. –.13 –.03 .03 –.08 .62

Face Factor Loading

Face Factor (r2 = .55) 1 2

Fear Mianziloss (α = .82) F1. I fear making mistakes in class. .73 –.06 F2. I fear others ridicule me. .78 –.03 F3. I am worried that I might be embarrassed in class. .81 –.01

Desire Mianzigain (α = .73) G1. I like to associate myself with people who have prestige or status. –.02 .55 G2. I would like to have a position with high status. –.10 .68 G3. I like for people to think of me as a person having prestige or status. –.02 .86

Singapore (Winall mean = 3.06) was clearly less individualistic than Hong Kong (Winall mean = 4.00) and the United States (Winall mean = 3.93) along the Winall variable, F(2,646) = 28.24. Singapore was also more collectivistic than the other two countries on the Grpref variable, F(2,646)= 11.94. On the Stdalone variable, Singapore was also less individualistic than the United States but was not significantly different from Hong Kong, F(2,646) = 3.23. On the Sacrif variable, all three countries were significantly different from each other with Singapore highest (mean = 5.79), followed by the United States (mean = 5.51), and then Hong Kong, mean = 5.32; F(2, 646) = 14.28. Finally, on the Indivthk variable, Hong Kong was significantly higher (mean = 3.80) than Singapore (mean = 3.08) and the United States, mean = 3.36; F(2, 646) = 28.38.

On the two mianzi variables, Americans were most concerned with Mianzigain (mean = 5.03) and Singaporeans least concerned (mean = 3.42). Regarding Mianziloss, Hong Kong people had the highest mean score (4.95) and Americans the lowest (3.63).

Finally, there were also differences on the three feedback variables. American students were most likely to ask questions in class (mean = 3.59) and Hong Kong students were least likely (mean = 2.75). Hong Kong and U.S. participants were equally more likely than Singaporeans to ask the professor questions outside of class (means of 2.76 vs. 2.36), and Hong Kong students most frequently checked with fellow students outside of class (mean = 3.48), whereas U.S. students were least likely to do this (mean = 2.83).

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 81

Feedback-Seeking Factor Loading

Feedback-Seeking Factor (r2 = .62) 1 2 3

OutCheck (α =.89) Q1. I meet my classmates after class to check on unclear answers. .82 –.01 –.01 Q2. I meet my classmates after class to check on answers to questions

the instructor did not answer adequately. .80 .00 .03 Q3. I meet my classmates after class to check on answers to confusing

questions. .89 .00 .07 Q4. I meet my classmates after class to discuss areas I do not understand. .77 .10 .04

InAsk (α = .84) Q5. I ask questions in class. .00 .10 .72 Q6. In class, I raise problems that I cannot resolve on my own. .03 .06 .87 Q7. In class, I seek clarification on questions in my mind. .07 .11 .66 Q8. I raise issues that are of concern to me in class. .02 .07 .77

OutAsk (α = .85) Q9. I check my views with my instructor after class .00 .62 .14

Q10. I privately seek out my instructor for solutions to questions. .05 .79 .04 Q11. I raise questions with my instructor after class. .08 .79 .13 Q12. I take time to meet my instructor after class .04 .83 .03

Normalized Best Grades (r2= .75; α = .92) S1. Best grade .76 S2. 2nd-best grade .88 S3. 3rd-best grade .94 S4. 4th-best grade .87

NOTE: Bolded figures indicate primary factor loadings of an item. a. Maximum likelihood extraction and varimax rotation after performing Leung and Bond (1989) double normaliza- tion procedure across countries.

TABLE 1 (continued)

We next used LISREL path models to test for significance of hypothesized causal effects1

in each of the three countries. Results from these individual country tests revealed significant causal effects in some but not all areas. Based on the significant results, we proceeded to test for invariant causal effects across the three samples. A series of “stacked” path models (Van de Vijver & Leung, 1997) that were based on the significant beta and gamma effects in the individual country models were tested (Jöreskog & Sörbom, 1989). The first fully causally invariant stacked path model across all three countries showed reasonable fit indices, χ2(124) = 354.73; CFI = 0.72; GFI = 0.88; RMSEA = .10.

82 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

TABLE 2

Descriptive Statistics and Pearson Correlations of Variables by Country

Mean SD α 1 2 3 4 5 6 7 8 9 10

Hong Kong (n = 258) 1. Stdalone 4.49 1.04 .78 2. Winall 4 1.14 .83 .36 3. Grpref 3.97 1.19 .84 –.22 –.14 4. Sacrif 5.32 0.69 .76 .09 .03 –.21 5. Indivthk 3.8 1.23 .82 .12 .17 .09 –.08 6. Mianzigain 4.56 1.07 .79 .18 .39 .01 –.01 .09 7. Mianziloss 4.95 1.17 .82 .04 .15 –.09 –.02 .08 .03 8. InAsk 2.75 0.87 .72 –.01 –.04 –.06 –.01 .00 .05 –.22 9. OutAsk 2.76 1.10 .89 .17 .05 –.03 –.06 .01 .02 –.09 .25

10. OutCheck 3.48 1.16 .92 .05 –.04 .20 –.17 .11 .09 .00 .20 .29 11. NormGrd –0.01 0.92 .95 .07 .04 –.18 –.01 .06 .01 .06 .01 .14 .08

Singapore (n = 124) 1. Stdalone 4.19 1.29 .82 2. Winall 3.05 1.11 .80 .47 3. Grpref 4.63 1.29 .91 –.06 –.14 4. Sacrif 5.79 0.71 .85 .18 .02 –.03 5. Indivthk 3.08 1.14 .72 .12 .19 –.02 –.14 6. Mianzigain 3.42 1.19 .81 .23 .47 .08 .03 .11 7. Mianziloss 4.31 1.42 .90 .22 .07 .06 .03 .02 .12 8. InAsk 3.08 1.17 .92 –.03 .08 –.07 .08 .04 .05 –.48 9. OutAsk 2.36 0.87 .86 .09 .02 .04 .16 .13 –.02 –.08 .39

10. OutCheck 3.25 1.06 .89 –.19 –.10 .09 .00 .05 –.06 .00 .13 .01 11. NormGrd –0.02 0.89 .87 .07 .09 –.21 .09 .07 .11 .08 .01 –.05 –.04

United States (n = 216) 1. Stdalone 4.5 1.31 .81 2. Winall 3.93 1.40 .83 .39 3. Grpref 4.1 1.39 .81 –.13 –.06 4. Sacrif 5.51 1.00 .85 .14 .10 –.01 5. Indivthk 3.03 1.33 .72 .03 .05 .11 –.23 6. Mianzigain 5.03 1.38 .83 .19 .45 .13 .22 –.01 7. Mianziloss 3.63 1.66 .86 .21 .21 –.14 .14 .04 .15 8. InAsk 3.59 1.45 .92 –.05 –.13 .13 .01 .02 .11 –.25 9. OutAsk 2.76 1.29 .91 –.02 –.03 .10 .04 .01 .21 –.02 .40

10. OutCheck 2.83 1.29 .93 –.03 –.04 .27 –.07 .11 .19 .02 .30 .40 11. NormGrd –0.06 0.95 .91 .02 –.05 –.09 –.04 –.14 –.07 .05 .07 .01 .10

NOTE: See text for definitions of acronyms. Bolded coefficients are significant at p < .05, italicized coefficients are significant at p < .10.

Next, these models were examined country by country to determine invariant effects (beta and gamma) that did not fit well across countries. As a result of this, the following causal effects were allowed to vary in the model: the beta effects from the feedback-seeking vari- ables (InAsk, OutAsk, and OutCheck) to grade performance, model χ2(118) = 308.50; the gamma effect from Sacrif (IC) to Mianzigain, model χ2(116) = 293.55; the beta effect from Mianzigain to InAsk, model χ2(114) = 281.37; and finally, the beta effect from Mianziloss to OutCheck, model χ2(112) = 272.55. The final partially invariant model for each of the three countries, χ2(112) = 272.55, CFI = 0.80, GFI = 0.91, RMSEA = 0.088, with significant beta and gamma effects and phi and psi relationships is shown in Figures 1 to 3.

In reviewing the individual models, there were some commonalities across all three coun- tries. Although there were phi relationships that varied across the three countries, there were two common relationships in all of them. Stdalone was positively related to Winall (phis ranged from .34 to .40) and Sacrif (phis ranged from .21 to .49). In examining the

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 83

TABLE 3

Post Hoc Mean Individualism-Collectivism Differences Between Countries

Dependent Mean Variable I Location M J Location M Difference (I–J) Sig. F Value (646,2)

Stdalone Singapore 4.19 United States 4.50 –0.31 .056 3.23 Winall Singapore 3.06 Hong Kong 4.00 –0.94 0.00 28.24

United States 3.93 –0.88 .00 Grpref Singapore 4.64 Hong Kong 3.97 0.66 .00 11.94

United States 4.10 0.54 .00 Sacrif Hong Kong 5.32 Singapore 5.79 –0.47 .00 14.28

United States 5.51 –0.19 .02 Singapore 5.79 Hong Kong 5.32 0.47 .00

United States 5.51 0.28 .01 Indivthk Hong Kong 3.80 Singapore 3.08 0.72 .00 28.38

United States 3.36 0.77 .00

NOTE: Sig. = significance.

GRPPREF

SACRIF

STDALONE

WINALL

OUTASK

MIANZILOSS

MIANZIGAIN

INDIVTHK

OUTCHECK

INASK

GRADE

.24-.23

.19

.28

.11

.17

.17

.36

.21

.17

.15

.16

.19.13

Figure 1: Hong Kong Individualism-Collectivism-Face-Feedback-Grade Model NOTE: Covariances are modeled, but common metric completely standardized coefficients are presented here.

hypothesized causal effects, invariant gamma effects from Stdalone (γ = .19), Winall (γ = .28), and Grpref (γ = .11) were seen to predict Mianzigain across all three countries. Stdalone and Sacrif also predicted Mianziloss with respective invariant gamma effects of .13 and .17. In turn, Mianziloss predicted InAsk with an invariant beta of –.23. The freed correlational psi relationships among InAsk, OutAsk, and OutCheck showed a consistent relationship between InAsk and OutAsk across all three countries (psis ranging between .16 and .48).

In addition, there were effects and relationships that were similar in only two of the three countries. For example, IC variables Grpref and Indivthk were positively related only in Hong Kong and the United States (phis of .15 and .22, respectively). Also, Winall and Indivthnk were positively related only in Hong Kong and Singapore (phis of .17 and .27,

84 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

GRPPREF

SACRIF

STDALONE

WINALL

OUTASK

MIANZILOSS

MIANZIGAIN

INDIVTHK

OUTCHECK

INASK

GRADE

-.23

.28

.11

.17

.40

.41

.27

.41

.24

-.41

.28

.19

.13

Figure 2: Singapore Individualism-Collectivism-Face-Feedback-Grade Model NOTE: Covariances are modeled, but common metric completely standardized coefficients are presented here.

GRPPREF

SACRIF

STDALONE

WINALL

OUTASK

MIANZILOSS

MIANZIGAIN

INDIVTHK

OUTCHECK

INASK

GRADE

.28

.11

.34

.49 .48

.35

.24

.13

-.23

.19

.19

.32

.24

.22 .45

-.23

.17

.14

Figure 3: United States Individualism-Collectivism-Face-Feedback-Grade Model NOTE: Covariances are modeled, but common metric completely standardized coefficients are presented here.

respectively). Sacrif predicted Mianzigain in both Hong Kong and the United States (respec- tive gammas of .17 and .32). On the feedback aspect of the model there were positive rela- tionships between InAsk and Outcheck in Singapore and the United States (respective psis of .34 and .35) and between OutAsk and OutCheck in Hong Kong and the United States (respective psis of .19 and .45).

One of the most interesting differences in the three models is seen in the effect of feedback channel on performance. Grade performance was positively related to InAsk in Hong Kong (β = .24) and OutAsk in the United States (β = .19). In contrast, OutCheck was negatively related to grade performance in Singapore ((β = –.41), and InAsk was negatively related to grade performance in the United States (β = –.23). Other unique effects and relationships included a single positive beta effect from Mianziloss to OutCheck in Singapore (β = .28), a psi relationship between Mianzigain and Mianziloss in the United States (ψ = .14), a beta effect between Mianzigain and InAsk in the United States (β = .24), and a phi relationship between Winall and Sacrif in the U.S. (φ = .24).

DISCUSSION

The aim of this study was to examine a complex linkage of various cultural variables to feedback-seeking behaviors and learning outcomes across three countries. The results have provided partial support for the hypothesized model of interrelationships among the vari- ables and raise interesting questions for future study.

Using Wagner’s (1995) interpretations of the five IC factors, Singapore was shown to be less individualistic than the United States along the three individualistic variables, Stdalone, Winall, and Indivthk, while simultaneously being more collectivistic on the Grpref and Sacrif dimensions. This pattern supported Hofstede’s (1980) findings on the more individu- alistic nature of the United States compared to Singapore. Singapore was also more collectivistic than Hong Kong on Sacrif and Grpref and less individualistic on Winall and Indivthk as well. Thus, Singapore was clearly the least individualistic and most collectivistic country in this study.

An interesting result was the higher Indivthk score for Hong Kong compared to the United States. Thus, along the Indivthk dimension, Hong Kong students in this sample were slightly more individualistic than American students. Although not generally discussed, this is not a new finding. For example, Hui (1988) found similar results in comparing Hong Kong and American students. Ho and Chiu (1994) found both individualistic and collectivistic themes in a content analysis of popular Chinese sayings, and a recent study comparing uni- versity students in the United States, Hong Kong, Australia, and China found that the Chi- nese were more vertically individualistic than the students in either of the Western countries (Chen, Li, Wasti, & Bachrach, 2000). The influence of capitalism and exposure to Western ideas seems to have brought about more individualism in places like Hong Kong, yet the emphasis on teamwork that has developed in American schools and corporations in recent years may have engendered more collectivistic trends there (cf. Hui, 1988).

An examination of the relationships between IC and mianzi indicates that the country pat- terns were in line with traditional expectations. Americans (more individualistic in this sam- ple) had the highest mean Mianzigain and the lowest average Mianziloss scores. This is con- sistent with Hallahan et al.’s (1997) findings on possible relationships between individualism and desire for face. Therefore, Hypothesis 1 was generally supported, as evi- denced by the consistent prediction of Mianzigain by Stdalone and Winall.

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 85

A similar weak but consistent effect from Grpref to Mianzigain was at first puzzling, but upon reflection, it is revealing of the nature of Mianzigain. When both Winall and Grpref were viewed together for their impact on Mianzigain, this suggested that perhaps desire to gain face had two underlying values: the first a motivation to get ahead of others so as to have a basis for recognition (Winall) and the second the desire for the presence of others (Grpref) so as to acknowledge one’s accomplishments. This balancing of the desire to get ahead, an individualistic tendency, against the need for others’ acknowledgement, a group tendency, is illustrated in the Redding and Ng (1982) study in which Hong Kong elites paraded their accomplishments (expensive cars and houses) before others so that they could gain face before them. Yet, they would normally maintain a very focused, individualistic approach in pursuing wealth and success. Thus, Mianzigain may have both an individualistic and collectivistic value basis.

Although recent studies have highlighted the importance of gaining face in collectivistic societies (Bond, 1996; Gabrenya & Hwang, 1996), the consistency of the effects of Winall and Grpref on Mianzigain in all the three samples showed that the face concept could be rele- vant in societies regardless of individualistic or collectivistic inclinations. This is consistent with the extant literature suggesting that face may be a universally relevant concept that is influenced by the IC value orientation of the individual and society (cf. Earley, 1997; Goffman, 1967; Ho, 1976; Ting-Toomey, 1994).

Hypothesis 2 on the relationship between collectivistic values and fear of losing face was generally not supported. Contrary to expectation, we found a consistent Stdalone to Mianziloss effect. One possible explanation for the role of Stdalone (an individualistic fac- tor) in fear of losing face could be found in a study of face by Earley (1997), who pointed out that mianzi is mostly a result of social interaction and can be lost through interacting with others or on a personal basis. Thus, for an individual who has a deep fear of losing face, one natural avenue to prevent that is to avoid others and thus lean toward the individualistic end of the IC orientation. Alternatively, because face loss can only come about as a result of others observing or becoming aware of the face-losing act, those who prefer to do things themselves (individualistically oriented people) would be (a) less likely to have others directly observe their shortcomings and (b) less likely to be blamed for others’ mistakes or shortcomings.

If we take the Wagner (1995) interpretation of Sacrif as a collectivistic factor, then the consistent positive effect on Mianziloss found across countries does provide some support for Hypothesis 2. However, Sacrif also positively predicted Mianzigain in Hong Kong and the United States. Perhaps the nature of the Sacrif factor is not clearly individualistic or collectivistic, or maybe the Mianziloss construct and its consequent still-unexplored rela- tionship to Sacrif does not have a clear IC basis. More studies are needed on Sacrif and Mianziloss to define their nature, boundaries, and relationships to other constructs.

The consistent negative effect of Mianziloss on InAsk in all the three countries showed that fear of losing face may cause students to avoid asking questions in class, regardless of country. This supported Hypothesis 3 and is consistent with findings from the educational literature on the fear of embarrassment arising from question-asking activities in class (Fassinger, 1995).

The positive relationship from Mianzigain to InAsk in the U.S. sample showed that InAsk was a possible avenue for gaining face and recognition before others, at least in the United States. This supported Hypothesis 4, but only in the United States. Although traditional edu- cational research would likely attribute question-asking behaviors to the desire for knowl- edge (Dillion, 1986), the direct relationship from Mianzigain to InAsk showed that gaining face before fellow students and the professor in class may be an additional motive.

86 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

How could questioning behaviors lead to gaining face? One possibility is that questioning behaviors provided opportunities for individuals to show their depth of knowledge through profound questions that others may not have thought of. In so doing, students who raise these questions may gain the admiration of others for their deep knowledge. This logic is consis- tent with Gabrenya and Hwang’s (1996) argument that “ostentatious displays” of one’s capa- bilities and resources are needed for one to gain face before others. When we contrast the stu- dents’ desire to gain face with known educational research findings on student fear of embarrassment from asking questions in class (Fassinger, 1995; Smith, 1977), the emerging finding here is intriguing for educators who have traditionally assumed that question-asking behaviors arise from desire for knowledge.

As further support that face could be a real influence on feedback-seeking behaviors, we draw from the research in the organizational literature that has shown desire for recognition as a motivator of consequent feedback-seeking activities (Leary & Kowalski, 1990; Rao & Schmidt, 1995). The positive Mianziloss to OutCheck effect in the Singapore sample sug- gested that checking with other students for knowledge needs arises out of fear of losing face before others. When considered with the negative Mianziloss to InAsk effect, it implies that Singapore students used this informal channel to meet their need for knowledge rather than the formal channel where there would be a risk of losing face.

Among the feedback-seeking activities, InAsk was positively related to OutAsk in all three countries, although it was related to OutCheck only in Singapore and the United States. OutAsk was also positively related to OutCheck in Hong Kong and the United States. The strong relationships among the three feedback-seeking channels (InAsk, OutCheck, OutAsk) indicated that students in the three countries are likely to use more than one channel to seek feedback. However, InAsk and OutAsk were the channels most likely to be used together. These relationships support arguments that feedback-seeking activities in a learn- ing environment should not only be limited to formal learning channels in class but need to include informal channels to meet students’ needs (Topping, 1996). Therefore, in addition to in-class question-asking behaviors that have been argued to be critical outcome-related, feedback-seeking behaviors (Carner, 1963; Dillion, 1986; Fassinger, 1997), the results point to the need for informal avenues, especially in the United States where professor-to-student interactions outside of class (OutAsk) had a positive impact on grade performance.

In the United States, InAsk had a negative effect on grades, whereas OutAsk had a posi- tive effect. Thus, Hypothesis 5 was partially supported here. However, in contrast, InAsk had a positive impact on grade performance in Hong Kong. The opposite effect of InAsk on grade performance in the United States and Hong Kong implies that in-class questions in Hong Kong led to the desired grade-performance knowledge, whereas this was not so in the United States.

Our experience has been that Hong Kong students ask questions relatively infrequently in class (as reflected in our data and supported by anecdotal evidence from professors in Hong Kong), and those who do tend to receive positive recognition in the form of higher grades. On the other hand, because the general level of in-class question asking is higher in American classes, the frequency of questions need not necessarily be correlated with the importance of those questions. Apart from taking up time, the greater number of questions may lead to loss of focus on the really important information in each lesson and, consequently, affect grades in a negative way (as shown in our data). This contrast is interesting, and further research is needed on how students in Hong Kong and the United States use these three feedback pro- cesses to gain knowledge and how the use of feedback might influence grading.

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 87

In the Singapore sample, it was neither InAsk nor OutAsk that had an impact on grades but rather OutCheck with a negative impact on grade performance. The implication was that checking with fellow students for desired knowledge is not a good move in Singapore; pre- sumably, the knowledge so acquired is inadequate or wrong in its substantive content. The lack of significant effects of InAsk and OutAsk on grade performance also leaves an open question: whether these two feedback channels might be useful to Singapore students for knowledge acquisition. To conclude, differences across countries showed that the impact of feedback forms on learning processes is highly dependent on cultural context.

LIMITATIONS AND CONCLUSION

Although the findings of this study are interesting as seen in some consistent as well as different patterns across countries, there are limitations that need to be acknowledged in this study. First, due to the emerging nature of many of the constructs that were used in this study, we could not clearly test for their external validity against more established constructs. Although internal consistencies and pooled-within factor analysis after the Leung-Bond double normalization procedure indicated reliability and support for internal validity of the constructs, the partially variant models across the three countries (see Figures 1 to 3) showed that these constructs need more testing and study.

Second, as the sample was drawn from three sets of undergraduate students, each from only one university in one country, it is possible that the students themselves were not repre- sentative of students in their respective countries or students in general. Also, the perfor- mance measures obtained, previous semester’s grades, were based on self-reports. So it is possible that students did not report their actual grades accurately, either intentionally or unintentionally. Because the grades were given by three different institutions, we normalized them to make them comparable. However, it is possible that the assumed normal distribu- tions did not, in fact, represent the actual performance of the students in our samples.

Based on the above limitations, future research could perhaps make use of different approaches to performance measurement so that the problem of possible nonequivalent data could be avoided. Perhaps some non-self-reported measures could be used in addition to the self-report measures to assess the reliability of performance outcomes.

To conclude, the contribution of this study has been to provide initial evidence of some underlying cultural influences on feedback and learning processes. The findings suggest that whether in a university or business-training environment, educators need to be sensitive to the face concerns of learners and must provide multiple avenues for them to seek feedback.

NOTE

1. Causal direction in path models is theoretically inferred but empirically undetermined.

REFERENCES

Ashford, S. J. (1986). Feedback-seeking in individual adaptation: A resource perspective. Academy of Management Journal, 29(3), 465-487.

Bentler, P. M., & Chou, C. (1987). Practical issues in structural modeling. Sociological Methods and Research, 16, 78-117.

88 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

Bond, M. H. (1996). Chinese values. In M. H. Bond (Ed.), The handbook of Chinese psychology (pp. 208-227). New York: Oxford University Press.

Bond, M. H., & Hwang, K. K. (1986). The social psychology of Chinese people. In M. H. Bond (Ed.), The psychol- ogy of the Chinese people (pp. 231-266). New York: Oxford University Press.

Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, UK: Cambridge Uni- versity Press.

Byrne, B. M., & Campbell, T. L (1999). Cross-cultural comparisons and the presumption of equivalent measure- ment and theoretical structure: A look beneath the surface. Journal of Cross-Cultural Psychology, 30(5), 555- 574.

Carner, R. L. (1963). Levels of questioning. Education, 83, 54-55. Chen, X. P., Li, S., Wasti, S. A., & Bachrach, D. G. (2000). Are Chinese more vertically individualistic? Evidence

from Hong Kong, the PRC, USA, and Australia. Working paper. Chickering, A. W., Gamson, Z. F., & Barsi, L. M. (1987). Principles of good practice in undergraduate education.

Milwaukee, WI: American Association for Higher Education, the Education Commission of the State, and the Johnson Foundation.

Dillion, J. T. (1986). Student questions and individual learning. Educational Theory, 36, 333-341. Earley, P. C. (1989). Social loafing and collectivism: A comparison of the United States and the People’s Republic of

China. Administrative Science Quarterly, 34, 565-581. Earley, P. C. (1993). East meets West meets Mideast: Further explorations of collectivistic and individualistic work

groups. Academy of Management Journal, 36(2), 319-348. Earley, P. C. (1997). Face, harmony, and social structure. New York: Oxford University Press. Earley, P. C., & Erez, M. (1997). The transplanted executive: Why you need to understand how workers in other

countries see the world differently. New York: Oxford University Press. Earley, P. C., & Gibson, C. B. (1998). Taking stock in our progress on individualism-collectivism: 100 years on soli-

darity and community. Journal of Management, 24(3), 265-304. Erez, M., & Earley, P. C. (1987). Comparative analysis of goal setting strategies across cultures. Journal of Applied

Psychology, 72, 658-665. Fassinger, P. A (1995). Understanding classroom interaction: Students’ and professors’ contribution to students’

silence. Journal of Higher Education, 66, 82-96. Fedor, D. B., Rensvold, R. G. & Adams, S. M. (1992). An investigation of factors expected to affect feedback seek-

ing: A longitudinal field study. Personnel Psychology, 45, 779-805. Gabrenya, W. K., & Hwang, K. (1996). Chinese social interaction: Harmony and hierarchy on the good earth. In

M. H. Bond (Ed.), The handbook of Chinese psychology (pp. 309-321). New York: Oxford University Press. Gall, M. (1970). The use of questions in teaching. Review of Educational Research, 40, 707-721. Goffman, E. (1967). Interaction ritual: Essays on face-to-face behavior. New York: Random House. Hallahan, M., Lee, F., & Herzog, T. (1997). It’s not just whether you win or lose, it’s also how you play the game: A

naturalistic, cross-cultural examination of the positivity bias. Journal of Cross-Cultural Psychology, 28(6), 768- 778.

Ho, D. Y. F. (1976).On the concept of face. American Journal of Sociology, 81(4), 867-884. Ho, D. Y. F., & Chiu, C. Y. (1994). Component ideas of individualism, collectivism, and social organization: An

application in the study of Chinese culture. In U. Kim, H. C. Triandis, C. Kag *ItçIbas77I, S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and applications (pp. 137-156). Thousand Oaks, CA: Sage.

Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage.

Holtgraves, T., & Yang, J. (1992). Interpersonal underpinnings of request strategies: General principles and differ- ences due to culture and gender. Journal of Personality and Social Psychology, 62, 246-256.

Hsu, F. L. K. (1981). Americans and Chinese: Passage to differences (3rd ed.) Honolulu: University of Hawaii Press.

Hu, H. C. (1944). The Chinese concept of “face.” American Anthropologist, 46, 45-64. Hui, C. H. (1988). Measurement of individualism-collectivism. Journal of Research in Personality, 22(1), 17-36. Hwang, K. K. (1987). Face and favor: The Chinese power game. American Journal of Sociology, 92(4), 944-974. Jöreskog, K. G., & Sörbom, D. (1989). LISREL 7: A guide to the program and applications. Chicago, IL: SPSS

Publications. Kag*ItçIbas77I, C. (1987). Individual and group loyalties: Are they compatible? In C. Kag *ItçIbas77I (Ed.), Growth and

progress in cross-cultural psychology (pp. 94-104). Lisse, Netherlands: Swets & Zeitlinger. Kag*ItçIbas77I, C., & Berry, J. W. (1989). Cross-cultural psychology: Current research and trends. Annual Review of

Psychology, 40, 493-531. Kim, M. S. (1993). Culture-based interactive constraints in explaining intercultural strategic competence. In R.

Wiseman & J. Koester (Eds.), Intercultural communication competence (pp. 132-150). Newbury Park, CA: Sage.

Larson, J. R., Jr. (1989). The dynamic interplay between employees’ feedback-seeking strategies and supervisors’ delivery of performance feedback. Academy of Management Review, 14, 408-422.

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 89

Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two-component model. Psychological Bulletin, 107, 34-47.

Lerner, G. H. (1996). Finding face in the preference structures of talk-in-interaction. Social Psychology Quarterly, 59, 303-321.

Leung, K., & Bond, M. H. (1989). On the empirical identification of dimensions for cross-cultural comparisons. Journal of Cross-Cultural Psychology, 20(2), 133-151.

Marsh, H. W., & Hocevar, D. (1988). A new more powerful approach to multitrait-multimethod analyses: Applica- tion of second-order confirmatory factor analysis. Journal of Applied Psychology, 73, 107-117.

Maznevski, M. L., Nason, S., & DiStefano, J. J. (1993, October). Fourteen faces of culture: A new instrument for understanding cultural differences. Paper presented at the Academy of International Business Annual Meeting, Maui, HI.

Morisaki, M., & Gudykunst, W. B. (1994). Face in Japan and the United States. In S. Ting-Toomey (Ed.), The chal- lenge of facework: Cross-cultural and interpersonal issues (pp. 47-93). Albany: State University of New York Press.

Morrison, E. W., & Bies, R. J. (1991). Impression management in the feedback-seeking process: A literature review and research agenda. Academy of Management Review, 16(3), 522-541.

Newcastle, H. (1970). A study of teachers’classroom questioning and responding techniques. Rockville, MD: Edu- cational Resources Information Center (ERIC Document Reproduction Service No. ED 079 260)

Ng, H. A., & Ang, S. (1997, August). Keeping “mum” in class—Feedback seeking behaviors and cultural anteced- ents of “Kiasu” and face in an Asian learning environment. Paper presented at the Academy of Management Conference, Boston.

Rao, A., & Schmidt, S. M. (1995). Upward impression management: Goals, influence strategies, and consequences. Human Relations, 2, 147-167.

Redding, S. G., & Ng, M. (1982). “The role of face” in the organizational perceptions of Chinese managers. Organi- zation Studies, 3, 201-219.

Schumacker, R. E., & Lomax, R. G. (1996). A beginner’s guide to structural equation modeling. Mahwah, NJ: Lawrence Erlbaum.

Schwartz, S. H. (1990). Individualism-collectivism: Critique and proposed reformulation. Journal of Cross- Cultural Psychology, 21, 139-157.

Singelis, T. M., Triandis, H. C., Bhawuk, D. P. S., & Gelfand, M. J. (1995). Horizontal and vertical dimensions of individualism and collectivism: A theoretical and measurement refinement. Cross-Cultural Research, 29(3), 240-275.

Sinha, D., & Tripathi, R. C. (1994). Individualism in a collectivist society: A case of coexistence of opposites. In U. Kim, H. C. Triandis, C. Kag*ItçIbas77I, S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and applications (pp. 123-156). Thousand Oaks, CA: Sage.

Smith, D. G. (1977). College classroom interactions and critical thinking. Journal of Educational Psychology, 69, 180-189.

Tesser, A., & Rosen, S. (1975). The reluctance to transmit bad news. In L. Berkowitz (Ed.), Advances in experimen- tal social psychology (Vol. 8, pp. 193-232). New York: Academic Press.

Ting-Toomey, S. (1988). Intercultural conflict styles. In Y. T. Kim & W. B. Gudykunst (Eds.), Theories in intercultural communication. Newbury Park, CA: Sage.

Ting-Toomey, S. (Ed.) (1994). The challenge of facework: Cross-cultural and interpersonal issues. Albany: State University of New York Press.

Topping, K. J. (1996). The effectiveness of peer tutoring in further and higher education: A typology and review of the literature. Higher Education, 32, 321-345.

Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview. Triandis, H. C., Bontempo, R., Villareal, M., Asai, M., & Lucca, N. (1988). Individualism and collectivism: Cross-

cultural perspectives on self-ingroup relationships. Journal of Personality and Social Psychology, 54, 323-338. Van de Vijver, F. J. R. & Leung, K. (1997). Method and analysis in cross-cultural research. In J. W. Berry, Y. H.

Poortinger, & J. Pandey (Eds.), Handbook of cross-cultural psychology (Vol. 1, pp. 257-300). Boston: Allyn & Bacon.

Wagner, J. A., III. (1995). Studies of individualism-collectivism: Effects on cooperation in groups. Academy of Management Journal, 38(1), 152-172.

Wagner, J. A., III. (1999, August). Ontological individualism-collectivism: A suggested direction for future research. Paper presented at the annual meeting of the Academy of Management, Chicago.

Wagner, J. A., III, & Moch, M. K. (1986). Individualism-collectivism: Concept and measure. Group and Organiza- tion Studies, 11, 280-303.

Alvin Hwang is an associate professor of management at the Lubin School of Business, Pace University. He holds a Ph.D. in management from the University of California, Los Angeles. Dr. Hwang publishes in the behavioral and management development and education areas. His research interests include cross-

90 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

cultural differences in learning and feedback processes, outdoor experiential learning and teamwork, lead- ership comparison across countries, use of learning networks, and stress. E-mail: [email protected].

Anne Marie Francesco is a faculty member in the Department of Management, School of Business, Hong Kong Baptist University, and serves as the associate director of the Bachelor of Business Administration program and coordinator of the Human Resource Management option. She has a Ph.D. in industrial/organi- zational psychology from The Ohio State University. Dr. Francesco is the coauthor of the book International Organizational Behavior: Text, Readings, Cases, and Skills (1998). Her current research interests include cross-cultural management and organizational behavior, life balance, and culture and feedback processes. E-mail: [email protected].

Eric H. Kessler is an associate professor of management at the Lubin School of Business, Pace University, and currently serves as the director of the Lubin Leaders and Scholars Program and vice president of the Eastern Academy of Management. He holds a Ph.D. in organization management from Rutgers University. Dr. Kessler’s research has been widely published in leading management and technology journals and prac- titioner book series. His current scholarship focuses on decision making in organizations, particularly as it applies to the management of innovation, and investigates related phenomena such as accelerated product development, wisdom and knowledge management, learning and alliance strategies, pedagogical uses of technology, virtual teams, and Internet-leveraged innovation processes. E-mail: [email protected].

Hwang et al. / INDIVIDUALISM-COLLECTIVISM 91