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

Values. Twenty-two questions assessed work goals: the characteristics of an ideal job, like "have an opportunity for high earnings" or "have security of employment," were each rated on a 5-point scale of importance. These were taken from the earlier cross-national research project and from later extensions of it. The interviews revealed no additional goals to add to the list.

Twenty-eight questions assessed general beliefs, like "competition between employees usually does more harm than good," each rated on a 5-point scale from "strongly agree" to "strongly disagree." Twenty-five of these stemmed from earlier cross-national research, mostly from the IBM studies and from Laurent (1983). Three were added based on the interviews.

Both work goals and general beliefs deal with values, but work goals represent "values as the desired" (what people claim to want for themselves) while general beliefs represent "values as the desirable" (what people include in their world view) (Hofstede, 1980: 20). Although items from the two categories tend to intercorrelate, answers are not necessarily logically consistent from the first category to the second (Hofstede, 1980: 21), and neither of the two is a perfect predictor of actual value-driven behavior in a choice situation. However, differences in verbal behavior (in questionnaire answers) between cultures do correlate with measures of collective actual behavior, at least in the national case (Hofstede, 1980: 328 ff).

Seven other questions were included, in a variety of formats, on other items statistically correlated with the previous values items, including questions on desired and actually perceived decision-making styles in one's boss. Five of these occupied a key role in the earlier cross-national research; the other two were added on the basis of the interviews.

Practices. Fifty-four questions assessed perceived practices in one's work situation. The first fifteen of these were inspired by Reynolds (1986), who did a thorough scan of the anecdotal U.S. literature on corporate cultures for suggested dimensions of differences. To Reynolds' questions, we added another 39 based on the interviews. We then cast all 54 questions into a bipolar format under the general heading "where l work . . ." and used 5-point scales on which, for example, 1 = "meeting times are kept very punctually" and 5 = "meeting times are only kept approximately." These 54 questions mostly cover symbols and rituals.

Seven questions asked about the "behavior of a typical member of the organization," using a 5-point "semantic differential" scale on which, for example, 1 = "slow" and 5 = "fast."

Thirteen questions asked about reasons for promotion and dismissal, rated on 5-point scales of importance or frequency.

Both the typical-member and the promotion-and-dismissal questions cover the category of "heroes" and were inspired by the interviews.

Four demographic questions asked about the respondent's sex, age group, seniority with the employer, and education level. Finally, there was an open question, asking the respondent for any additions or remarks.