For three weeks through May and June I ran a Web Survey directed at the active members of the Ubuntu community. I got more than 290 valid responses, and I’ve set up a page to share the results with the Ubuntu community.
Not having done any major statistical surveys before, this was quite a change of pace for me – and I’ve certainly learned a lot about the finer points of quantitative analysis. Among other things, how to phrase your questions to get the kind of answers you’re looking for.
I started out with lots of comment boxes, soft interpretation of questions and giving the respondees the opportunity to write in the margin, make small comments and generally give feedback at every step. It was a good way to find out what worked in the survey, but it also made it very difficult to process the bits that didn’t work as well. As the quote goes:
â??There are three kinds of lies: lies, damned lies, and statisticsâ?ť
And once you start “massaging” data to get nice pie diagrams and meaningful percentages, it is very easy to just modify the data as well. Naturally, I didn’t do so, I just noted how easy it is to do – even unknowingly. Quantitative data seems to be just as liable to suffer from over- interpretation as qualitative data.
Now it’s out there, and hopefully it’ll be of some use. In any case I’ll be able to use for my thesis.
Now I’m off to LĂ¦sĂ¸ for a few days, and will be completely off the grid until Sunday – something that has become increasingly difficult to do as I get more involved in this brave new on-line world.