Page 122 - Cultures and Organizations

I, We, and They 101

clusters, two of these were found to be highly significantly correlated with
IDV: autonomy versus embeddedness, and egalitarianism versus mastery. 9
The GLOBE study defined and tried to measure two categories of

collectivism: institutional collectivism and in-group collectivism—both “as
is” and “should be.” Ten out of GLOBE’s eighteen dimensions were sig-
nificantly correlated with IDV, but the dominant correlation was with

in-group collectivism “as is.” GLOBE’s questions in this case dealt with
relatively simple aspects of human behavior, which explains why its mea-
sure came closer to ours than in the case of the other dimensions. IDV
explained 58 percent of the country differences on in-group collectivism
“as is.” In Chapter 3 we saw that in-group collectivism “as is” was also the
10
strongest correlated GLOBE dimension for PDI, but the correlation with
IDV was slightly stronger.
From GLOBE’s other three measures of collectivism, only institu-
tional collectivism “should be” was weakly negatively correlated with IDV
but more strongly with our uncertainty avoidance index (UAI, Chapter
6). Institutional collectivism “as is” was exclusively correlated with our
UAI. In-group collectivism “should be” was correlated with our long-term
orientation index (Chapter 7). 11
Peter Smith’s analysis of the Trompenaars database produced two
major dimensions. Both were correlated with IDV; the second one was also,
12
and even more, correlated with PDI. However, the correlation with PDI

was influenced by the fact that there were no Eastern European, high-PDI
countries in the IBM sample. In fact, the second dimension opposed most
Eastern European countries to East Asian countries, and the question-
naire items involved focused mainly on teamwork, which received positive
associations in China and negative associations in most Eastern European
countries.
Subsequently, an ingenious study by Smith compared not the results

of the various international studies but rather the degree of acquiescence
in their answers. Acquiescence occurs in all paper-and-pencil surveys: it
is the tendency among respondents to give positive answers regardless of
the content of the questions. Smith compared six studies that each covered
thirty-four or more countries, including studies by Geert, Schwartz, and
GLOBE. For sections of the questionnaires dealing with values, all six
studies demonstrated similar acquiescence patterns. Smith showed that the
common tendency to give positive answers in the six studies was stronger
in countries that, according to our measures, were collectivist and had