Page 389 - Cultures and Organizations

354 CULTURES IN ORGANIZATIONS

All sixty-one “Where I work . . .” questions were designed on the basis
of the information collected in the open interviews and were subjected to
a statistical analysis very similar to the one used in the IBM studies. They
produced six entirely new dimensions: of practices, not of values. What
was used was a factor analysis of a matrix of sixty-one questions by twenty
units; for each unit, a mean score was computed on each question across all
respondents (who comprised one-third managers, one-third professionals,
and one-third nonprofessionals). This analysis produced six clear factors

reflecting dimensions of (perceived) practices distinguishing the twenty
organizational units from each other. These six dimensions were mutually
independent; that is, they occurred in all possible combinations.
Choosing labels for empirically found dimensions is a subjective pro-
cess: it represents the step from data to theory. The labels chosen have been
changed several times. Their present formulation was discussed at length
with people in the units. As much as possible, the labels had to avoid sug-
gesting a “good” and a “bad” pole for a dimension. Whether the score of a
unit on a dimension should be interpreted as good or bad depends entirely
on where the people responsible for managing the unit wanted it to go. The

terms finally arrived at are the following:
1. Process oriented versus results oriented
2. Employee oriented versus job oriented
3. Parochial versus professional
4. Open system versus closed system
5. Loose versus tight control
6. Normative versus pragmatic

The order of the six cross-organizational dimensions (their number)

reflects the order in which they appeared in the analysis, but it has no

theoretical meaning; number 1 is not more important than number 6. A
lower number only shows that the questionnaire contained more questions
dealing with dimension 1 than with dimension 2, and so on; this can well

be seen as a reflection of the interests of the researchers who designed the
questionnaire.
For each of the six dimens ions, three key “Where I work . . .” questions
were chosen to calculate an index value of each unit on each dimension,
very much the same way as index values in the IBM studies were com-