Exploring the national and organizational culture mix in service firms

There are three advantages of this semi-idiographic approach to measuring organizational culture: (1) since each item is implicitly compared to every other item, the outcome of a Q-sort is a realistic profile with items arranged in an order that reflects the relative importance of each item to each other item; (2) culture strength is captured by assessing the similarity of members’ perceptions of organizational values (tested with reliability coefficients and interrater correlations) and the intensity with which values are held (examining the most extreme items such as the top and bottom items); and (3) meaningful comparisons across profiles (individuals, firms, or individuals and firms) are possible (Chatman 1989). On the other hand, one possible drawback of using a Q-sort method to assess organizational culture is that items are not strictly independent of one another. That is, raters are constrained in the number of discriminations they are allowed to make. But computing the number of different ways in which the 54 culture items can be arranged into the designated categories reveals that there are many different ways to sort the items (3.1 × 10⁴²). Thus, any two items are relatively independent (the intercorrelation of each item with each other item is approximately − .02). Researchers who use the Q-sort method argue that item analysis, in a conventional rating scale, is therefore acceptable (Block, 1978).

The individual samples were comprised of the following quadrads: 75 sets of two buyers from a Japanese vendor firm and two of its buyers; 75 sets of two sellers from a U.S. vendor firm and two of its buyers; 52 sets of two sellers from a Japanese vendor firm in Japan and two of its buyers and 52 sets of two sellers from a subsidiary of that firm located in the U.S. and two of its customers; 72 sets of two executives form a U.S. vendor firm in the U.S. and two of its customers; and 72 sets of two executives from a subsidiary of that firm located in Japan and two of its customers.

Several steps were taken to discover the linkages among organizational culture dimensions and employee behaviors. First, the common behavioral characteristics stemming from each cultural dimension were identified. Second, two randomly-selected employees (with at least five years’ employment with their respective firms) from each participating firm were asked to complete a questionnaire measuring the extent to which each of the behaviors characterize their firms (a nine-point scale was used, ranging from 1 to 9 “(un)characteristic of employees’ behavior in my firm”). An average of the two responses within each firm was used to build the scales (there was a significant correlation (p ≤ .05) on all measures within the dyad). The behaviors were then regressed on the cultural dimensions, and all results were significant: aggressive and the behaviors of experimental, uses nonstandard approaches, and takes risks, R² = .84, p = .01; stable and the behaviors of predictable, secure, unchanging, R² = .91, p = .00; people oriented and the behaviors of friendly and helpful, R² = .96, p = .00; outcome oriented and the behaviors of goal oriented and demanding, R² = .90, p = .00; detail oriented and the behaviors of precise, analytical, careful, emphasize quality, R² = .89, p = .00; team oriented and the behaviors of collaborative, shares information freely, and lengthy decision making, R² = .88, p = .00; and aggressive and the behaviors of takes initiative, action oriented, decisive, opportunistic, conflictual, and impatient, R² = .94, p = .00. Thus, it appears that the organizational culture items in the scale used here reflect the actual behaviors of employees.

The authors thank an anonymous reviewer for these comments.