Research – from data to knowledge

As an Action Researcher I belong to a group of people who distinguish research on nature from research on social systems – while being aware that social systems are a part of nature’s incredible variety. Before I introduce you to Action Research as such, I’ d like to take a closer look to research. Especially on the process of generating knowledge from data, and what makes good research.

An example. My nick Nae73 is data. It includes the 73. When you see it you might – or might not – interpret it as information. E.g., one could assume that 73 is my year of birth. If you read elsewhere that I am 40 it is not too far fetched to say: I know that Nae was born in 1973. You can now even generate new knowledge, if you can calculate and know your year: that I am not born between January and July. Which is true; I am scorpio. TADA, good research! Right?

Wrong! Although in this case the conclusions you draw were right, they are based on the false assumption that I chose 73 to indicate my year of birth. The real reason why I chose the 73 is that 73 is the best number. That I happen to be born in the best of years is a coincidence.

But what is wrong with the initial interpretation, if the knowledge we generated was correct? 

Nothing, actually. You might even find that it meets crucial criteria for good research, which are objectivity, reliability and validity. To keep it short and simple:

  • If the majority of people you ask also concludes that I am born in 1973, you assume reliability.
  • As many people use a number in their nick which does indicate their age, it is appropriate to assume I did so, too. That’s (totally superficial) validity.
  • Ignoring your own fling for Sheldon Cooper when evaluating my nick would show you’re unbiased: objectivity.

In traditional research traditions, it goes like this:

If you want to evaluate me (research topic + questions), you look at the data (methodology): Nae73. You evaluate the reference frame: online nickname, often structured as ‘name+age’ or ‘name+year’ (in research papers: discussion). You check secondary literature, supporting your interpretation of ‘name+year’. You present the finding: Nae was born in 1973 (question answered), in the third trimester of the year (area for further research). Everyone is happy, you met all quality criteria and get your paper published. Whoohoo!

Some Action Researchers of course share this more tradional approach, but we will see later on that researching human beings challenges the quality criteria on all levels. For today, let me focus on objectivity. The thing is not that the unbised approach led you to wrong findings, but it led you to a limited view of your research object. 

In AR, we do not assume objectivity. Some interpret this approach as arbitrary, because a researcher might interpret what she wants, but that is not the case. Denying the possibility of an objective researcher does not mean I can do whatever I want; it means I am doomed to interpret data based on my existing knowledge. You cannot interpret 73 as the best number if you have never heard of the idea of a ‘best number’. But if you have, you are a nerd lover and it is likely that you try to ignore it to stay objective. But why? Maybe your particular frame of reference is exactly my frame of reference, and you are the only researcher who can interpret me correctly? Which would be pointless, if your fellow researchers would not believe you, but that is the fate of all brilliant minds, isn’t it?

And here we approach the ‘poodle’s core‘: knowledge is only knowledge if others share it. This is as true for the traditional research community as it is for the relationship between action researchers and their research object. As an Action Researcher you would aks me which of your potential interpretations is correct – something no traditional researcher would even consider because part of the scientific word view is that ‘we know, and they don’t’. But that makes for another post.

I love research. But I do not like the elitist approach many scientists show. I think that they should write in a language which allows non-scientists to question their work. Especially because some say that politics should consider scientifc findings to a greater degree. Poltics – in democratic societies at least – should not be based on concepts nobody understands. And, for heaven’s sake, it should not be based on concepts treating human beings as ‘objects’!

Cheers, Nae73