Cross-cultural variations in leadership perceptions and attribution of charisma to the leader

Organizational Behavior and Human Decision Processes 92 (2003) 52–66

ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES www.elsevier.com/locate/obhdp

Cross-cultural variations in leadership perceptions and attribution of charisma to the leader Nurcan Ensaria,* and Susan Elaine Murphyb,1 a

California School of Organizational Studies, Alliant International University, 1000 South Fremount Avenue, Unit 5, Alhambra, CA 91803, USA b Department of Psychology, Claremont McKenna College, Kravis Leadership Institute, 850 Columbia Avenue, Claremont, CA 91750-6420, USA

Abstract Leadership perception is based on (a) a recognition-based process that involves categorization of leadersÕ characteristics into relevant stereotypes; and (b) an inference-based process that involves making attributions for leadersÕ characteristics based on outcomes of salient events (Lord & Maher, 1993). The present study examined the interactive effects of these two alternative processes of leadership perceptions on attributions of charisma cross-culturally. Groups of participants from either a collectivistic culture (Turkey) or an individualistic culture (United States) read a vignette about a prototypical or antiprototypical leader (manipulation of recognition-based process) whose company produced a slight or significant increase in sales (manipulation of inference-based process). The results showed that the co-occurrence of these two processes produced optimal attribution of charisma to the leader. In addition, the leadersÕ prototypical characteristics were more effective in forming a leadership impression in an individualistic culture, whereas collectivistic people made attributions based on the company performance outcome. Ó 2003 Elsevier Science (USA). All rights reserved.

Introduction The power of leaders is largely dependent on how they are perceived by others (Hollander & Julian, 1969; Maurer & Lord, 1991; Pfeffer, 1977). The extent to which an individual is perceived as a leader can increase employeesÕ acceptance of organizational decisions and policies, followersÕ organizational commitment, and positive affect among employees (Pfeffer, 1977). Positive perceptions help leaders accentuate their important characteristics to manage their public impressions (Foti, Fraser, & Lord, 1982). Perceptual processes are also important influences on the measurement of leader behavior (Hogg, 2001; Maurer & Lord, 1991). For instance, to select leaders, or to provide developmental feedback to managers, most assessment centers use behavioral measures that can be distorted by observersÕ perceptual inaccuracies (Maurer * Corresponding author. E-mail addresses: [email protected] (N. Ensari), susan.murphy @claremontmckenna.edu (S.E. Murphy). 1 Also corresponding author.

& Lord, 1991; Murphy & Jones, 1993). Therefore, research on leadership perceptions expands our views of how leaders gain and maintain power, helps leaders to improve their abilities to execute and manage their roles in organizations, and guides applied psychologists and managers in the use of behavioral measurements. To understand how people perceive leaders, it is essential to understand how they process information and interpret organizational performance in different cultural contexts. The present study investigated leadership perceptions by comparing recognition-based and inference-based information processing, and primarily examined how their co-occurrence increases attributions of charisma to leaders in collectivistic and individualistic cultures. We sought to obtain experimental evidence for the interactive effects of these two alternative forms of information processing on the attributions of charisma to the leader. The other purpose of this study was to investigate both the moderating and mediating processes involved. In particular, we examined culture as the potential moderator, and dispositional attributions as the potential mediator.

0749-5978/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0749-5978(03)00066-9

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

53

Leadership information processing

Attribution of charisma to the leader

A cognitive-attribution approach has been developed to explain the link between leadership perceptual processes and performance (Lord & Maher, 1993). Research suggests that leadership perceptions are based on both the leaderÕs traits (recognition-based processing) (Lord, De Vader, & Alliger, 1986), and outcome of events (inference-based processing) (Calder, 1977; Meindl & Ehrlich, 1987; Meindl, Ehrlich, & Dukerich, 1985). Recognition-based processing involves categorization of the leader by making a connection between the leaderÕs characteristics and the prototype of a leader held in memory. A prototype of a leader (i.e., an abstract representation of the most representative features of leaders), such as someone who is goal-oriented and intelligent, is accessed from long-term memory when triggered by a stimulus in the environment (Mischel, 1979). ‘‘Once a stimulus person is categorized as the leader, the activated leader prototype causes followers to selectively attend to, encode, and retrieve schema-consistent information, and to provide consistent information where such information does not exist’’ (Kenney, Blascovich, & Shaver, 1994, p. 411; Phillips & Lord, 1982). When there is a good fit between the leaderÕs characteristics and the abstract ideas of what leaders are, then that leader is categorized into the relevant category (Lord & Maher, 1993), perceived as more powerful and charismatic, and given more credit for work outcomes (Lord, Foti, & De Vader, 1984). Both experimental (e.g., Cronshaw & Lord, 1987; Fraser & Lord, 1988; Lord et al., 1984; Maurer & Lord, 1991) and correlational (e.g., Foti et al., 1982) studies indicated that categorization affects perceptions of leaders and descriptions of their actual behavior. Inference-based processing involves making attributions for leadersÕ characteristic based on outcomes of salient events (Lord, 1985; Lord & Maher, 1993; Meindl & Ehrlich, 1987; Meindl et al., 1985; Phillips & Lord, 1981; Rush, Phillips, & Lord, 1981). A groupÕs level of performance is a salient piece of information (Binning, Zaba, & Whattam, 1986). Rush et al. (1981) showed that participants who were told that group performance was good provided higher leadership ratings than those who were told performance was poor. A leader is perceived as charismatic when the business is successful (Shamir, 1992), while business failures are usually attributed to a lack of leadership of the executives, and moreover these failures can detract from the executivesÕ perceived leadership qualities (Lord & Maher, 1993). Repeated organizational successes can enhance the perceptions of CEOsÕ leadership, increasing their opportunities for further successes, whereas failures severely limit the potential actions of CEOs, diminishing their capacity to improve future organizational performance (Lord & Maher, 1993).

Charismatic leadership is a popular and much researched approach to understanding effective leadership (Hughes, Ginnett, & Curphy, 2003). Sociologist Max Weber described charismatic individuals as possessing magical abilities and supernatural power of the mind and speech (Etzioni, 1961). A number of researchers have produced complementary conceptualizations of charismatic leadership (House, 1977; Sashkin, 1988; Shamir, House, & Arthur, 1993) as well as incorporating charisma as a part of transformational theory of leadership (Bass & Avolio, 1994). One of the most useful and well-researched models for studying the behaviors of charismatic leaders and how followers attribute charismatic leadership is Conger and KanungoÕs (1994) theory that focuses on six behavioral factors exhibited by a leader: strategic vision and communication behavior, sensitivity to the environment, unconventional behavior, personal risk, sensitivity to organizational membersÕ needs, and a deviation from the status quo. Past research that utilized this model used both the composite and separate analyses of these behavioral factors (Conger & Kanungo, 1998). This model is important because it focuses on the behavioral aspects of charismatic leadership (such as demonstrating heightened sensitivity to environmental constraints and followersÕ needs, formulating and articulating an ideological vision that is discrepant from the status quo) that form the basis for followersÕ attribution of charisma rather than a set of traits associated with charismatic leadership (Conger & Kanungo, 1987).

The present study Much of the research on charismatic leadership relies on perceptual ratings, and therefore, it is important to understand what factors are likely to contribute to attributions of charisma. Although a recent study found that the type of attributions played a mediating role between a leaderÕs actual behavior and whether he or she was seen as influential, the study did not look specifically at whether attributions affected ratings of charisma (Yorges, Weiss, & Strickland, 1999). Also, no studies to date have specifically examined which types of information processing play a role in affecting attributions of charisma. Thus, the primary purpose of the present study was to examine the interactive effects of inference-based and recognition-based processing on attributions of charismatic leadership. We hypothesized that the cooccurrence of these two processes does not merely produce an additive combination of their individual benefits, but instead, produces an interaction. Specifically, we expected the combined effects of recognitionand inference-based processing to produce an interaction

54

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

such that optimal attributions of charismatic leadership will be formed when the leader is presented as highly prototypical, and when the companyÕs performance outcome is successful (Hypothesis 1). The rationale of our prediction is based on KelleyÕs (1973) attribution theory, which asserts that people try to yield causal explanation for factual outcomes by assembling relevant information in the social environment to reach accurate conclusions (Fiske & Taylor, 1984). If the amount of information level is inadequate, such as when information about an issue is poor or ambiguous, then people are likely to seek out additional information to develop an explanation for the outcome (Kelley, 1973). However, when an effect occurs in the presence of a facilitative cause that provides relevant and plausible information justifying the effect, the process of searching for meaningful explanations for the effect becomes easier. Therefore, people tend to give the facilitative cause more extreme weight in producing the outcome (Kelley, 1973). According to KellyÕs principle, when the given characteristics of a leader are contextually appropriate and congruent with the successful outcome (i.e., when the leader is prototypical), this information provides a justification mechanism such that people are more likely to make attributions of charismatic qualities to the leader when they perceive his or her characteristics to be prototypical. Thus, we predict that when two facilitative pieces of information (the leader as highly prototypical, and the companyÕs successful performance outcome) are presented to the perceiver at the same time, then they will cause more extreme attributions of charismatic leadership and create an interactive effect. We do not expect, however, that this principle applies in conditions where the two facilitative pieces of information do not provide a justification for one another. For example, when the company makes a high profit, but the leader is presented as being antiprototypical, then the perceiver encounters a facilitative (successful outcome), and an inhibitory (antiprototypical leader) explanation. In this case, we expect that the presence of inhibitory information attenuates the impact of the facilitative information, and thus decreases attribution of charisma to the leader as compared to a situation where both types of information are facilitative. The present study is important in three ways. First, prior research provides no experimental confirmation of the interactive effect of inference and recognition-based processes within leadership literature. Research in the area of leadership perceptions has focused solely on the independent contributions of these two informationprocessing methods rather than a combination (e.g., Lord et al., 1984, 1986; Meindl & Ehrlich, 1987; Meindl et al., 1985). Why is the examination of the interactive effects of these two approaches important? Understanding simultaneous modes of information processing is important within more comprehensive models of

information processing. Recent theorizing by Lord, Brown, Harvey, and Hall (2001) and Hanges, Lord, and Dickson (2000) sheds light on leadership perceptual processes such as simultaneous information processing within a connectionist architecture. This model of information processing replaces the more static and serial processing with a dynamic model. For example, according to the connectionist approach, leader schemata are constructed each time a leader is perceived rather than a person holding a static schema. Additionally, schemas for leaders do not exist as separate non-overlapping structures, but instead the schema for a leader shares processing units that are weighted to produce a particular schema. Therefore, the model points to the need for studies that uncover these complicated effects. Secondly, the present study investigates information processing specifically within the context of charismatic leadership. Although followerÕs perceptions play a large role in determining leadership influence, additional understanding of what types of information processing affect the attribution of charismatic leadership will help researchers delineate true variation in charismatic behavior from attributional effects. The romance of leadership theory (Meindl et al., 1985) presents a potential problem for many theories of leadership that rely solely on leadership ratings. It argues that leaders have little influence on performance, and people have difficult time delineating the true effects leaders have on organizations (Calder, 1977; Meindl & Ehrlich, 1987; Pfeffer, 1977). People are biased to attribute more responsibility to leaders for organizational performance while disregarding the influence of external determinants that are beyond the leaderÕs control, such as economic conditions and governmental policies (Meindl et al., 1985). Although there are studies that examined how the attribution of charisma is affected by the tendency to overattribute characteristics to leaders based on organizational outcomes (Awamleh & Gardner, 1999; Pillai & Meindl, 1991; Shamir, 1992), past research did not simultaneously examine the effects of prototypicality and outcomes on charisma attributions. Finally, the present study will take into account the context of culture as a backdrop for the two types of information processing of charismatic leadership. A large contribution of the connectionist model of information processing is that it allows for larger consideration of situational factors such as culture (Hanges et al., 2000) and other context effects (Lord et al., 2001). Next, we discuss the moderating effects of culture, and its effects on attribution of charisma. Cross-cultural variations in the processes of leadership perceptions Cultural factors affect the basic processes underlying leadership relations (Markus & Kitayama, 1991).

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

Leadership prototypes and conceptions vary from one culture to another (Bass, 1990), and they have dramatically different implications in different cultural contexts (Chemers & Ayman, 1993). Specific research has looked at how dimensions of leader behaviors, such as task structuring behavior, might remain the same across cultures, but the specific behaviors within that dimension varying in meaning (Misumi, 1985). For example, Smith and Peterson (1988) showed that with respect to the Performance–Maintenance (PM) model of leadership (Misumi, 1985), criticizing someone privately was seen as considerate in the US (individualistic culture) but inconsiderate in a collectivist culture (Japan). One of the most comprehensive studies of cross-cultural leadership, the GLOBE study, has sought to understand how implicit leadership theories vary across cultures (Den Hartog, House, Hanges, Ruiz-Quintanilla, & Dorfman, 1999; House et al., 1999). In general, this research reveals large common themes, but as in previous research, more specific differences regarding the detail of leader behavior. The GLOBE study found that three factors of leader behavior, Charismatic/Value Based, Team-Oriented, and Participative, were prototypical for leaders across 60 different cultures. This is consistent with other studies (Bass, 1997; Bass & Avolio, 1993; Carl & Javidan, 2001) that showed that transformational leadership components, including charisma, are found consistently across many cultures, and that charismatic leadership is valued in many different countries. The cross-cultural application of charismatic and transformational leadership coupled with findings of specific behavioral differences are best summarized by Smith (1997) who writes, ‘‘Charisma may be best thought of as a quality that is global but impute to leaders on the basis of behaviors that are culturespecific’’ (p. 628). Past cross-cultural leadership studies, however, have not examined specific variations in perceptions of a leader with respect to information processing (Lord & Emrich, 2001). Culture undoubtedly affects the content of the leadership prototype and information processing of leader behavior (Hanges et al., 2000; Lord et al., 2001; Shaw, 1990). Thus, our purpose was to examine cultural variations in leadership perceptions by focusing on the information processing aspects of culture, and to improve our understanding of the impact of leadership on culture, and how culture, in turn, influences leadership (Lord & Emrich, 2001). The present study specifically examined differences between individualistic vs. collectivistic societies on the social cognitive processing of leadership perceptions. In individualistic cultures, there is a greater possibility that causal attributions will be made to dispositional factors (Markus & Kitayama, 1991) where the cause of peopleÕs successes and failures typically is to be found within themselves (Moghaddam, 1998). One explanation for

55

this tendency is that trait concepts are highly accessible for individualistic people, therefore, they are more likely to apply dispositional characteristics in interpreting outcomes (K€ uhnen, Hannover, & Schubert, 2001). On the other hand, dispositional explanations are less accessible for collectivistic people, thus they tend to underestimate dispositional causes in explaining events (K€ uhnen et al., 2001). They derive much of their self-esteem from group identity that forms their ‘‘interdependent self, and make attributions to collective activities or groups (Markus & Kitayama, 1991). Therefore, individualistic cultures focus on the leader as the primary determinant of subordinate motivation and performance, whereas peer pressures and collective action are stronger influences than leadership in collectivistic cultures (Hofstede, 1980, 1984). Experiencing interdependence in a collectivistic culture entails seeing the leader as part of an organization structure, and recognizing that his or her behavior is determined and contingent on organizational, collective actions. In accord with this view, we expected that people in individualistic cultures will make their judgments based on the leadersÕ characteristics and behaviors such that charismatic leadership will be perceived through reference-based processing (Hypothesis 2). Alternatively, collective outcomes will be the focus of attention in making judgments in collectivistic societies (i.e., inference-based processing) (Hypothesis 3). In individualistic cultures where the leaderÕs image is highly consistent with the prototypes, we expect a stronger tendency toward attributions of charismatic qualities to the leader than when the leader was antiprototypical. That is, prototypes of the leader are more effective in forming a leadership impression than basic factual outcomes in individualistic countries. Conversely, collectivistic participants will attribute the causes of outcomes to the collective action and thus make attributions of charismatic qualities to the leader based on organizational outcomes. Mediating and moderating role of the dispositional attributions The present study also explored whether dispositional attributions mediated the cross-cultural effects of recognition-based processing on attributions of charisma to the leader. As indicated, individualistic cultures are more likely to make attributions to dispositional factors than collectivistic cultures (Markus & Kitayama, 1991). Phillips and LordÕs (1981) study showed that American participants merely go to a simplified dispositional processing as long as the leaderÕs behavior is typical and fits within their implicit view of leadership. Further, KelleyÕs (1973) attribution theory suggests that people acting in a typical manner are seen as more responsible, while those acting in an antiprototypical manner are

56

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

seen as distinctive and therefore lead to an external attribution. Correspondingly, the extent to which people in an individualistic culture make dispositional attributions for the companyÕs outcomes, the stronger the prototypicality (recognition-based processing) manipulation will be. Thus, it was predicted that the effect of recognition-based processing on attributions of charisma to the leader should be mediated by the extent to which participants in an individualistic culture made attributions to dispositional factors (Hypothesis 4). A final hypothesis, suggested by the previous research work of Phillips and Lord (1981), is that the more likely a person is to make dispositional attributions, the more likely this will interact with performance outcomes and increase attributions of charisma. We expected this to be true across both cultural groups (Hypothesis 5).

Method Pilot study Knowledge structures that exist in long-term memory (i.e., schemas, stereotypes, prototypes etc.) guide information processing in several domains, such as problem solving, social perceptions, and social interactions (Lord & Maher, 1993). These knowledge structures are generated by automatic processes and they offer insight to the nature of culture, because culture specifies what behaviors are appropriate or inappropriate for leaders (Schein, 1985). Therefore, as a first step, we examined the cultural differences in leadersÕ prototypical behaviors in an attempt to select protototypical and antiprototypical behaviors of leaders in collectivistic and individualistic cultures. The other purpose of this pilot study was to distinguish ideal, effective leader behaviors from prototypical, representative leader behaviors. People may organize experiences in terms of cognitive categories based on ideals rather than prototypes (Barsalou, 1985). In addition, peopleÕs prototypes of leaders can be confounded

with effectiveness. That is, prototypical leader behaviors may be perceived as more effective and favorable than non-prototypical leader behaviors (Foti et al., 1982). The pilot study allowed us to make such a distinction by asking participants to rate both the prototypicality and the effectiveness of the leader behaviors, and to control the extremity of effectiveness by carefully selecting moderately effective prototypical and antiprototypical leader behaviors. To select a balanced set of two moderately effective prototypical and antiprototypical leader behaviors, an independent sample of 49 American students and 26 Turkish students rated both the prototypicality and the effectiveness of each of the 76 leader behaviors used in past research (Cronshaw & Lord, 1987; Foti et al., 1982; Kenney et al., 1994; Lord et al., 1984; Phillips & Lord, 1982). Specifically, they were asked to indicate how well each behavioral statement fit their image of a leader (i.e., the level of prototypicality), and the extent to which each of the leader behaviors is prototypic of effective leadership (i.e., the level of effectiveness) using a 5-point scale (1 ¼ not at all to 5 ¼ extremely). Among these 76 statements, 4 prototypical and 4 antiprototypical behaviors were selected separately for the two cultures (see Appendix A). The mean of the 4 prototypical behaviors used in the American version of the vignette (M ¼ 3:31) was greater than the mean of the 4 antiprototypical behaviors (M ¼ 2:72, tð47Þ ¼ 4:86, p < :001) (see Table 1). Similarly, Turkish students rated the 4 prototypical behaviors to be more prototypical (M ¼ 3:85) than the antiprototypical behaviors (M ¼ 2:80, tð25Þ ¼ 4:89, p < :001). Further, all of these behaviors were rated as being moderately effective in both cultures (p < :05). These results support Foti et al.Õs (1982) study that showed that prototypicality is indeed measuring something besides favorability or effectiveness. Participants A total of 187 undergraduate students (87 American students from a small liberal arts college representing

Table 1 Mean prototypicality and effectiveness of the selected prototypical and antiprototypical leader behaviors in the individualistic (US) and collectivistic (Turkey) cultures Culture

Dependent measure

Condition

Mean

SD

Individualistic

Prototypicality

Prototypical Antiprototypical

3.31 2.72

.58 .76

Effectiveness

Prototypical Antiprototypical

3.02 2.88

.53 .67

Prototypicality

Prototypical Antiprototypical

3.85 2.80

.64 .64

Effectiveness

Prototypical Antiprototypical

3.62 3.64

.51 .64

Collectivistic

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

the individualistic culture, and 100 Turkish students from the Bogazici University representing the collectivistic culture) participated in this study for partial credit in their psychology course. Design The design constituted a 2 (recognition-based processing: prototypical vs. antiprototypical leader behavior)
2 (inference-based processing: high vs. low outcome)
2 (culture: individualistic vs. collectivistic) between subjects factorial design. Participants were randomly assigned to prototypical/antiprototypical and high/low outcome conditions prior to the experiment. Procedure The study was conducted during class time. The participants received a questionnaire containing instructions, followed by a vignette, and questions regarding the vignette on the next page. The instructions indicated that this study is a large-scale study designed to investigate how people perceive leaders, and asked the participants to read the story about John Smith (or Ahmet Acar in the Turkish version) carefully, and then complete the questions without referring back to the story. All instruments within the questionnaire used in Turkey were translated into Turkish by a professional translator, and then back translated by the first author who is a native speaker to ensure equivalence. Prototypicality was manipulated by using either the prototypical or antiprototypical leader behaviors incorporated into a stimulus vignette (similar to Fraser & Lord, 1988; Lord et al., 1984; Shamir, 1992) (see Appendix A for the prototypical and antiprototypical items used in the vignettes, and Appendix B for the original vignettes used in the US and Turkey). The performance outcome was manipulated by inserting into the vignettes one of the following bogus pieces of information on the companyÕs performance: a slight (5%) (low performance outcome) or high (25%) increase (high performance outcome) in sales performance. The performance outcome manipulation was adopted from the vignettes used in Meindl et al. (1985) and Shamir (1992). Measures Dispositional attribution After reading the vignettes, participants rated the extent to which John Smith (or Ahmet Acar) was responsible for the performance outcome of the company (dispositional attribution) on a single 7-point scale (1 ¼ not at all, 7 ¼ extremely). This measure has been typically used as a measure of internal/dispositional attributions in the literature (e.g., Hewstone, 1989; Weiner, 1974).

57

General leadership impression Participants assessed whether the leader in the vignette was exhibiting leadership behavior. The one item (‘‘How well does John Smith (or Ahmet Acar) fit your image of a typical leader?’’) was modeled after other studies that used measures of general leadership impression (Maurer & Lord, 1991). Charismatic leadership perceptions Participants rated their perceptions of the leaderÕs charisma on a six-item scale based on the six dimensions of Conger and Kanungo (1994) CK scale for charismatic leadership. The CK Scale is a well-respected measure of charisma that is highly correlated with other measures of charisma or inspirational leadership (Anderson & Wanberg, 1991; Conger & Kanungo, 1994). For the purposes of the present study, using the full 25-item scale was not feasible given the brief leader vignette, so we modified the scale to only include the main dimensions of charisma: inspiration, vision, environmental sensitivity, group member sensitivity, personal risk, and unconventional behavior. Others (e.g., Maurer & Lord, 1991; Shamir, 1992) have used specially constructed scales to measure perceived leader behavior in research using vignettes or videotapes. These scales are modified to reflect the typical leader behavior content in a laboratory simulation. The questions represented the attribution of charisma were the followings: (1) ‘‘What is the likelihood that John Smith (or Ahmet Acar) is an inspirational leader?’’ (2) ‘‘What is the likelihood that John Smith (or Ahmet Acar) has a vision for his work groupÕs future?’’ (3) ‘‘What is the likelihood that John Smith (or Ahmet Acar) may readily recognize constraints that may stand in the way of his company meeting its objectives?’’ (4) ‘‘What is the likelihood that John Smith (or Ahmet Acar) easily recognizes abilities and skills of his subordinates?’’ (5) ‘‘What is the likelihood that John Smith (or Ahmet Acar) takes personal risks for the sake of his company?’’ (6) ‘‘What is the likelihood that John Smith (or Ahmet Acar) is likely to engage in unique, unconventional behaviors in order to achieve companyÕs goals?’’ These six items were combined to form a total charisma score. Research on charismatic leadership using the full 25-item CK scale has examined both the composite of the scale and the sub-dimensions in answering research questions (Conger & Kanungo, 1998). One study reported correlations between the items and the total scale ranging from .42 to .64, except for the items measuring unconventional behavior which are uncorrelated (Conger & Kanungo, 1994). The resulting CronbachÕs alpha for the modified six-item scale used in this study was .65. This is lower than the typical .90 reported by Conger and Kanungo for the full 25-item scale. We conducted a confirmatory factor analysis for a single factor for the six charisma items using AMOS

58

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

(Arbuckle & Wothke, 1995). The one factor model produced good fit indices CFI ¼ .96. NFI ¼ .91 and RMSEA ¼ .06, with a X2 (9, N ¼ 185) ¼ 14.03, p ¼ :12. Therefore, the six items were combined and examined as a composite rather than six single item scales. Manipulation check measures After the participants completed the leadership ratings, they were given 7-point scales (1 ¼ not at all, 7 ¼ extremely well) to check the manipulation of prototypicality and performance outcome. They indicated how well John Smith (or Ahmet Acar) fit their image of a typical leader, and how well his company performed this year. Individualism–collectivism Using a 20-item individualism–collectivism scale (Triandis, Bontempo, Villareal, Asai, & Lucca, 1988), participants indicated the extent to which they agreed on a 7-point scale with statements designed to measure individual orientations of the participants along the individualism–collectivism dimension. The reliability analysis of the response of the 187 participants indicated that this scale was highly reliable (a ¼ :82). To measure individual orientation of each participant along the individualism–collectivism dimensions, the items were averaged to form a single individualism–collectivism score for each participant (higher scores indicated greater individualistic orientation). Next, the participants were divided into two groups (individualistic vs. collectivistic) based on a median-split method (median ¼ 3.55). The resulting split fell along cultural groups. Of those in the high individualistic group, 0% were from Turkey, and 97.7% were from the US.

Results Manipulation checks Manipulation check of prototypicality All inferential tests were evaluated using p value of .10. Results of a 2 (prototypicality)
2 (outcome)
2 (culture) ANOVA with the manipulation check item as the dependent variable revealed only a main effect for prototypicality, F ð1; 179Þ ¼ 14:21, p < :01, g2 ¼ :08. The leader behaviors in the prototypicality condition were rated as more prototypical (M ¼ 4:88) than in the antiprototypical condition (M ¼ 4:17). Clearly, the prototypicality manipulation was successful. We also performed the same analysis with the individualism– collectivism scores as a covariate, however, the results remained the same (i.e., the main effect for prototypicality was significant, F ð1; 174Þ ¼ 13:10, p < :01, g2 ¼ :07). In other words, the participantsÕ general impressions of whether the leaderÕs behavior was typical

was not affected by the other experimental manipulations. Another manipulation check of prototypicality concerned the degree to which Turkish and American students perceive the prototypical and antiprototypical leader as charismatic. We collected additional data from both cultures to test whether the leader behaviors in the vignettes included only prototypical (or antiprototypical) leader descriptions, or they also included the breath of charismatic behaviors. To be able to test our hypotheses, attribution of charisma should not be influenced by the degree of charismatic behaviors embedded in the items, rather it is the prototypicality of the leader, as we predicted, that influenced the charisma attributions. For that reason, a questionnaire was distributed to 41 students (21 Turkish and 20 American). This questionnaire included a total of 12 leader behaviors (6 prototypical and 6 antiprototypical) that were used in the vignettes in the original experiment (see Appendix A). The students were asked to read the given leader behaviors carefully, and then complete the questions without referring back to the previous item. The 6 questions aimed to assess attribution of charisma and were the same as the charisma measure used in the original experiment (see above). Before the analysis, we first combined the 6 prototypical items (a ¼ :75), and the 6 antiprototypical items (a ¼ :74). Next, we performed a 2 (prototypicality: prototypical vs. antiprototypical items)
2 (culture: American vs. Turkish) ANOVA factorial with the first factor manipulated within subjects. The results showed that there was neither a main effect of prototypicality (M ¼ 4:07 for the prototypical items and M ¼ 3:95 for the antiprototypical items, F ð1; 39Þ ¼ :44, p > :05), nor a main effect of culture (M ¼ 4:12 for Americans and M ¼ 3:90 for Turkish students, F ð1; 39Þ ¼ :65, p > :05). Thus, students did not differentially perceive the prototypical and antiprototypical leaders as charismatic, and American and Turkish students did not differ in how they perceive the leader as charismatic. In sum, the vignettes used to characterize prototypical or antiprototypical leader did not vary as a function of the amount of charismatic leader behavior they contained. Manipulation check of outcome To assess the effectiveness of the outcome manipulation, participants were asked to indicate how well the leaderÕs company performed. Confirming a successful manipulation of outcome, the results of a 2 (prototypicality)
2 (outcome)
2 (culture) ANOVA with the outcome manipulation item as the dependent variable revealed only a main effect for outcome, F ð1; 179Þ ¼ 109:17, p < :001, g2 ¼ :38. Participants correctly perceived the performance outcome of the company as higher in the high outcome condition (M ¼ 5:97)

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

than in the low outcome condition (M ¼ 4:45). We also performed the same analysis with the individualism– collectivism scores as a covariate, however, the main effect for performance outcome remained significant, F ð1; 178Þ ¼ 108:58, p < :001, g2 ¼ :38.

59

when the leader was highly prototypical, the high outcome led to higher attribution of charisma (M ¼ 5:13, SD ¼ :72) than the low outcome (M ¼ 4:70, SD ¼ :76, tð96Þ ¼ 2:86, p < :001) (see Fig. 1). This difference, however, disappeared under the antiprototypical condition (M ¼ 4:62, SD ¼ :74 in the high outcome, and M ¼ 4:77, SD ¼ :82 in the low outcome conditions,

Attribution of charismatic leadership In the present study, it is also important to determine whether information about prototypicality of behavior and outcomes truly have an effect on attributions of charisma beyond general impressions of leadership. Therefore, we predict that Hypothesis 1 should hold true even when we control for a perceiverÕs rating of whether the target leader is generally seen as leader-like. In other words, sometimes just having the leader assume a leadership role is enough to make attributions of specific leader behaviors whether or not the leader displayed the behaviors. By controlling for this general leadership perception, the remaining variance in charisma should be due to the specific manipulations of prototypicality and outcomes. Previous studies used a measure that asked how much the leader behaved as a leader (General Leadership Impression—GLI) as a check on whether or not the effects they obtained for manipulations of leader behaviors affected merely the perception of being leader-like or more importantly, affect actual perceptions of specific leader behaviors (Lord et al., 1984; Maurer & Lord, 1991). We found in the present study that GLI was correlated with charisma ratings, rðN ¼ 185Þ ¼ :29, p < :001 across all conditions. Therefore, we have included GLI as a covariate in our experiment to determine whether the attributions to charisma for the experimental manipulations are beyond that of just a general impression of leadership. A 2 (prototypical vs. antiprototypical leader)
2 (high vs. low outcome)
2 (individualistic vs. collectivistic culture) ANCOVA factorial design using GLI as a covariate yielded a two-way interaction between prototypicality and outcome on ratings of leader charisma, F ð1; 176Þ ¼ 7:26, p < :01, g2 ¼ :04 (see Table 2 for the full ANCOVA results). In accord with the Hypothesis 1,

Fig. 1. Shape of interactive effects on attribution of charisma.

Table 2 Analysis of covariance on attribution of charisma to the leader (N ¼ 184) Source

Sum of squares

df

General leadership Impression covariate prototypicality Outcome Culture Prototypicality
outcome Prototypicality
culture Outcome
culture Prototypicality
outcome
culture

6.002 1.01 .34 .27 3.86 1.62 2.36 .00

1, 1, 1, 1, 1, 1, 1, 1,

176 176 176 176 176 176 176 176

Mean square

F

p

g2 effect size

6.002 1.01 .342 .27 3.86 1.62 2.36 .00

11.30 1.91 .65 .51 7.26 3.05 4.44 .001

.00 .17 .42 .47 .01 .08 .04 .97

.060 .011 .004 .003 .040 .017 .025 .000

60

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

p > :05). This interaction was also significant across the two cultures, F ð1; 81Þ ¼ 3:69, p ¼ :06, g2 ¼ :04 in the individualistic culture, and F ð1; 96Þ ¼ 2:95, p ¼ :07, g2 ¼ :03 in the collectivistic culture. Hence, as hypothesized, the co-occurrence of prototypicality and high performance outcome produced optimal attributions of charisma. The results also revealed an interaction between outcome and culture, F ð1; 176Þ ¼ 4:44, p < :05, g2 ¼ :025, and an interaction between prototypicality and culture, F ð1; 76Þ ¼ 3:05, p ¼ :08. g2 ¼ :02. As shown in Fig. 1, individualistic students attributed charisma to the prototypical leader (M ¼ 5:12, SD ¼ :74) more than the antiprototypical leader (M ¼ 4:68, SD ¼ :86, tð83Þ ¼ 2:51, p ¼ :01), For collectivistic students, the prototypicality of the leader did not play a role in making attributions of charisma (M ¼ 4:79, SD ¼ :75, and M ¼ 4:73, SD ¼ :73, respectively). These results support Hypothesis 2. On the other hand, as expected, collectivistic students made their judgments based on the outcome of the company (see Fig. 1). When the company made a high profit (M ¼ 4:93, SD ¼ :69), study participants made higher attributions of charisma than when the companyÕs outcome was low (M ¼ 4:58, SD ¼ :75, tð98Þ ¼ 2:43, p < :05) providing support for Hypothesis 3. Individualistic students, on the other hand, made equal attributions of charisma to the leader in high (M ¼ 4:92, SD ¼ :85) and low outcome conditions (M ¼ 4:92, SD ¼ :80, p > :05). In conclusion, individualistic culture focused on the leaderÕs behaviors more than the companyÕs outcome, whereas collectivistic culture made attributions based on collective outcome rather than the leaderÕs dispositions. Dispositional attributions When we examined the effects of prototypicality, outcome, and culture on the extent to which the participants believed the leader to be responsible for the performance outcome of the company, a 2
2
2 ANOVA revealed a three-way interaction, F ð1; 178Þ ¼ 3:40, p ¼ :06, g2 ¼ :02 (see Table 2). To fully understand this interaction, we performed separate 2 (prototypicality)
2 (outcome) ANOVAs for the two cultures. For the individualistic sample, this analysis revealed an interaction, F ð1; 81Þ ¼ 4:08, p < :05, g2 ¼ :055. The results of the tests of simple effects showed that the individualistic students were more likely to give credit to the leader for the high outcome when the leader was prototypical (M ¼ 5:41, SD ¼ :82) than when he was not (M ¼ 4:43, SD ¼ :85), tð41Þ ¼ 3:63, p < :01. Thus, the leaders who behaved prototypically with a high outcome were rated as more responsible than those acting antiprototypically. On the other hand, this difference

disappeared when the outcome was low (M ¼ 5:17, SD ¼ :86 and M ¼ 4:92, SD ¼ :72, respectively, tð40Þ ¼ 1:02, p > :05). For the collectivistic sample, neither the interaction between prototypicality and outcome, nor the main effect of outcome was significant. However, there was a main effect of prototypicality, F ð1; 178Þ ¼ 6:43, p < :05, g2 ¼ :035. Consistent with earlier findings, the participants made higher person attributions when the leader was prototypical (M ¼ 5:35, SD ¼ :81) than when he was not (M ¼ 5:00, SD ¼ :88). Mediating role of dispositional attributions To understand more fully how attributional categories affected attributions of charisma and provide a test of Hypothesis 4, we used the attributional measure that asked to what extent the person was responsible for performance as a covariate in separate 2 (prototypicality)
2 (outcome) ANCOVAs for the individualistic culture, and compared those results to an ANOVA without the attribution rating as a covariate. For the individualistic sample, dispositional attribution was a significant covariate, F ð1; 80Þ ¼ 12:44, p < :01, g2 ¼ :14, but after taking into account this variable, there were no main effects or interactions of experimental conditions on charisma ratings. However, without attribution to person as a covariate there was a significant main effect for typicality, F ð1; 81Þ ¼ 7:31, p < :01, g2 ¼ :08, and an outcome by typicality interaction, F ð1; 81Þ ¼ 3:69, p ¼ :05, g2 ¼ :04. These ANCOVA results guided the mediation analysis for the high outcome condition only within the individualistic culture group. To investigate this mediation, we followed the procedure of Baron and Kenny (1996). We looked at mediation for the high outcome condition only, because our results revealed significant differences between the prototypical and antiprototypical conditions only when the companyÕs sales were high. Mediation was assessed by following the four-step regression equations. First, the mediator was regressed on the independent variable; second, the dependent variable was regressed on the independent variable to test for direct effects; third, the dependent variable was regressed simultaneously on both the independent variable and on the mediator to examine the link between the mediator and the dependent measure; finally, the dependent variable was regressed on the independent variable controlling for the mediator (Baron & Kenny, 1996; Judd & Kenny, 1981). First, regression of the dispositional attribution on the prototypicality showed that individualistic participants made more dispositional attributions for the prototypical leader than the antiprototypical leader, F ð1; 41Þ ¼ 13:20, p < :01, R2 ¼ .24. Second, we regressed the dependent measure (attribution of cha-

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

risma) on the prototypicality, and found a significant relationship, F ð1; 41Þ ¼ 10:89, p < :01, R2 ¼ .21. Consistent with the previously reported ANCOVA results, the greater attribution of charisma to the leader was found in the prototypical than in the antiprototypical condition. At the third step of the mediation analysis, with prototypicality simultaneously entered in the equation, dispositional attributions influenced the charisma attributions, F ð2; 40Þ ¼ 8:74, p < :01, R2 ¼ .30. Supporting the mediation role of the dispositional attributions, after insertion of the dispositional attributions into the equation, the direct effect of prototypicality on charisma no longer remained significant, p > :05. Thus, after controlling for the dispositional attribution, prototypicality was not a significant predictor of charisma. Utilization of SobelÕs (1982) procedure for testing the significance of the indirect mediation relationship provided evidence of a reliable change in b from simple (b ¼ :46) to multiple regression (b ¼ :28), zð41Þ ¼ 30:67, p < :05. Thus, these analyses show that the dispositional attributions mediated the effect of prototypicality on the attributions of charisma. Moderating role of dispositional attributions As a test of Hypothesis 5, a moderated regression with charisma as the dependent measure and the dispositional attribution, outcome level, and the dispositional attribution by outcome level cross product revealed a significant interaction, (R2 ) changed for the interaction .037, F ð1; 181Þ ¼ 7:61, p < :001, and for the overall equation, R2 ¼ .346, F ð3; 181Þ ¼ 8:19, p < :001. An identical analysis using the GLI measures as a control variable in the regression described above, produced the same result. Exploring the interaction as suggested by Aiken and West (1991), showed that if a person tended to make dispositional attributions, then those who were in the high outcome condition rated that leader as more charismatic than those in the low outcome condition. The standardized slope (b) for this regression line was .46, p < :001. If a person made less dispositional attributions, then the outcome level did not affect his or her ratings or charisma, and they tended to rate charisma low across both outcomes. The standardized slope was .05 ns.

Discussion In this study, we primarily investigated how the combined effects of leader prototypicality and performance outcome determine the attribution of charisma. Across both individualistic and collectivistic cultures, we found that prototypical behavior combined with a high outcome produce greater attribution of charisma

61

to the leader as compared to their individual contributions. One possible explanation for this interaction is that when leadership traits and behaviors match a prototype, trait inferences made spontaneously by the perceiver help him and her assimilate other information, such as outcome of events. Support for this interpretation comes from Van Overwalle, Drenth, and Marsman (1999) who argued that ‘‘dispositional inferences often occur spontaneously without any intention or awareness on the perceiverÕs part to make them’’ (p. 450), and they occur in reaction to both the actor and to his or her actions. This explanation is also consistent with the results of the moderator analysis of dispositional attributions. We found that only when a person makes dispositional attributions, the outcome information affects his or her ratings of charisma. Consistently, Van Overwalle et al. (1999) showed that information disconfirming a dispositional attribution served to eliminate spontaneous trait inferences. Although the vignettes used in the present study provided an impoverished context for judging leadership, participants attributed charisma to the leader based merely on the degree of prototypicality of leader and company outcomes even in the absence of specific charismatic behavior. Moreover, we found that the manipulation of the different types of information affected charisma ratings above and beyond just a general impression of leadership. As mentioned previously, charismatic leadership research must take into account attributional effects to separate out real behavioral aspects. Although the perception of the leader plays an important role in making judgments about the leader, examining the conditions under which these judgments and impressions are made extends our understanding of leadership perceptions. Another important aspect of our study is that our manipulation of outcome was successful even with a limited range of outcomes (i.e., 5% vs. 25%). Most studies of inferential perceptions compare an increase vs. decrease in sales, rather than two levels of increase. If we increased the range of outcomes (for example 25% decrease vs. a 25% increase), we might have had seen even stronger effects. Our results showed that in the individualistic society leadership perceptions were predominantly determined by the recognition-based processing, whereas in the collectivistic culture, inference-based processing was more likely to affect charisma attributions. This study is important because it simultaneously looked at both types of information processing to understand their interactive effects on attributions in two separate cultures. In prior research the experimental paradigms have never orthogonally manipulated both perceptual processes, with one exception (Shamir, 1992). While previous studies have compared implicit leadership

62

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

theories across cultures and what constitutes an ideal leader with respect to leader prototypes and scripts (Ayman, 1993; Den Hartog et al., 1999; Gerstner & Day, 1994), only a few studies systematically examined information processing issues. Thus, the present study provides the first cross-cultural assessment of the effects of two types of information processing on charisma attributions. The importance of understanding crosscultural context in social cognitive processing has large implications for the study of cross-cultural leadership as well as understanding issues of leadership diversity within the US. Our results also provide support for the ‘‘Semantic– procedural Interface Model’’ by K€ uhnen et al. (2001). This model proposes a mechanism by which the nature of self-knowledge affects perceptions and attributions. It suggests that members of a collectivistic culture value being responsive to the needs and expectations of others to maintain interdependence with them (K€ uhnen et al., 2001). When interdependent self-knowledge is highly accessible, it increases the tendency to focus attention on the social situation (K€ uhnen et al., 2001). On the other hand, for individualistic members, independent self-knowledge is more accessible, therefore judgments are likely to be assimilated toward autonomous contents. Our study confirms this model in that individualistic participants processed the information unaffected by the context in which it appeared (‘‘context-independent mode of thinking’’), and exhibited a tendency to focus more on individual-based explanations, whereas collectivistic participants processed the information while attending to its relations to the social context (‘‘context-bound mode of thinking’’) (K€ uhnen et al., 2001, p. 406). Furthermore, this study is important because it applies social-cognitive perspective to executive leadership and organizational performance. As mentioned previously, attributions of charisma are influenced by leader or organizational performance (Shamir, 1992), and by specific behaviors that the leader exhibits (Conger & Kanungo, 1987). Charisma researchers acknowledge the importance of perceptual processes involved in making attributions of charisma, but have not fully investigated the exact processes by which attributions are affected. Our results show that the romance of leadership (Meindl et al., 1985; Pillai & Meindl, 1991) is alive and well because of the tendency for individuals in our study to rate leaders as more charismatic when the outcome for the company was good even though the leader was not exhibiting charismatic leader behavior in either culture. Limitations Studies that use paper and pencil descriptions of leaders cannot fully capture many features of a real life

leadership situation in the absence of face-to-face interaction (Lord & Maher, 1993), however, these studies can help unravel the social cognitive processing of leadership perceptions (Lord et al., 1984). For example, although Lord and Alliger (1985) found that leadership perceptions in rich information environments such as a group interaction context were less affected by leader prototypic behavior than they were in impoverished information environments, Maurer and Lord (1991) showed that both high and low information environments did affect leadership perceptions. Future studies of leaders portrayed through videotaped or face-to-face interactions as suggested by Maurer and Lord (1991) and Lord and Alliger (1985) can increase the realism of this type of approach and allow us to examine other cues that differentially affect the attributions of charisma across cultures. In addition, future research should examine how prototypicality affects the acceptance of leader behavior by group members (Lord et al., 1984) and how it affects the perception of subordinates who have a longstanding history with leaders within organizations (Phillips & Lord, 1982). In summary, although laboratory research on charismatic leadership is criticized for lack of realism, it is an important step in understanding the causes and effects of charismatic leadership (Brown & Lord, 1999; Wofford, 1999) and can help understand information processing effects for charismatic leadership (Hogg, 2001). One potential limitation in this study is related to the equivalence of the prototypicality manipulation. Based on a pilot study, we have employed slightly different manipulations in the two cultures. Although two of the four items were actually the same in the two cultures, there could be a potential confound due to the differences in the other items. However, we think that this is not a serious concern for our study due to three reasons. First of all, the strength of the prototypicality manipulation relies on the meaning of the stimulus items that is created by the interpretation of items by participants, not just by the surface features. In other words, the strength of the manipulation depends on the joint function of items and interpretation of participants. Secondly, the results of the pilot study indicated that both American and Turkish students rated the prototypical behaviors to be more prototypical than the antiprototypical behaviors. The absolute difference in prototypicality for the two stimuli is actually greater (although not significant) for the Turkish sample (1.05) than for the American sample (.59), thus it is difficult to argue that the manipulation of prototypicality for the American sample was stronger than for the Turkish sample. Finally, the manipulation checks of prototypicality also suggest that the prototypicality manipulation functioned equally in the two cultures.

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

Implications for future research A recent study by Awamleh and Gardner (1999) showed that a strong vision delivery produces effects larger than that of organizational outcomes. Their research along with that of Holladay and Coombs (1993, 1994) points to specific ways that a leader can increase perceptions of charisma, for example, through the content and delivery of speeches. Other important research looks specifically at the role of stress or crisis situations as a contextual factors in the attribution of charisma (Halverson, Murphy, & Riggio, 2003; House, Spangler, & Woycke, 1991; Hunt, Boal, & Dodge, 1999; Pillai, 1996; Shamir et al., 1993). Both areas of research, specific behaviors and context, are important areas for future research on charismatic leadership. Within the current study, we only chose to examine the effects of one specific type of cultural value, collectivism/individualism. While it is important to limit research questions in this way to understand specific cultural variation, future research should expand this effort by examining how other cultural dimensions (such as power distance, masculinity, and uncertainty avoidance; Hofstede, 1980) affect the social cognitive processing of leadership information. In addition, including other dependent variables are important to understanding these processes. For instance, Phillips and Lord (1982) examined how these types of information processing mechanisms affected recall of leader behavior. Understanding these deeper processes will inform future cross-cultural research on leadership. In this study we examined charismatic leadership in two cultures to understand the generalizability of perceptual process of leader behavior across cultures. We believe that culturally diverse analysis of attribution can offer much in coming to understand the social cognitive factors affecting leadership judgment, and in turn, cognitive processes involved in leadership perceptions of different cultures.

Acknowledgments We thank Prof. Ercan Alp, and Prof. Diane Sunar in the Department of Psychology, Bogazici University for their cooperation, Prof. Ron Riggio and Kevin Groves at the Kravis Leadership Institute, Claremont McKenna College for their helpful comments, and Elif Kara for her assistance in collecting data. We would like to thank a number of Kravis Leadership Institute research assistants for work on this project including Alexa Burns for data collection. We also thank two anonymous reviewers for their invaluable suggestions.

63

Appendix A. Prototypical and antiprototypical behaviors used for the American and Turkish versions of the vignette Prototypical

American

1. He (the leader) exercises great influence on his subordinates 2. He wants his own way on issues at hand 3. He asks his subordinates to work harder 4. He talks to his subordinates frequently

Antiprototypical 1. He tends not to change established and comfortable practices of his workgroup 2. He usually requests his subordinatesÕ approval on issues at hand 3. He admits it when he makes a mistake 4. He usually lets his subordinates decide what to do Prototypical

Turkish

1. He explains what the goal of the group should be 2. He assigns specific tasks to the group members 3. He asks his subordinates to work harder 4. He talks to his subordinates frequently

Antiprototypical 1. He lets everyone work at their own pace 2. He usually requests his subordinatesÕ approval on issues at hand 3. He admits it when he makes a mistake 4. He usually lets his subordinates decide what to do

Appendix B. Vignettes used in different culture for different prototype conditions Vignette used in the prototypical condition in US John Smith is the Director of Sales for a major appliance firm. John assumed his position two years ago following his attainment of an MBA degree with specialization in marketing. John is currently in charge of

64

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

12 subordinates. He exercises great influence on his subordinates, and asks them to work harder. John always wants his own way on issues at hand, and talks to his subordinates frequently in each stage involved in marketing a new product. Twice a year, John writes detailed reports concerning his suggestions to his superior. These reports generally take him a week to write, and they are read carefully by JohnÕs supervisor, since they are frequently used for making future product changes. At the end of the year sales records have shown a slight (5%) (or high 25%) increase during the year, over the last yearÕs performance. Vignette used in the antiprototypical condition in US John Smith is the Director of Sales for a major appliance firm. John assumed his position two years ago following his attainment of an MBA degree with specialization in marketing. John is currently in charge of 12 subordinates. He usually lets his subordinates decide what to do in each stage involved in marketing a new product, and tends not to change established and comfortable practices of his workgroup. His subordinates indicated that John usually requests their approval on issues at hand, and admits it when he makes a mistake. Twice a year, John writes detailed reports concerning his suggestions to his superior. These reports generally take him a week to write, and they are read carefully by JohnÕs supervisor, since they are frequently used for making future product changes. At the end of the year sales records have shown a slight (5%) (or high 25%) increase during the year, over the last yearÕs performance. Vignette used in the prototypical condition in Turkey John Smith is the Director of Sales for a major appliance firm. John assumed his position two years ago following his attainment of an MBA degree with specialization in marketing. John is currently in charge of 12 subordinates. He explains what the goal of the group should be to his subordinates, and asks them to work harder. John always assigns specific tasks to the group members, and talks to his subordinates frequently in each stage involved in marketing a new product. Twice a year, John writes detailed reports concerning his suggestions to his superior. These reports generally take him a week to write, and they are read carefully by JohnÕs supervisor, since they are frequently used for making future product changes. At the end of the year sales records have shown a slight (5%) (or high 25%) increase during the year, over the last yearÕs performance. Vignette used in the antiprototypical condition in Turkey John Smith is the Director of Sales for a major appliance firm. John assumed his position two years

ago following his attainment of an MBA degree with specialization in marketing. John is currently in charge of 12 subordinates. He usually lets his subordinates decide what to do in each stage involved in marketing a new product, and lets everyone work at their own pace. His subordinates indicated that John usually requests their approval on issues at hand, and admits it when he makes a mistake. Twice a year, John writes detailed reports concerning his suggestions to his superior. These reports generally take him a week to write, and they are read carefully by JohnÕs supervisor, since they are frequently used for making future product changes. At the end of the year sales records have shown a slight (5%) (or high 25%) increase during the year, over the last yearÕs performance.

References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications Inc. Anderson, S. D., & Wanberg, K. W. (1991). A convergent validity model of emergent leadership in groups. Small Group Research, 22, 380–397. Arbuckle, J. L., & Wothke, W. (1995). AMOS 4.0 User’s Guide. Chicago, IL: SPSS. Awamleh, R., & Gardner, W. L. (1999). Perceptions of leader charisma and effectiveness: The effects of vision content, delivery, and organizational performance. The Leadership Quarterly, 10, 345–373. Ayman, R. (1993). Leadership perceptions: The role of gender and culture. In M. Chemers, & R. Ayman (Eds.), Leadership theory and research: Perspectives and directions (pp. 137–166). San Diego, CA: Academic Press. Baron, R. M., & Kenny, D. A. (1996). The moderator/mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Barsalou, L. W. (1985). Ideals, central tendency, and frequency of instantiation as determinants of graded structure in categories. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 629–654. Bass, B. M. (1990). Bass & StodgillÕs handbook of leadership: Theory, research, and applications (3rd ed.). New York: The Free Press. Bass, B. M. (1997). Does the transactional–transformational leadership paradigm transcend organizational and national boundaries? American Psychologist, 52, 130–139. Bass, B. M., & Avolio, B. J. (1993). Transformational leadership: A response to critiques. In M. M. Chemers, & R. Ayman (Eds.), Leadership theory and research: Perspectives and directions (pp. 48– 80). Orlando, FL: Academic Press. Bass, B. M., & Avolio, B. J. (1994). Increasing organizational effectiveness through transformational leadership. Thousand Oaks, CA: Sage Publications Inc. Binning, J. F., Zaba, A. J., & Whattam, J. C. (1986). Explaining the biasing effects of performance cues in terms of cognitive categorization. Academy of Management Journal, 29, 521–525. Brown, D. J., & Lord, R. G. (1999). The utility of experimental research in the study of transformational/charismatic leadership. The Leadership Quarterly, 10, 531–540. Calder, B. J. (1977). Endogenous–exogenous versus internal–external attributions: Implications for the development of attribution theory. Personality and Social Psychology Bulletin, 3, 400–406.

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66 Carl, & Javidan (2001). Universality of charismatic leadership: A multi-nation study. Academy of Management Best Papers. Chemers, M. M., & Ayman, R. (1993). Leadership Theory and Research: Perspectives and directions. San Diego, CA: Academic Press. Conger, J. A., & Kanungo, N. (1987). Toward a behavioral theory of charismatic leadership in organizational settings. Academy of Management Review, 12, 637–647. Conger, J. A., & Kanungo, N. (1994). Charismatic leadership in organizations: Perceived behavioral attributes and their measurement. Journal of Organizational Behavior, 15, 439–452. Conger, J. A., & Kanungo, N. (1998). Charismatic leadership in organizations. Thousand Oaks, CA: Sage Publishing. Cronshaw, S. F., & Lord, R. G. (1987). Effects of categorization, attribution, and encoding processes on leadership perceptions. Journal of Applied Psychology, 72, 97–106. Den Hartog, D. N., House, R. J., Hanges, P. J., Ruiz-Quintanilla, S. A., Dorfman, P. W., et al. (1999). Culture specific and crossculturally generalizable implicit leadership theories: Are attributes of charismatic/transformational leadership universally endorsed? The Leadership Quarterly, 10, 219–256. Etzioni, A. (1961). A comparative analysis of complex organizations. New York: Free Press. Fiske, S. T., & Taylor, S. E. (1984). Social cognition. Reading, MA: Addison-Wesley. Foti, R. J., Fraser, S. L., & Lord, R. G. (1982). Effects of leadership labels and prototypes on perceptions of political leaders. Journal of Applied Psychology, 67, 326–333. Fraser, S. L., & Lord, R. G. (1988). Stimulus prototypically and general leadership impressions: Their role in leadership and behavioral ratings. Journal of Psychology, 122, 291–303. Gerstner, C. R., & Day, D. V. (1994). Cross-cultural comparison of leadership prototypes. Leadership Quarterly, 5, 121–134. Halverson, S. K., Murphy, S. E. & Riggio, R. E. (2003). Charismatic leadership in crisis situations: The role of social and emotional intelligence. Unpublished manuscript. Claremont, CA: Kravis Leadership Institute, Claremont McKenna College. Hanges, P. J., Lord, R. G., & Dickson, M. W. (2000). An informationprocessing perspective on leadership and culture: A case for connectionist architecture. Applied Psychology: An International Review, 49(1), 133–161. Hewstone, M. (1989). Causal attribution: From cognitive processes to collective beliefs. Oxford: Basil Blackwell. Hofstede, G. (1980). CultureÕs consequences. Newbury Park, CA: Sage Publishing Inc. Hofstede, G. (1984). National cultures revisited. Behavior Science Research, 18, 285–305. Hogg, M. A. (2001). A social identity theory of leadership. Personality and Social Psychology Review, 5, 184–200. Holladay, S. J., & Coombs, W. T. (1993). Communicating visions: An exploration of the role of delivery in the creation of leader charisma. Management Communication Quarterly, 6, 405–427. Holladay, S. J., & Coombs, W. T. (1994). Speaking of visions and visions being spoken: An exploration of the effects of content and delivery on perceptions of leader charisma. Management Communication Quarterly, 8, 165–189. Hollander, E. P., & Julian, J. W. (1969). Contemporary trends in the analysis of leadership processes. Psychological Bulletin, 71, 387–397. House, R. J. (1977). A 1976 theory of charismatic leadership. In J. G. Hunt, & L. L. Larson (Eds.), Leadership: The cutting edge. Carbondale, IL: Southern Illinois University Press. House, R. J., Hanges, P. J., Ruiz-Quintanilla, S. A., Dorfman, P. W., Jaivdan, M., Dickson, M., Gupta, V., and 170 co-authors (1999). Cultural influences on leadership and organizations: Project GLOBE. In W. Mobley (Ed.), Advances in Global Leadership (Vol. l, pp. 171–233). San Francisco: JAI Press.

65

House, R. J., Spangler, W. D., & Woycke, J. (1991). Personality and charisma in the US presidency: A psychological theory of leadership effectiveness. Administrative Science Quarterly, 35, 364–396. Hughes, R. L., Ginnett, R. C., & Curphy, G. J. (2003). Leadership: Enhancing the lessons of experience. New York: Irwin McGrawHill. Hunt, J. G., Boal, K. B., & Dodge, G. E. (1999). The effects of visionary and crisis-responsive charisma on followers: An experimental examination of two kinds of charismatic leadership. The Leadership Quarterly, 10(3), 423–448. Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation in treatment evaluations. Evaluation Review, 5, 602–619. Kelley, H. H. (1973). The processes of causal attribution. American Psychologist, 28, 107–127. Kenney, R. A., Blascovich, J., & Shaver, P. R. (1994). Implicit leadership theories: Prototypes for new leaders. Basic and Applied Sociology Psychology, 15, 409–437. K€ uhnen, U., Hannover, B., & Schubert, B. (2001). The semantic– procedural interface model of the self: The role of self-knowledge for context-dependent versus context-independent modes of thinking. Journal of Personality and Social Psychology, 80, 397–409. Lord, R. G. (1985). An information processing approach to social perceptions, leadership, and behavioral measurement in organizations. Research in Organizational Behavior, 7, 87–128. Lord, R. G., & Alliger, G. (1985). A comparison of four information processing models of leadership and social perceptions. Human Relations, 38, 47–65. Lord, R. G., Brown, D. J., Harvey, J. L., & Hall, R. J. (2001). Contextual constraints on prototype generation and their multilevel consequences for leadership perceptions. The Leadership Quarterly, 12, 311–338. Lord, R. G., De Vader, C., & Alliger, G. (1986). A meta-analysis of the relation between personality traits and leadership perceptions: An application of validity generalization procedures. Journal of Applied Psychology, 71, 402–410. Lord, R. G., & Emrich, C. G. (2001). Thinking outside the box by looking inside the box: Extending the cognitive revolution in leadership research. Leadership Quarterly, 11, 551–579. Lord, R. G., Foti, R. G., & De Vader, C. (1984). A test of leadership categorization theory: Internal structure, information processing, and leadership perceptions. Organizational Behavior and Human Performance, 34, 343–378. Lord, R. G., & Maher, K. (1993). Leadership and information processing. New York: Routledge. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253. Maurer, T. J., & Lord, R. G. (1991). An exploration of cognitive demands in group interaction as a moderator of information processing variables in perceptions of leadership. Journal of Applied Social Psychology, 21, 821–839. Meindl, J. R., & Ehrlich, S. B. (1987). The romance of leadership and the evaluation of organizational performance. Academy of Management Journal, 30, 91–109. Meindl, J. R., Ehrlich, S. B., & Dukerich, J. M. (1985). The romance of leadership. Administrative Science Quarterly, 30, 78–102. Mischel, W. (1979). On the interface of cognition and personality: Beyond the person/situation debate. American Psychologist, 34, 740–754. Misumi, J. (1985). The behavior science of leadership: An interdisciplinary Japanese research program. Ann Arbor: University of Michigan Press. Moghaddam, F. M. (1998). Social psychology: Exploring universals across cultures. New York: Freeman. Murphy, M. R., & Jones, A. P. (1993). The influences of performance cues and observational focus on performance rating

66

N. Ensari, S.E. Murphy / Organizational Behavior and Human Decision Processes 92 (2003) 52–66

accuracy. Journal of Applied Social Psychology, 23(18), 1523– 1545. Pfeffer, J. (1977). The ambiguity of leadership. Academy of Management Review, 2, 104–112. Phillips, J. S., & Lord, R. G. (1981). Causal attributions and perceptions of leadership. Organizational Behavior and Human Performance, 28, 143–163. Phillips, J. S., & Lord, R. G. (1982). Schematic information processing and perceptions of leadership in problem-solving groups. Journal of Applied Psychology, 67, 486–492. Pillai, R. (1996). Crisis and the emergence of charismatic leadership in groups: An experimental investigation. Journal of Applied Social Psychology, 26, 543–562. Pillai, R., & Meindl, J. R. (1991). The effect of a crisis on the emergence of charismatic leadership: A laboratory study. Academy of Management Best Paper Proceedings, 235–239. Rush, M. C., Phillips, J. S., & Lord, R. G. (1981). Effects of a temporal delay in rating on leader behavior descriptions: A laboratory investigation. Journal of Applied Psychology, 66, 442–450. Sashkin, M. (1988). The visionary leader. In J. Conger, & R. Kanungo (Eds.), Charismatic leadership: The elusive factor in organizational effectiveness (pp. 122–160). San Francisco: Jossey-Bass. Schein, E. N. (1985). Organizational culture and leadership. San Francisco: Jossey-Bass. Shamir, B. (1992). Attribution of influence and charisma to the leader: The Romance of Leadership revisited. Journal of Applied Social Psychology, 22, 386–407. Shamir, B., House, R., & Arthur, M. B. (1993). The motivational effects of charismatic leadership: A self-concept based theory. Organizational Science, 4, 577–594.

Shaw, J. B. (1990). A cognitive categorization model for the study of intercultural management. Academy of Management Review, 15, 626–645. Smith, P. B. (1997). Cross cultural leadership: A path to the goal? In P. C. Barley, & M. Erez (Eds.), New perspectives on international industrial/organizational psychology (pp. 239–626). San Francisco: The New Lexington Press. Smith, P. B., & Peterson, M. F. (1988). Leadership, organizations and culture: An event management model. Thousand Oaks, CA: Sage Publications Inc. Sobel, M. (1982). Asympototic confidence intervals for indirect effects in structural equations models. In S. Leinhart (Ed.), Sociological methodology (pp. 290–312). San Francisco: Jossey-Bass. Triandis, H. C., Bontempo, R., Villareal, M. J., 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 Overwalle, E., Drenth, R., & Marsman, G. (1999). Spontaneous trait interferences: Are they linked to the actor or to the action? Personality and Social Psychology Bulletin, 25(4), 450– 462. Weiner, B. (1974). Achievement motivation and attribution theory. Morristown, NJ: General Learning Press. Wofford, J. C. (1999). Laboratory research on charismatic leadership: Fruitful or futile? Leadership Quarterly, 10, 523–530. Yorges, S. L., Weiss, H. M., & Strickland, O. J. (1999). The effect of leader outcomes on influence, attributions, and perceptions of charisma. Journal of Applied Psychology, 84, 428–436. Received 31 August 2001