I couldn’t help but repost this Aaron Levenstein quote from Brand Autopsy. I’ve been thinking about writing a post on the death of critical thinking for a while, but this pretty much sums it up. Statistics – and research in general – is carried out by *people*. People have motivations. Understand their motivations and you understand a lot about what they have done (either unwittingly or deliberately) to achieve the results presented to you.
Related posts:
[...] via 90% of everything und dieser auf Brand Autopsy. Written by PW – the webconsulSeptember 22nd, 2010 at 2:31 pmPosted [...]
The same is true of any human activity, including design. So we can also say the following: “Interaction design is carried out by *people* People have motivations. Understand their motivations and you understand a lot about what they have done (either unwittingly or deliberately) to achieve the results presented to you.”
Kev – sure, but there’s something about stats that give a feel of ‘irrefutable, objective fact’ when in fact they’re just numbers. It’s the details of the research from which they’re generated that matters.
Harry- do you have any examples to illustrate what you mean?
Kev, how about this for a recent example?
This is why Theory first is so important. What this means is that you state your theory, and then test it. Where allot of UX research is challenged is that most methods claim to be theory last. This is where you observe and then come out with a theory. I say claim as I believe that allot of practitioner have a theory before they start. Sure it is possible to do theory last, with rigour, but it takes more time than most of people in our industry have.
It gets very dangerous when you mix theory last with statistics. “Correlation does not imply causation”.
Most explanations, the why, of the natural world have come about through “theory last”. I get slightly annoyed when people claim that “theory last” can’t explain the why.
Can’t wait for your post on Critical Thinking.
Inside the pages of Freakonomics and other books like that are countless examples of “bikini statistics” that reveal interesting stuff but the vital information is in what those statistics don’t reveal.