How Much Woe When We Go? Predicting Culture Shock.

How much woe when we go? International Journal of Stress Management, Spring 1996

1

How Much Woe When We Go: A Quantitative Method for Predicting Culture Shock Lawrence R. Zeitlin1

Running head: How much woe when we go: A quantitative method for predicting culture shock. Mailing address: Lawrence Zeitlin; 12 Brook Lane; Peekskill, NY 10566-6502 Telephone: (914) 737-4905 FAX: (914) 737-4905 e-mail:[email protected] 1

Graduate Center - City University of New York; New York, NY 10010

ABSTRACT Culture shock is defined as the confusion and discomfort caused by the conflict in perceived motives and expected behaviors between the home culture and the foreign culture. Several quantitative and graphical methods employing techniques of cluster analysis and similarity mapping are offered for predicting the magnitude of culture shock between pairs of countries using data extracted from Hofstede's 1980-83 studies of national cultural values. Implications for business, politics, and personal stress management are discussed.

Keywords: Culture shock, national values, cross cultural research, cluster analysis, similarity mapping.

How much woe when we go? International Journal of Stress Management, Spring 1996

2

INTRODUCTION Anyone with first hand experience in foreign travel understands the woe of culture shock, the realization that expected behaviors and perceived values of the new environment are disturbingly dissimilar from those of home. Most of us expect that there will be physical and climatic differences in the environment when we travel. The language and the coinage may be unfamiliar. The food may be unusual. Foreigners may even drive on the other side of the road. But, given a little time, we can adjust to all these differences and even look forward to meeting the challenge as one of the joys of travel. It is much harder to deal with the subtle differences in motives and values encountered during residence in a foreign culture. We spend a lifetime learning what to expect from others and what others expect from us. We strive to meet the demands of our society and our culture. Other cultures make different demands and other people have different expectations. Culture shock may be defined as the confusion and discomfort caused by the conflict in perceived motives and expected behaviors between the home culture and the foreign culture. Individuals within all cultures have unique and idiosyncratic personalities which are often a source of misunderstanding. Cultural values, however, are common to almost everyone in a given culture and establish a general theme of behavior. Personality differences are simply variations on that theme. Is there a way to predict the amount of culture shock that someone would experience upon travel to another country? Let us ignore the physical differences between societies and restrict the culture shock concept to cultural ideas and values. Such a prediction depends upon assessment and quantification of value systems of different countries. The amount of culture shock that a traveler will experience will be proportionate to the quantitative difference in value systems between the home country and the visited country. In a summary article, Adler (1984) classified a variety of approaches taken by researchers of cultural values. Parochial research is that conducted in a single culture, generalizing the findings to all other cultures. It assumes universal motives and behaviors. It is the model used by businessmen and politicians in presuming that the values of one culture can be transferred without modification to another. Ethnocentric research is the search for similarities between cultures, exploring the features of home country culture that can be used abroad. Comparative research searches for both similarities and differences between cultures in order to determine which values hold across cultures and which do not. Comparative research seeks different prescriptions for behavior in each culture depending on the values and motives uncovered. The most encompassing comparative research study in recent years was that conducted by Geert Hofstede (1980, 1983). In his book "Culture's Consequences", and in a follow up paper, he describes a massive 53 country, 160,000 subject study on a multinational corporation. Based on the data collected from an attitude survey of cultural values administered to employees, Hofstede was able to isolate the following four dimensions of culture:

How much woe when we go? International Journal of Stress Management, Spring 1996

3

INDIVIDUALISM/COLLECTIVISM - the nature of the relationship between an individual and fellow human beings. POWER DISTANCE - the acceptance of the institutionalized difference in power between individuals. UNCERTAINTY AVOIDANCE - the degree of tolerance toward the unpredictable, or to deviations from the norm in behavior. MASCULINITY/FEMININITY - the acceptance of masculine (assertive, acquisitive) values or feminine (supportive, cooperative) values. After isolating these dimensions, Hofstede quantified the dimensions for each country based on the specific values exhibited by survey respondents. National cultures were categorized by the interplay of these dimensions. While it is true that Hofstede's subjects were all employees of large businesses, he felt that they were representative of "middle class" values in the societies in which they belonged. Most of the people whom a typical business traveler would meet would come from this broad group.

METHOD A graphic display of the similarity between cultures can be generated by performing a cluster analysis of Hofstede's numerical measures for the cultural dimensions of each country. Cluster analysis is a technique often used in the organization of complex data into meaningful structures. (Schiffman, et. al., 1981) A quantitative index of cultural differences is achieved through the pairing and matching of the foreign cultures into related groups. In this method the numerical scores for individual dimensions are normalized. A computer program searches through the list of countries and finds the two whose values are most alike. The "normalized" distance between the numerical values for this pair are calculated and stored. As soon as two countries are paired together, they form one single unit with averaged values and the list of items is reduced by one. This new combined unit is allowed to find its own matches. Each group always starts off with one pair. The pair grows by adding new countries and country combinations. The program continues to compare the groups and in the end it relates all of the countries in a single unit. The graphic result of this matching is a tree diagram based on the similarities between items. Figure 1 shows such a diagram plotting the relationship between national cultures. For the sake of clarity, the tree is shown lying on its side. The numbers at the nodes or junction points represents the value distance between the joined groups. The actual number is arbitrary, depending upon the scale used to quantify the values but it is linear and provides the ability to make ratio scale judgments of cultural distance. Thus a node with a value of 2 represents twice the cultural distance of a node with the value of 1.

How much woe when we go? International Journal of Stress Management, Spring 1996

.21

4

.88

.66

1.13 1.54

.68 .73

.10

.57

.6

2.80 2.95

3.47

.73

5.56

.70

.46 1.11

6.50

2.10 1.28 .08

.40

.50

.74

6.82

.85 .88

1.91 3.36

.13

6.49

.86 2.80 .79 1.69 .41

.87

Figure 1. Cluster tree of cultural similarities. Hofstede ‘80 value dimensions. Figure 1 supports the classification of national cultures into groups, with similar cultures clustered together. Anglo-Saxon , Western European, Hispanic and Asian clusters emerge. Japan is revealed as culturally unique, more distant from all other cultures than all are distant from each other. An estimate of probable cultural shock can be made by finding the numerical value of the node on the shortest path joining the home country and the visited country. The smaller this distance, the greater the cultural similarity and the less the culture shock that one would experience upon travel to this country. Thus the United States (USA) and Australia (AUL) with a cultural distance of .08 are quite similar, the United States and Japan (JAP), with a distance of 6.82 are culturally remote.

How much woe when we go? International Journal of Stress Management, Spring 1996

5

Our factor analysis of Hofstede's 1980 data discovered that three of the dimensions were highly correlated and represented essentially the same factor. This factor (F1), was labeled AUTHORITARIANISM. It contains most of the components that Hofstede ascribes to the individualism/collectivism dimension, the uncertainty avoidance dimension, and the power distance dimension. The second factor (F2), was labeled "MACHISMO" or GENDER DIFFERENTIATION and consists of most of the masculine/feminine components of assertiveness, aggressiveness, acquisitiveness, and achievement. The emergence of only two factors permits plotting the results on a two dimensional map. Table 1 below, shows these factors. COUNTRY LABEL

Argentina Australia Austria Belgium Brazil Canada Chile Columbia Denmark Finland France Great Britain Germany Greece Hong Kong India Iran Ireland Israel Italy Japan Mexico Netherlands Norway New Zealand Pakastan Peru Phillipines Portugal South Africa Singapore Spain Sweden Switzerland Taiwan Thailand Turkey United States Venezula Yugoslavia

ARG AUL AUT BEL BRA CAN CHL COL DEN FIN FRA GBR GER GRE HOK IND IRA IRE ISR ITA JAP MEX NET NOR NZL PAK PER PHI POR SAF SIN SPA SWE SWI TAI THA TUR USA VEN YUG

HOFSTEDE'S Individual Power Collective Distance 45 50 90 35 55 11 75 65 38 67 80 40 22 62 13 66 75 14 62 35 70 69 89 31 67 35 33 56 24 68 48 78 40 55 71 26 55 13 77 49 45 55 30 82 81 36 68 30 79 19 14 54 15 62 31 94 25 61 65 48 20 75 53 58 71 30 69 35 15 56 21 61 36 64 91 38 13 82 26 76

DIMENSIONS Uncert. Masculine Avoidance Feminine 84 54 51 63 68 79 94 52 77 48 46 52 84 29 79 65 26 16 59 29 84 41 34 68 64 62 110 57 29 59 39 57 60 42 34 69 78 46 75 72 95 98 83 72 52 20 51 11 45 60 68 50 88 40 44 67 103 30 45 67 8 47 84 40 28 7 59 68 70 45 64 34 84 44 43 65 78 75 88 19

Table 1. Factor weightings of Hofstede ‘80 value dimensions.

FACTOR ANALYSIS Factor 1 Factor 2 Authoritarian Gender Diff. 0.24 0.343 0.769 -1.28 1.39 -0.923 0.393 0.245 -0.109 0.749 0.156 -1.046 -1.098 1.051 0.57 1.208 -1.747 -1.759 -1.009 -0.621 -0.216 0.296 0.916 -1.585 0.646 -0.671 0.449 1.051 0.099 0.365 0.21 0.271 -0.486 0.224 0.849 -1.357 -0.157 -0.717 1.262 -0.413 2.433 0.567 1.051 1.285 -1.363 -1.043 -1.89 -0.936 0.509 -1.499 -0.218 0.792 -0.588 1.228 0.633 0.994 -0.938 1.233 0.795 -0.619 -0.604 0.311 -0.38 0.377 -2.18 -1.305 0.924 -0.782 -0.442 0.848 -0.963 0.77 -0.284 0.821 0.836 -1.345 1.068 1.531 -1.523 1.341

How much woe when we go? International Journal of Stress Management, Spring 1996

6

Figure 2 is a "similarity" map of world cultural patterns using the new factors. The distance between countries represents their factorial cultural value distance. Most of the countries are near their neighbors on the branches of the cultural tree. Japan is alone on the right hand side of the map. The greatest linear distance is between Japan and Sweden, implying that the greatest mutual cultural shock would be between those countries.

Figure 2. “Similarity map” of world cultures based on value factor weightings.

The national cultures still cluster together according to similarity groupings. While not quite as amenable to prescriptive use as Hofstede's original conception, an astute traveler could assume that culture shock would be small between countries close together on the graph, but might be greater for national cultures more distant. A different, but perhaps more basic, approach to graphic representation of culture differences is to forego the use of interpreted or factor analyzed data entirely. Such processed data filters subject responses through the bias of our own culture. While the values survey module in Hofstede's original data collection instrument was nominally exploratory, his interpretation of the results and the derived dimensions showed strong influences of then contemporary (1970's) management theory.

How much woe when we go? International Journal of Stress Management, Spring 1996

7

Fortunately Hofstede published the national average raw data for his studies. Using the national means of the 14 value relevant questions in the survey for the 53 covered countries, it is possible to compute the distances of countries from each other in a hypothetical value space. By plotting the geometric distances between this n-dimensional set of data points and presenting them as a two dimensional map, it is easier to visualize the relationships inherent in this complex data set. (Kendall, 1971; Spencer, 1986) Distances between cultural value points I and J are computed using the formula: Dist. = ((value 1 i - value 1j)^2 + ((value 2 i - value 2j)^2 + ((value 3 i - value 3j)^2 + ...((value ni - value nj)^2)^.5

Such a computation leads to a very large table of relative distances. For a list of N data points the number of distances computed is N(N-1)/2. Ten points yield 45 distances, 20 points yield 190 distances, 40 points yield 780 distances. The table of 1378 cells generated for the 53 countries in Hofstede's data make for difficult interpretation. However, just as a highway map can be generated from the table of distances in the highway atlas, a cultural distance map can be generated from the value distances. Figure 3 is just such a map. The national cultural groups are arranged so that the distances between each are proportionate to their geometric distance apart in the value space. Representation of this multidimensional space on a two dimensional surface is subject to error. In plotting such a distance map, the computer attempts to arrange points in a form which satisfies the requirements of the table. Since it is unlikely that the two dimensional map projection will match the multidimensional nature of the data, the computer rearranges the plotted points in successive iterations until the closest match is achieved. The criterion for ending the iterative process is a decrease in percentage error between map and table of less than 0.1% in successive iterations. The computer program that generated the map adjusted distances in an iterative fashion to minimize the overall root mean square error. In this case, it took 39 iterations to reduce the error to 23.4%. The fact that this error is relatively small gives credence to the two factor interpretation of Hofstede's data. Comparison of Figures 2 and 3 show that many nations are in the same relative position. Similar affinity blocks emerge. The bullseye superimposed on the map represents the variability of cultural distance. Individual national value distances are compared with a hypothetical average set of human cultural values. In essence, this value set is that of the trunk of the value tree of Figure 1. Each ring of the bullseye is one standard deviation. The inner ring of the double ring pair is at 1.95 standard deviations. The center of the bullseye is at the mean of human values. Ironically, the nation with the value set closest to the average of all humanity was pre-revolutionary Iran. If the standard deviation bullseye is centered over the home country, it is possible to graphically depict the possibility of culture shock. This is a simple linear transformation of the data. By adding a constant to each value set, the origin of the bullseye is shifted without changing the standard deviation.

How much woe when we go? International Journal of Stress Management, Spring 1996

Figure 3. World cultural map based on Hofstede ‘83 raw data. The additional countries included in this analysis are: ARA COS EAF EQA GUA

Saudi Arabia Costa Rica East Africa Equador Guatemala

IDO JAM KOR MAL

Indonesia Jamaica Korea Malasia

PAN SAL URU WAF

Panama El Salvador Uruguay West Africa

8

How much woe when we go? International Journal of Stress Management, Spring 1996

9

Figure 4. World cultural map. Standard deviation rings centered on USA. Figure 4 shows the bullseye centered over the value set of the USA. Relatively little culture shock will be experienced for countries within the one standard deviation ring. The traveler to Australia, New Zealand, Great Britain, Canada, or Ireland may find some value differences but hardly enough to be described as shocking. Most western European countries lie within the two standard deviation ring, France, Spain, and Belgium being the exceptions. As expected, most Asian and third world countries are three or more standard deviations distant in value space from the USA. The traveler can anticipate a high degree of culture shock on an extended visit to such countries.

How much woe when we go? International Journal of Stress Management, Spring 1996

10

Minimization of culture shock will be a matter of everyday concern when the European economy is combined by the end of this decade. Figure 5 is a culture tree showing the members of the European Community plus several neighboring countries. The United States is included for comparison. The tree shows that Europe is divided into three main cultural blocks, probably determined by geography, religion, and amount of industrialization.

Figure 5. European culture tree. Table 2 is the value distance table of European countries. By entering Table 2 with the home country and reading across to the visited country the probable amount of culture shock can be predicted. Both the actual value distances calculated from Hofstede's national average data and the adjusted distances are shown. Because European countries share many values, the adjustment necessary to plot a map has been reduced to 19.5%. This is analogous to the reduced distortion that a two dimensional geographic map displays when representing a smaller portion of the earth's surface.

How much woe when we go? International Journal of Stress Management, Spring 1996

11

Table 2. Value distance table for European countries. The scale units on the table are arbitrary measures of distance in the value space but they are clearly proportionate to the cultural adjustment that the traveler would be required to make on a trip from the home to the visited country. An American visitor to Great Britain would be required to make only 12 units of adjustment. Visiting Yugoslavia, the same traveler would be required to make 99 units of adjustment. Switzerland and Germany present the smallest value difference between countries. Except for passing through a passport control on the German - Swiss border, the traveler might be hardly aware of the eight units of cultural difference. Figure 6 is a "similarity" map of Europe based on the data of Table 2. With the standard deviation bullseye centered on the hypothetical mean European culture, the more industrialized and/or Protestant ethic countries are found in the left two quadrants, the less industrialized and/or Catholic countries are found in the right two quadrants. The adjustment necessary to draw the map in two dimensions can be seen from examining the difference between equivalent cells of Table 1. The observed value distance of 12 units between the USA and GBR has been adjusted to 10 units on the map. The distance of 8 units between GER and SWI is adjusted to 7 units. The distance of 99 units between USA and YUG is represented on the map as 102 units. Given the typical imprecision of value measurement, the 19.5% distortion of the value space is negligible. Small differences between countries infer little culture shock; large differences, significant culture shock.

How much woe when we go? International Journal of Stress Management, Spring 1996

12

Figure 6. Cultural similarity mape of Europe. Hofstede ‘83 data. The map is interesting from a geopolitical perspective as well as a cultural one. The major country closest to the value average is Germany, a fact which helps explain its position as the major player in the European Community. France and Italy are more than one standard deviation from the European mean and Great Britain is exactly on the 1.95 standard deviation ring. Given the accepted statistical interpretation, this implies that the British perceive their cultural values as distinctly different from those of the average European and may well explain their emotional reluctance to join the Common Market. Denmark, even more distant from the average than Britain has already voted to reject total consolidation. Yugoslavia, geographically close to the rest of Europe, is perceived as a cultural outsider. European diffidence to the political crisis in Yugoslavia may be attributed, in part, to a lack of cultural empathy. An American perspective on probable culture shock is graphically represented by the similarity map of Figure 7. Recentering the bullseye on the USA reveals that the American traveler would experience negligible culture shock visiting Ireland and Great Britain and significant culture shock in Yugoslavia, Portugal, and Greece.

How much woe when we go? International Journal of Stress Management, Spring 1996

13

Figure 7. Cultural similarity map of Europe, standard deviation rings centered on USA.

DISCUSSION The concept of the value distance of cultures, as presented in this paper, is useful for many purposes. Nominally addressing the culture shock issue, it has applicability whenever products, processes, or politics from one culture are to be transferred to another. Marketing managers need the information to predict the successful transfer of products or consumer goods. Organizational psychologists could make use of the concept of cultural distance to determine the acceptability of management policies and/or the additional personal stresses encountered when employees native to one culture are transferred to another. American political scientists could use quantitative cultural information to determine which US mediated interventions have the best chance of success.

How much woe when we go? International Journal of Stress Management, Spring 1996

14

The described approach to the quantitative and graphic representation of culture is a universal one but the specific representations of cultural groups in the value space and the tables, maps and prescriptions derived therefrom are dependent on the database of comparative cultural values used. Hofstede's data set is oriented primarily toward middle class values important in management decision making. It was used in this paper both because it was familiar to the behavioral science community and because it was available in a sufficiently quantitative form to be readily analyzed. Other value data sets, such as the more general ones included in the Cross Cultural Index, would almost certainly lead to a modified arrangement of cultures in the value space. Additionally, national cultures are not as monolithic as represented in this paper. Witness the diversity in values between Canadian French and Canadian English; between Californians and New Englanders; between New Yorkers and everyone else. Given suitable data, quantitative representations of the intranational value space can be as provocative as the international value space. Technical note: Cluster analysis and similarity mapping are performed by a matched set of computer programs available from the author.

How much woe when we go? International Journal of Stress Management, Spring 1996

15

REFERENCES Adler, N. J. (1984). Understanding the ways of understanding: Cross cultural management methodology reviewer. Advances in International Comparative Management, Vol. 1, pp. 31-67. Hofstede, G. (1980). Culture's Consequences: National Differences in Thinking and Organizing. Beverly Hills, CA: Sage. Hofstede, G. (1983). The cultural relativity of organizational practices and theories. Journal of International Business Studies, Vol. 14 (No. 2), pp. 75-90. Kendall, D. G. (1971). Construction of maps from 'odd' bits of information. Nature, 231, pp. 279-284. Schiffman, S. S., Reynolds, M. L., and Young, F. L. (1981). Introduction to Multidimensional Scaling. New York: Academic Press. Spencer, R. (1986). Similarity mapping. Byte, Aug. '86, pp. 85-92