Measuring organizational cultures: a qualitative and quantitative study across twenty cases

Administrative Science Quarterly, June, 1990 by Geert Hofstede, Bram Neuijen, Denise Daval Ohayv, Geert Sanders

METHOD

Sample. We attempted to cover a wide range of different work organizations, to get a feel for the size of culture differences that can be found in practice, which would then enable us to assess the relative weight of similarities and differences. A crucial question is what represents "an organization" from a cultural point of view. One organization may include several culturally different departments, and these departments may consist of culturally different work groups. Determining what units are sufficiently homogeneous to be used for comparing cultures is both a theoretical and an empirical problem. We took the pragmatic approach, to accept as units of study both entire organizations and parts of organizations and to follow management's judgment as to whether a unit was culturally homogeneous. In a few cases, the research results later gave us reason to doubt a unit's cultural homogeneity, but it is unlikely that the results have been substantially affected by this. In the end, we got access to 20 units from 10 different organizations, five in Denmark, five in the Netherlands. These 20 units were from three broad kinds of organizations: (1) private companies manufacturing electronics, chemicals, or consumer goods (six total divisions or production units, three head office

or marketing units, and two research and development units); (2) five units from private service companies (banking, transport, trade); and (3) four units from public institutions (telecommunications, police). Unit sizes varied from 60 to 2,500 persons. Twenty units was a small enough number to allow studying each unit in depth, qualitatively, as a separate case study. At the same time, it was large enough to permit statistical analysis of comparative quantitative data across all cases.

Design. The project consisted of three phases. In the first phase, we conducted in-depth interviews of two to three hours' duration each with nine informants per unit, for a total of 180 interviews. These interviews allowed us to get a qualitative feel for the gestalt of the unit's culture and to collect issues to be included in the questionnaire for the subsequent survey. Informants were chosen nonrandomly in a discussion with our contact person(s) in the unit. They included, in all cases, the unit top manager and his (never her) secretary, and then a selection of men and women in different jobs from all levels, sometimes a gatekeeper or doorman, an oldtimer, a newcomer, an employee representative (equivalent to a shop steward). A criterion in their selection was that they were assumed to be sufficiently reflective and communicative to be valuable discussion partners. The interview team consisted of 18 members (Danish or Dutch), most of them with a social science training but deliberately naive about the type of activity going on in the unit studied. Each unit's interviews were divided among two interviewers, one woman and one man, so that the gender of the interviewer would not affect the observations obtained. All interviewers received the same project training beforehand, and all used the same broad checklist of open-ended questions. Interviews were taped and reports were written in a prescribed sequence, using respondents' actual words.