From cognition to behavior: a cross cultural study for global business effectiveness.

Strubler, David ; Agarwal, Atul ; Park, Sung-Hee 等

INTRODUCTION

With globalization expanding at an exponential rate, the need for cross-culturally effective managers and workers has increased significantly. Professionals with strong and favorable intercultural attitudes, skills, knowledge, and behavior add value to their organizations. Cross-cultural effectiveness builds organizational learning, making it possible to expand and integrate marketing, sales, supply chains and operations to other countries. Whether professionals are working overseas as expatriates or conducting short-business trips, parent and host organizations need methods to ensure the success of multinational business ventures.

Black, Mendenhall, & Oddou (1991) first supported the idea that international adjustment was the result of an interaction effect of multiple variables in sequence. They found that training and previous experience in the pre-departure stage are prescriptions for improving the accuracy of an assigned expatriate's perception of themselves and the situation prior to departure. Some other studies found pre-departure selection and preparation to be important factors for adjustment while identifying variables that predict intercultural effectiveness for either selection or training purposes (Hutchings, 2002; Liu & Lee, 2008). It should be noted that many competency trait variables, such as flexibility, are cross-culture general traits so that they apply regardless of which culture the expatriate is entering (Adler, 1974; Bochner, 1973).

However, in the late 1980s, much of the trait research started to shift toward integrating cognitive and behavioral research primarily due to limited support for the links between traits and performance (Hammer, 1987). Other research showed that a combination of these trait and behavioral factors, in addition to culture specific preparation, may play an important role in expatriate success (Pires, Stanton, & Ostenfeld, 2006; Elmer, 1986). For example, in a recent study, Smith and Reynolds (2009) set out to "assess differences between cognitive and affective measures and their ability to predict behavioral intentions and the impact of service features on these measures" among a population of subjects from a variety of cultures. They found that for all cultural groups, overall quality, satisfaction, and positive affect predict behavioral intentions."

Building on and modifying the model of Black et al. (1991), Strubler & Park (2009) have argued that assigned expatriates should engage in (and their parent corporations ought to provide) cross-cultural effectiveness testing for selection purposes and to ensure the accuracy of pre-departure expectations as shown in Table 1. While testing should not be the only criteria for expatriate selection, testing does make both the individual and organization aware of strengths and weaknesses prior to selection and departure. Testing also makes it possible to design appropriate training for the expatriate.

Black et al. (1991) identified key factors affecting intercultural effectiveness and adjustment. They define 1) intercultural effectiveness as the ability of a person within the intercultural environment, and 2) adjustment as the overall multifaceted process through which expatriates develop an increasing degree of satisfaction in being able to cope with a cross-cultural environment. Therefore, it should be noted that intercultural effectiveness and adjustment are not equivalent concepts. Intercultural effectiveness may predispose an individual toward adjustment and without it the individual may not be able to adjust. However, adjustment is a complex process involving a large number of variables, many of which are outside the control of the individual, e.g., family and organization support. Therefore, intercultural effectiveness is necessary but insufficient for adjustment.

Still, according to Black et al. (1991), intercultural effectiveness and adjustment possess four common dimensions, namely, self-orientation, other-orientation, perceptual skills and cultural toughness. Self-oriented individuals engage in activities and have attributes that increase their self-esteem and confidence while finding replacements for their home interests and activities. They, in turn, handle stress well and demonstrate efficacy in both the work and social environments. Other-oriented individuals have the ability to develop relationships with host nationals and actively seek and find mentors. Individuals with strong perceptual skills tend to engage in non-judgmental, non-evaluative mental processing about their situation. Successful expatriates are often required to adjust to cultural toughness, which refers to differences in standards of living that expatriates experience: the greater the difference, the more difficult the adjustment. Therefore, these individual factors seem to contribute to adjustment.

[TABLE 1 OMITTED]

Fisher and Hartel (2003) further assert that three personal factors play a significant role in intercultural effectiveness: 1) ability to communicate effectively, 2) to establish relationships, and 3) to cope with psychological stress. Still another recent study proposed that emotions, especially for individualists working in collectivistic cultures for long periods of time, play a major role in cross-cultural success. More specifically, emotional demands caused by cultural differences in expatriate encounters impact negatively on their experience (Tan, Hartel, Panipucci & Strybosch, 2005). Therefore, emotional maturity may be a major factor, at least when cultural differences are great, in determining competency and possibly whether an expatriate will complete a long-term assignment. Emotional maturity or intelligence is defined as "an array of capabilities, competencies, and skills that influence one's ability to cope with environmental demands" (Tan et al., 2005). Four key factors are cited: emotional appraisal and expression, emotional regulation in self and others, promotion of intellectual and emotional growth, and generation of emotions to assist in problem solving (Salovey & Mayer, 1990; Tan et al., 2005). For example, expatriates who can deal with negative emotions in a positive manner can experience continued job satisfaction in a cross-cultural environment. Further, expatriates with idiocentric personalities, i.e., individualists who view the environment as unstable and themselves as stable, will experience a higher degree of emotional labor. Also, individualists (regardless of gender) prefer to be frank about their emotions and will fare better in feminine cultures where there is more freedom to express even their feelings of frustration (Hofstede, 1980; Mumby & Putnam, 1992; Ollilainen, 2000). Finally, high status expatriates serving in collectivistic cultures will experience less emotional labor than low status expatriates. A deeper look at the role of emotion in cross-cultural competency implies that multiple intervening variables, including cultural dimensions and individual personality factors, create a more complicated interaction effect.

The Intercultural Competency Scale, used in this study and developed by Elmer (1986), identified twelve intercultural competency traits from an extensive literature review. These include the following twelve dimensions: Approachability, Intercultural Receptivity, Positive Realistic Orientation, Forthrightness, Social Openness, Enterprise, Shows Respect, Perseverance, Flexibility, Cultural Perspectivism, Venturesome, and Social Confidence. From these dimensions, the Intercultural Competency Scale (ICS) was developed. It tests an individual for the presence of these cross-culture-general traits. None of traits are specific to a particular cultural situation. A cognitively- measured trait approach, the ICS was employed in this study as a measure of intercultural effectiveness.

A similar concept, also employed in this current study, is Cross Cultural Social Intelligence (CCSI). Combining the social intelligence and cross cultural communication literature, Ascalon, Schleicher, and Born (2008) developed a comprehensive situational judgment test. While encompassing emotional intelligence, social intelligence is defined as the "ability to understand the feelings, thoughts, and behavior of persons, including oneself, in interpersonal situations and to act appropriately upon that understanding" (Marlow, 1986). Extending that concept, socially intelligent people can adapt their behavior in a wide array of social situations (Cantor & Kihlstrom, 1987). Because social intelligence is specific to a particular culture, it may not be able to explain interpersonal effectiveness across cultures. Therefore, empathy and ethnocentrism are assumed as the basis for judgment of social intelligence in cross-cultural interactions. Three abilities are measured in the CCSI test: 1) recognition and understanding of (non)verbal cues of people from multiple cultures; 2) ability to accurately infer social references in multiple cultural encounters; and 3) by accepting and understanding multiple cultures, achieve relevant social objectives across cultural negotiations. CCSI is an example of a systematic, interdisciplinary measurement for integrating and examining cognitive and behavioral dimensions of cross-cultural effectiveness.

However, literature is scant on studies that have explored the relationship between cognition/perception and the ability to accurately read cultural situations or take action successfully in cross cultural interactions. Therefore, this research focuses on identifying correlations between two testing instruments: ICS and CCSI. Based on this exploratory study, we measure respondents' intercultural competency traits through the ICS (a cognitive and perceptual measure) and also determine the extent to which possession of these traits can predict the presence of cross-cultural social intelligence (CCSI, a behavioral assessment). This research is intended to support future research on the relationship between intercultural effectiveness, cross-cultural background and success in adjusting to intercultural living situations.

METHODOLOGY

This research focuses on correlating two instruments that address intercultural effectiveness: the Intercultural Competency Scale (Elmer, 1986) and the Cross Cultural Social Intelligence test (Ascalon, Schleicher, & Born, 2008). The Intercultural Competency Scale was designed to evaluate the presence of cross-cultural abilities to interact with people from other cultures by testing individuals on twelve variables previously mentioned. The second study, developed by Ascalon, Schleicher, and Born (2008), measures Cross-Cultural Social Intelligence (CCSI Assessment) through a set of scenarios for which participants are asked to select the most appropriate interaction between two people in various cross-cultural situations.

Email requests were sent to approximately 100 working professionals to undertake both the ICS and CCSI surveys, both of which were made available on a website. To ensure data integrity for the correlation study, it was important to track and match completion of both surveys by individual respondents. After ignoring respondents who completed only one of the ICS or CCSI surveys, we found that 52 respondents undertook both of the surveys. Since seven out of these 52 respondents had missing data, the pool of respondents who filled out both the surveys completely was reduced to 45.

For the ICS test, each of the twelve factors, determined by a total of 45 items that are scored using a five-point Likert scale, is rated high, medium, and low. With 225 possible points, the mean score is 158 points, as calculated in the original research by Elmer (1986). These high, medium and low divisions are set from a range of scores. For example, a score of 200 is a high score and 120 is a low score. One's level is determined from the mean of the individual's test results. A total score is also assigned that explains one's overall capacity to build strong abiding relationships with host country people in another culture and to perform one's task in keeping with their organizational goals. A feedback page is provided to each participant that colors each of the twelve categories for easy recognition. It also provides the mean score obtained for each category. Then participants, by scrolling over that number, can have a more detailed explanation of what the score means and what a person can do to improve upon it (Elmer, 1986). Therefore, ICS measures personal characteristics that contribute to intercultural effectiveness. It is a cognitive scale which relies on self-reported preferences for engaging in general interpersonal situations, all of which have been determined to be indicators of intercultural effectiveness (Elmer, 1986).

By contrast, CCSI measures one's ability to correctly interpret and then select the best interaction scenario in situations with people from different cultures. For each question on the CCSI, four answers are given that must be rated by the participant on effectiveness. The respondent also indicates whether he/she would use the response. The responses offered are: Empathetic-Ethnocentric=EE,Empathetic-Non-ethnocentric=EN,Non-empathetic- Ethnocentric = NE, and Non-empathetic-Non-ethnocentric=NN. For each of the questions asked, each answer is assigned a CCSI style, is given a rating of 0-5 on effectiveness, and is rated as a best or worst answer. Each of these ratings is developed from the evaluation of cultural experts. Unfortunately, many of the offered answers are considered inconclusive on the CCSI style and best/worst categories, making analysis difficult. Feedback for this survey is not immediately given back to participants, although the data on CCSI style, best/worst scoring, and effectiveness for each response can be provided to the participant after completion if requested.

As suggested by Ascalon et al. (2009), we scored the CCSI instrument based on each respondent's likelihood of performing the best and worst alternatives. The scores can range from 1 (not at all likely) to 5 (extremely likely) for each item and scale. First, respondents' likelihood ratings for each of the worst alternatives for each scenario were reverse scored. Then item scores were computed by averaging respondents' likelihood ratings on the best and worst alternatives. The total scale score was calculated by averaging across all the items. Higher the CCSI score on the total scale, greater would be the cross cultural social intelligence. Higher CCSI scores would be related to empathy, openness to experience, conscientiousness, and emotional stability.

RESULTS

As indicated in Figure 1 below, the majority of respondents (90.383) were U.S. Citizens. The remaining respondents were from Australia, Europe, Asia, Central America, and the Middle East. All of the respondents were fluent in English, possessed or were completing a bachelor's or master's degree, and were currently working or had worked in either full or part-time professional positions. Virtually none were traditional students without professional work experience.

[FIGURE 1 OMITTED]

From all participants, factors of each test will be based on background information. The analysis of the data was performed using Minitab software. The current focus is to determine whether those with a high level of overall intercultural competency are able to produce a high overall score in actual implementation of that knowledge and their traits.

[FIGURE 2 OMITTED]

Figures 2 and 3 show the distribution of CCSI and ICS scores for the respondents. According to Figure 2, only 6% respondents scored low (< 3) whereas about 60% respondents scored high (> 4) on CCSI scale, thus implying possession of higher levels of cross cultural social intelligence by majority of the respondents in this sample.

[FIGURE 3 OMITTED]

Figure 3 shows that ICS scores for the same respondents ranged from 120 to 200. The majority (60%) of respondents scored between 150 and 180 on ICS scale. Furthermore, only 6% respondents scored very high (above 180) while 20% respondents scored low between 120 and 140 on ICS scale. It is interesting to note that none of the respondents scored very low (< 120) on the ICS scale.

To determine if respondents who scored higher on CCSI scale also scored higher on ICS scale, we next performed the correlation analysis between CCSI and ICS scores of all respondents using Minitab statistical software. The Pearson correlation (r) was found to be statistically significant at 0.05 level with a value of 0.36. It implies a positive relationship of moderate strength between ICS and CCSI scores. Since the ICS score is the cumulative score on 12 intercultural competency factors, we decided to examine the relationship of each specific factor with the CCSI score.

Table 2 shows Pearson correlation coefficients between all twelve intercultural competency factors and the CCSI scores. The results showed that four ICS factors Approachable, Perseverance, Cultural Perspectivism, and Venturesome - were significantly correlated with CCSI scores. Cultural Perspectivism and Social Confidence were found to exhibit strongest (r=0.4) and weakest (0.1) correlation with CCSI, respectively. Enterprise was the only factor to be negatively correlated with CCSI but without any statistical significance. It is interesting to note the that intuitively obvious factors such as Social Openness, Shows Respect, and Flexibility showed no significant relationship with CCSI.

To test the stream of research which propounds the effect of cognitive intercultural traits on one's cross cultural social performance, we decided to identify which ICS factors, if any, can be useful and reliable predictors of CCSI scores. A regression model was developed next using CCSI as the response variable and all twelve ICS factors as the predictor variables with the objective to keep the model both parsimonious and statistically useful.

Table 3 shows the stepwise regression model approach used to identify the most useful predictor variables. The 4-ICS factors found to be the most useful in predicting the CCSI score are Cultural Perspectivism, Approachable, Venturesome, and Enterprise. Table 4 shows the best regression model based on highest [R.sup.2] (35.16%) and smallest Mallows [C.sub.p] value (0.8) which can be stated as: E(CCSI) = 2.302 + 0.116 (Cultural Perspectivism) + 0.05 (Approachable) + 0.069 (Venturesome) - 0.058 (Enterprise)

According to Table 4, this regression model is globally useful with an F value = 5.42 and a corresponding p-value =0.001. The partial tests of usefulness for individual ICS factors are all significant at 0.102 level.

Figures 4 and 5 show the residual analysis and normality plot for the above regression model. It is clear that there are no major violations of the underlying assumptions of normality, independence, and equal variance. Hence, the above regression model can be used for predicting CCSI scores based on 4 cognitive intercultural predictor variables.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

DISCUSSION

Beginning with the overall scores for both tests, why is the distribution of ICS (150-180, mean = 158) and CCSI, (> 4), higher for the majority of respondents in this sample? One possibility is that the education and experience level of the subjects for this study was greater than scores would be for a control group randomly selected from the general population. All subjects in the study had professional industry experience and most were already completing graduate studies while working full-time. In short, education and professional-level work experience may be correlated with intercultural competence and cross-cultural social intelligence. This explanation will be tested in the next study as the characteristics of the population are correlated with their CCSI and ICS scores. We should note here that the CCSI is a new instrument that has not been tested extensively. It may require more testing with larger samples to develop norms for low, medium and high scores. This may explain why the CCSI scores did not follow a normal distribution.

As indicated in Table 2, why are seemingly and intuitively obvious cognitive factors such as Intercultural Receptivity, Positive Orientation, Social Openness, Shows Respect, Flexibility, and Social Confidence not significantly correlated with CCSI? One possibility is that some cognitions do not necessarily translate into behavior. Secondly, effects may already be included in the four cognitive factors that have significant correlation with CCSI. Third, ICS may be measuring factors that simply were not taken into account by CCSI. Again, CCSI measures are based on behaviors related to ethnocentrism and empathy. ICS is measuring traits. For example, Intercultural Receptivity is defined as being "interested in people, especially people from other cultures" (Elmer, 1986). This concept appears, on the surface, to be the opposite of ethnocentrism. However, people who do not practice ethnocentrism (cognitively or behaviorally) still may not have an active interest in people from other cultures. In short, absence of malice toward others who are different than oneself does not necessarily constitute interest in others. Likewise, neither the absence of ethnocentrism nor the presence of empathy would necessarily correlate with social openness (the inclination to interact with people regardless of their differences), flexibility (open to culture learning), or social confidence (tends to be self-assured). People who are empathetic are understanding, aware of, sensitive to, and vicariously experience the feelings thoughts and experiences of others (Websters, 2003). Therefore, empathy may readily and reasonably correlate with Cultural Perspectivism, the capacity to imaginatively enter into another cultural viewpoint (Elmer, 1986). We may conclude then that certain factors as defined by both Elmer (1986) and Ascalon et al (2008) are correlated and useful for research purposes. Other factors such as Social Confidence are operationally-defined in such a way that there is no correlation. Therefore, they are not useful in predicting behavior from cognition. For future research, we can conclude from both the correlation and regression analysis of ICS and CCSI factors, that Approachable, Cultural Perspectivism, and Venturesome are useful ICS factors in predicting behaviors measured by the CCSI.

FUTURE RESEARCH

This empirical study is the first in a series to test Strubler & Park's (2009) revision of Black et al's (1991) Framework for International Adjustment. The purpose of the revision was two-fold: 1) to make the model more prescriptive so that organizations could improve their expatriate success rate and 2) to test hypotheses that would close multiple gaps in the literature. The model combines the concepts of intercultural effectiveness and adjustment. In this present study, the goal was to test and correlate two instruments that measure intercultural effectiveness. Having reliable instruments that test both the cognitive and behavioral aspects of effectiveness are necessary to test hypotheses regarding both effectiveness and adjustment. The next empirical study will determine the extent to which cross-cultural experience, i.e., language and culture education, training, family background, travel, and work experience, is correlated with intercultural effectiveness. Later studies will focus on the correlations between extended cross-cultural experience and success in international adjustment.

REFERENCES

Ascalon, Ma. E., D.J. Schleicher, & M.Ph. Born (2008). Cross-cultural social intelligence: An assessment for employees working in cross-cultural contexts. Cross Cultural Management, 15(2), 109-130.

Black, J.S. (1988). Work role transitions: A study of American expatriate managers in Japan. Journal of International Business Studies, 19(2), 277-294.

Black, J.S. & M. Mendenhall (1990a). Cross-cultural training effectiveness: A review and theoretical framework for future research. Academy of Management Review, 15(1), 113-136.

Black, J.S. & M. Mendenhall (1991b). The u-curve adjustment hypothesis revisited: A review and theoretical framework. Journal of International Business Studies, 22(2), 225-247.

Black, J.S., M. Mendenhall & G. Oddou (1991). Toward a Comprehensive Model of International Adjustment: An Integration of Multiple Theoretical Perspectives. Academy of Management Review, 16(2), 292-317.

Elmer, M. (1986). Intercultural Effectiveness: Development of an Intercultural Competency Scale. Ph.D. dissertation, Michigan State University, East Lansing, Michigan.

Fisher, G. B. & C.E.J. Hartel (2003). Cross-cultural effectiveness of western Expatriate-Thai client interactions: Lessons learned for IHRM research and theory. Cross Cultural Management, 10(4), 4-28.

Hofstede, G. (1980). Cultures Consequences: International Differences in Work-related Values. Sage Publications: Beverly Hills, CA.

Marlow, H.A. (1986). Social intelligence: Evidence for multidimensionality and Construct independence. Journal of Educational Psychology, 78(1), 52-58.

Mumby, D.K. & L.L. Putnam (1992). The politics of emotion: A feminist reading of bounded rationality. Academy of Management Review, 17(3), 465-486.

Ollilainen, M. (2000). Gendering emotions: gendering terms: Construction of emotions in self-managing teamwork. In N.M. Ashkanasy, Hartel, C.E.J., & Zerbe, W.J. (Eds.), Emotions in the workplace: Research, theory, and practice: 82-96, Quorum Books: London.

Salovey, P. & J. Mayer (1990). Emotional intelligence. Imagination, Cognition And Personality, 9 (185-211). Strubler, D.C. & S. Park (2009). Framework for International Adjustment: Refining and Extending a Model.

Proceedings of the North American Management Society, Midwest Business Administration Association Annual Conference.

Tan, J.A.C., C.E.J. Hartel, D. Panipucci, & V.E. Strybosch (2005). The effect of emotions in cross-cultural expatriate differences. Cross Cultural Management, 12(2), 4-15.

David Strubler, Kettering University

Atul Agarwal, University of Illinois at Springfield

Sung-Hee Park, Kettering University

Muriel Elmer, Independent Consultant

Table 2: Correlations between CCSI and ICS Factors

ICS Factors                    Correlation (r)

1. Approachable                      0.30 *
2. Intercultural Receptivity         0.24
3. Positive Orientation              0.15
4. Forthrightness                    0.17
5. Social Openness                   0.24
6. Enterprise                       -0.05
7. Shows Respect                     0.27
8. Perseverance                      0.34 *
9. Flexibility                       0.28
10. Cultural Perspectivism           0.40 **
11. Venturesome                      0.35 *
12. Social Confidence                0.01

* Significant at "=0.05

** Significant at "=0.01

Table 3: Stepwise Regression: CCSI Scores versus 12-ICS Factors

Alpha-to-Enter: 0.15     Alpha-to-Remove: 0.15

Step                       1         2        3        4

Constant                  2.947    2.042     1.885    2.302

Cultural Perspectivism    0.109    0.113     0.106    0.116
T-Value                   2.90     3.16      2.99     3.29
P-value                   0.006    0.003     0.005    0.002

Approachable                       0.058     0.043    0.050
T-Value                            2.43      1.66     1.97
P-value                            0.020     0.105    0.056

Venturesome                                  0.064    0.069
T-Value                                      1.53     1.69
P-value                                      0.133    0.099

Enterprise                                           -0.058
T-Value                                              -1.68
P-value                                               0.101

S                         0.491    0.0466    0.458    0.449
R-sq                     16.32    26.61     30.60    35.16
R-sq (adj.)              14.37    23.12     25.52    28.67
Mallows Cp                5.2      1.5       1.3      0.8

Table 4: Regression Analysis--CCSI Scores versus significant ICS
Factors

The regression equation is:

CCSI = 2.30 + 0.116 Cultural Perspectivism + 0.0504 Approachable
                          (0.002) *                  (0.056) *

+ 0.0690 Venturesome - 0.0581 Enterprise
          (0.099) *            (0.101) *

* Numbers in parentheses represent the P-value

Analysis of Variance

Source              DF          SS          MS           F         P
Regression           4        4.3629      1.0907       5.42      0.001
Residual Error      40        8.0467      0.2012
Total               44        12.4096

S = 0.448516 R-Sq = 35.2% R-Sq(adj) = 28.7%