What do you mean, “qualitative”?

If anyone paid attention to what I say, they’d probably notice that I tend to display some skepticism towards data and data-driven conclusions. I’ve even spoken (twice) about empathy as a business tool – this is probably some new-age thing, right?

Nope. I’m actually quite the science nerd, and have been all my life. However, I love science, not because it’s data driven, but because it’s about finding out. And, even more crucially, about finding out how to find out.

It’s just that, or my profession, I favor something called the Qualitative Method. 


What is that? Well, thanks for asking, ficticious reader! Since you’re obviously a bright person, you can probably guess it must have some relation to the concept of “quality”, and you’d be right (see? I said you were bright).

Its counterpart is called the Quantitative Method, and you can probably already see how this works, just by looking at those words:

Qualitative, as mentioned, has something to do with quality, specifically the qualities of experience.

Quantitative has to do with quantity, i.e. things you can count, weigh and measure.

These days there’s a lot of talk about data. Interestingly, while the concept of Big Data is obviously “a quantitative endeavour”, so to speak, using the qualitative method often involves what we sometimes call Small Data. Where Big Data can involve thousands or millions of people, Small Data might look at just a few dozen – but this small pool of data is very deep.

This has to do with the kind of data we’re talking about, and the ways it’s collected. The kind of data we collect is very much about the “why” of people’s actions and behavior, and about how they feel, which means it tends to be deeply subjective. Which is to say, if five people do the same thing, at the same time, in a way that appears identical to an observer, they can (and often will) have five individual experiences of why they’re doing it, or doing it in this particular way, at this time etc.

Subjectivity is usually a problem for data – but the thing is, in any market, business, organisation or society, very few things have as much impact as people’s subjective experiences. Dismissing them from research and knowledge simply because they’re difficult to handle with quantitative methods means blinding oneself to one of the most important factors of all. Hence we have this qualitative method. 



Qualitative research is carried out using inter-personal techniques – observation, conversation, participation and so on – in the field or under controlled circumstances. The lack of obvious technicalities in these techniques, which look a lot like “hanging out and talking”, often leads to concerns that it’s imprecise and biased. This is a valid point, but it has little to do with the method and techniques; the notion of “neutral data” is largely a myth, and proper care must be taken to address biases and sources of error, regardless of method. In qualitative research, this is done by preparing properly, using multiple observers,  recording the processes carefully so that subjective assumptions can be checked against the actual events, and in a host of other ways.

Just like with the “harder” methods, it’s all about applying discipline and rigor, in preparation, execution, and analysis & interpretation. 

Another concern is, of course, the small sample size. How can we assume, from talking to 20 people, that what we find can be applied to tens or hundreds of thousands? Well, again, this requires skill with the method because yes, it’s true that you can solicit opinions from 20 people that are completely their own and have no bearing on anyone else’s behaviors or choices. On the other hand, people are, in fact, people, and it’s not as if everything about any one of us is a complete mystery until we ask – each of us is a person ourselves, and we already know that we share certain experiences with people around us, just by that virtue.

If I may be literary for a moment, I read an interesting passage in a novel once that explains how this works – the book is “The Name of The Rose” by Umberto Eco, and the passage went something like this (from memory, so, sorry, Eco):

– from a certain distance, if you see an animate shape, you’ll be satisfied it’s some sort of animal, even if you can’t tell which kind. A little closer, you can see it’s a farmyard animal, though not which kind. Closer, and you see it’s a pack animal, though horse, ass or donkey, you can’t tell. Closer yet, and you can see it’s a horse, though not if it’s Brunellus or Favellus. Get close enough, and you can see the individuality of that animal, and that it can be no other.

So, translated to the topic at hand, it’s about choosing your distance; too far, and the results are too general, too close and they’re too specific. The right distance depends a lot on the particular job – carrying over from the  passage, one might consider the choice of distance depending on whether the question is “how many legs on a horse” or “what color is Brunellus”.

Another way I’ve heard this issue addressed was by prof. Peter Dahler-Larsen from the University of Copenhagen, who says “- if you do an experiment with, say, hydrogen, in the lab, you’re working with a very small sample of all the hydrogen in existence – but that doesn’t mean your research with that sample is irrelevant”. Using the proper experiment design, we can indeed say some things about all of it by looking at just some of it.

It bears mentioning that the qualitative research method features heavily in design thinking, UX, service design and certain parts of the medical world – basically, anywhere it may be interesting, or imperative, to find out about human actions, reactions and behavior.

So now you know. And, as usual, feel free to contact me if you want to know more.