I'm not a proponent of quantitative research in the process of studying organizational culture. I prefer qualitative research. The reason is simple. Quantitative research, based on the analysis of previously completed surveys, examines not the actual culture but rather the respondents' perceptions and opinions about it, or the effects of more or less unstructured, though unfortunately quite numerous, confabulations. Furthermore, opinions are usually tinged with preferences and emotions, which undermines their objectivity. Of course, one can put in the effort and develop quantitative research methodology over years, achieving results like those achieved by Cook and others in constructing the Organizational Culture Inventory. However, this doesn't solve all problems. Research in a specific unit can be designed so that even from a theoretically anonymous survey, it's possible to deduce who provided "problematic" answers. Of course, if an external entity (for example, a consultant hired for this purpose) has full control over the research, its development, and the presentation of results, then potentially effective manipulation of internal organizational forces is unlikely. Unfortunately, this fact has absolutely no impact on how survey respondents themselves perceive the research. My experience shows that even when convinced of the survey's 100% anonymity, respondents can consciously provide untruthful answers. Some do so out of fear of overly negative research results that could impact their team, department, or the entire organization. Others, on the other hand, see anonymous surveys as an opportunity to finally get even with others for the wrongs they experienced years ago. On the other hand, it's possible that an initial culture assessment (using qualitative methods) can be conducted in just a few minutes of walking around the company and a few dozen minutes of conversation with management. These first few or several dozen minutes are enough to conclude that you've got work to do for at least several months.

It Seems Easier, But…

I fully understand the interest many consultants have in quantitative research. You conduct surveys, receive responses, apply them to a template, and get the results. The computer calculates them. And the diagnosis reveals that we need to improve the well-being of half the staff. So we sculpt that half of the staff, and that's "only" 150 people. Instead of promoting one abusive manager from the factory floor to an independent expert position with a higher salary, because a few months ago someone decided that since he was a high-class data expert, he would be an equally good expert in interpersonal relations. But he probably won't. Statistically, such sets of radically different competencies don't exist in one brain.

I highly value the creators of OCI mentioned above, but I prefer to conduct a diagnosis of organizational culture through qualitative research, not quantitative research. OCI, as a quantitative survey, is a perfectly reasonable solution when the person diagnosing the organization doesn't plan to engage in long-term or more intensive cooperation with the organization being studied. When it's someone whose professional occupation isn't management, but rather research or teaching, for example, and visits the organization being studied "for a moment," conducts research, suggests some solutions, perhaps coaches for a while, returns a year later, conducts further research, and shows the difference and progress. I greatly respect this approach, but I believe that an organization is made up of people, and it's difficult to understand people by limiting yourself to examining and analyzing statistical data. And OCI, like any survey, is based on statistics. Human emotions are difficult to quantify. Of course, such an attempt is possible, and I appreciate the OCI creators for their contribution, as they did a truly excellent job. It offers a large comparative base, and the concept itself is not only interesting but also intriguing. It's worth exploring if you're not closely associated with the organization for which these activities are being carried out. However, if someone is involved as an interim manager in such an endeavor, not so much for the diagnosis itself, but generally to construct the organizational culture, they should, whenever possible, conduct this diagnosis through observation and conversations with people. It's better to devote more time to this, and even if you don't reach everyone, which is rather obvious in large organizations, the quality of the information obtained will be higher.

Subjectivity Doesn't Help

The most popular quantitative research tools were designed by researchers, some not necessarily from management and quality science disciplines, and certainly not by practitioners, for whom business was the primary activity, but rather research and teaching. What's the difference? The primary role of researchers is research and knowledge acquisition. The primary role of managers is to identify problems and solve them. Within months. Not years.

Moreover, if we divide the results of quantitative studies of organizational culture into those obtained from women and those obtained from men, we will see that women in non-constructive cultures perceive this "non-constructiveness" as greater than men and more severe. In turn, in constructive cultures, women perceive it as more constructive than men. This points to a very significant shortcoming of such studies. We perceive the same things differently, so there is little objectivity. The second problem is that if we combine these results into a single aggregate, we will obtain an average that will have no relation even to the respondents' subjective perception of reality. The overall result will likely be a result that, with a high probability, no one actually obtained, and if they did, it would be very few. We will be dealing with a statistical result similar to the one according to which the typical Polish family has one and a half children and half a dog. And we will focus on developing ideal solutions for such an entity. Will we design cars with three and a half seats? What about the dog? Which half do we have at home? Right or left? Front or back? Should we give him just a bowl or just a litter box? Or maybe half a bowl and half a litter box?

If that's not convincing enough, let's try to objectify the research results when the organization is 71% women and 29% men. Because if we know that the scale of perception of women is broader, and perhaps we should acknowledge that the deviations we observed in women or the narrowings we observed in men are fundamentally inaccurate, how can we determine which of them, or to what extent, are actually inaccurate? Or perhaps all of them? Both in women and men? And is our average meaningful because it is representative of the organization, but not of an individual woman and man in a single team? If we conduct such research on a team of five men and five women, and on another team of eight women and two men (from the same department in the same company), even if all men perceive reality the same way, and women perceive reality the same way, we will get two different results, even if everyone in the gender groups sees exactly the same thing! So what kind of organizational culture do we really have? And do we really need to research this, since all we can hope for is an approximation? The fact that the results we obtain won't be representative of anyone, or even if they are, it will be for a very narrow group, is indirectly confirmed by consultants using these tools.

Survey research can be useful for management teams when it aims to pinpoint a specific individual's leadership style. This allows us to obtain a relatively accurate picture, specific to each individual. For entire organizations, the picture is averaged, and usually asymmetrically averaged if the number of women and men is unequal. Surveying the entire organization yields an aggregated result that, in all likelihood, was not actually generated in any individual's mind.

Human Emotions Are Not Mathematics

I'm not a fan of statistical research for another reason. They average opinions and average emotions. It's never the case that in a company, one person with extremely negative emotions will be neutralized by one person with very positive emotions. Negative emotions have a very high leverage. Therefore, it's possible that focusing solely on quantitative research in a culture diagnosis could lead to biased results in the negative. People emotionally invested in certain phenomena, for example, in a social phenomenon like a company, will not perceive events and situations with a negative emotional tone in the same way as positive ones. Losing 10,000 euro hurts us more than gaining 10,000 euro gives us pleasure. These are theoretically two events that, on a scale from -10, through 0, to +10, should be exactly the same distance from "0." If we assigned a value of 3 to gaining 10,000, then we should "value" the loss of the same amount at -3. Right? Will we do that? No... It will be maybe -4, maybe -5. Survey results may indicate more negative opinions or moods than they actually are. If we place two experiences of equal intensity on a certain scale - one negative and one positive - they will not be equidistant from the neutral point. We experience negative experiences more acutely than positive experiences of equal intensity, value, or impact. I haven't yet encountered a survey that takes these differences into account. This is likely because they are unquantifiable and highly individual, stemming from personal preferences and mental state at a given moment. We can perceive the same negative phenomenon differently, depending on the context and, above all, our mental state at that moment. A survey doesn't examine reality. A survey examines the opinions about reality of people who are emotionally non-neutral towards the phenomenon being studied. Therefore, I am not a staunch supporter of 360-degree research. Unless we want the management in a non-constructive culture (which it wants to change), in a company employing mainly women, to be brought down to earth by the results of research on its management style and employee perception in a way reminiscent of the touchdown of a lunar lander with a damaged system responsible for the controlled slowdown of the descent.

The Best Moment... Usually Never Comes

But these aren't the only reasons I dislike surveys. One is that... those who are thrust into their faces with a suggestion/request/command to complete them don't like them. Have you ever filled out a survey? Did you feel excited when you received it? Did you feel like it was exactly what you needed at that moment? Did you feel like it was the perfect moment to complete the survey, and that nothing more pleasant/important/more anticipated could have happened to you at that moment? Did you feel like you had the time and inclination? Exactly... And do you remember your commitment to completing it? Exactly... How many such surveys are honestly completed? Exactly... Regardless of how well-thought-out and perfectly prepared a survey is, if the respondent isn't genuinely interested in completing it diligently, its results will be of little value. And the researcher is unlikely to be aware of this, because no one will admit that they simply ignored this (so important!) survey.

Everyone has his own preferences. Some organizational culture specialists use surveys in their diagnosis process. My approach is slightly different. I pick holes in the puzzle. I'm not interested in averaging opinions. I'm interested in identifying specific problems. Identifying a specific problem, proposing a specific solution, and implementing it is usually cheaper and faster than addressing that specific problematic factor by influencing the entire organization, because in the aggregated research results, we found a problem of a specific scale. If you own a company and have an abusive manager, give him another job instead of hoping he will change his attitude by influencing the mindset of the entire organization. The average person in the workplace won't change much. The responsibility for being abusive primarily rests with that manager. But the responsibility for making them a manager rests with someone else. Guess who? If they became abusive only as a manager, they shouldn't be fired. They should be given another job unless someone is already at risk of losing their health or life.

In summary: surveys don't reveal the state of organizational culture. They show asymmetrically averaged opinions about culture compared to other asymmetrically averaged opinions about other organizations. For example, the problem with OCI is that this study doesn't show us the absolute degree to which we are constructive or unconstructive in given styles or categories. The survey results only show us how constructive or unconstructive we are compared to other, primarily American, companies.

This implies that quantitative research is an attempt to quantify human emotions. As far as I know, no one has yet succeeded in doing so. Nor do I believe, contrary to most opinions, that quantitative research results are "hard data." Hard data exists in the balance sheet, cash flow statement, and profit and loss statement. If my actions don't translate into financial statement results, then I haven't accomplished anything.

Quantitative research results may indicate the existence of a problem. But they don't precisely identify it. They only indicate the area where the problem occurs, not the specific location. That is, they can indicate that an aggressive, violent individual is located somewhere within the organization. But they won't pinpoint the precise location. Survey research isn't GPS. It's a map, and not a very accurate one at that. To use a map, however, you need to know something about the area you're traveling through.