Re: [Usability] Gnome Research - Sociological Surveys

[Guillaume Ardaud <guillaume ardaud gmail com>]

I whole-heartedly agree with Celeste. As mentionned before, you should definitely have a full survey and not proceed questions by question; the resulting stats will have very little value in the second case.

For example, consider the two questions "How old are you" and "Do you feel you can work faster if the software allows it" (no matter how bad that question is, it's just for the sake of the example).

If you ask your questions separately, you won't be able to draw any conclusion from the results. However if you ask them together, you might be able to see a certain age interval feels limited by the interface and not the other; and from there you'd be able to investigate why this is the way it is.

It is really important that you know WHY you're doing this survey (ie what do you expect to learn from it). Once you have a specific goal in mind, you'll be able to define a certain number of questions that all fit together, instead of a pot-pourri of vague interrogations (for instance "What would you change in Gnome 3.0?" - I'm assuming your survey concerns UI design; imagine that people answered you with technical answers. What use would that be to you?)

Good luck,