From fuzziness to clarity

In order to decrease fuzziness and to increase clarity one needs to meet at least three important requirements:

Define organisational culture in an operational manner.
Measure culture(s) precisely; i.e. in numbers.
Create a balance between over-simplification and too much complexity.

The first requirement is to define organisational culture in an operational manner. Our definition is: Organisational culture describes the way people in an organisation relate to each other, to their work and to the outside world in comparison to people in other organisations. This definition includes: "The way we do things over here", but it covers more ground. It also includes attitudes and beliefs. "Core values" are not included in our definition. To us "core values" equate beliefs, but "core values" sounds so much better. The problem with "core values" is that these words give the wrong signal. People rarely change their "values", let alone their "core values", whatever that may be. People may sometimes change their beliefs if there exists a compelling reason to do so. The good news is, that this does not mean that culture cannot be changed. Change is especially about practices, i.e. about the way we do things over here. What employees believe is for the work situation less important than that they do a proper job. Our definition also includes the word "comparison". Without comparative data it is hard if not impossible to give meaning to the data collected about just one culture. It is the same as in the case of a world in which everything is blue. In such a world the notion of color would not exist, making the characteristic "blue" irrelevant. In order to compare one needs a data bank. We have collected data since we started our quick scans at the end of the eighties of last century.

The second requirement is to measure culture(s) precisely; i.e. in numbers. With the help of our quick scan we can position the culture(s) of organizations in precise numbers. Yet, two warnings should be given:

Complexity of life cannot be captured in two numbers behind the comma. Numbers should be understood as representations of comparative tendencies. Thus a score of 86 on the dimension of our Hofstede model, called "internally versus externally driven", should be read as "this culture supports customer focus to a very strong degree compared to most other organizations".
Given that groups of people belong to the most complex systems we know, one way of data collection may not always suffice to get the culture of groups fully into focus. Thus, we are prepared and we are able to use additional data collection processes next to on-line measurement of the actual culture, such as in-depth interviews.

The third requirement is to create balance between over-simplification and too much complexity. It is hard to strike exactly a border line between over-simplification and too much complexity. Use of "good" models will make it easier to strike such a balance than by only collecting descriptive information e.g. with help of in-depth interviews. A good model is characterised by the following:

It should consist of autonomous dimensions (variables). Autonomous means that scores on one dimension will not predict the scores on another dimension and the other way round. Only by creating autonomous dimensions will one ensure that work reality will be described to its fullest extent possible.
It should indeed cover work reality in the broadest sense possible. If a client would require measurement of an aspect which is not already covered by the model, one cannot compare the findings in the case of this specific client, making it hard if not impossible to give meaning to those specific findings.
Most psychologists nowadays agree among themselves, that a personality is being best described by five autonomous dimensions, called "The Big Five". A group of individuals is more complex than one individual.Therefore it seems to be a fair assumption that a good model should consist at least of six autonomous dimensions. This applies to our Hofstede model.