Re: GNOME user survey 2011

[Shaun McCance <shaunm gnome org>]

On Mon, 2011-08-01 at 16:33 +0300, Felipe Contreras wrote:
> On Mon, Aug 1, 2011 at 4:24 PM, Shaun McCance <shaunm gnome org> wrote:
> > On Mon, 2011-08-01 at 12:21 +0300, Felipe Contreras wrote:
> >> But what if you get:
> >> 2% users answered 'Too many options'
> >> 10% users answered 'just enough'
> >> 88% users answered 'few options'
> >>
> >> I repeat, the worst that could happen is that the results of the
> >> question don't provide any value, so you wasted one question... big
> >> deal. You remove it in the next one.
> >
> > That is still not useful information. Developers aren't going
> > to add options for the sake of adding options. "Users want
> > more options? I guess I better hunt through my program to see
> > what I can make configurable." That's absurd.
> >
> > We need to know what users want to change, and (importantly)
> > why they want to change it. Aggregate statistics on this, even
> > if accurate and significant, are not actionable.
> 
> Oh, so you agree that a lot of options are missing?

I didn't say that. And that statement illustrates to me that you
have a very strong agenda. I don't think somebody with an agenda
can design an unbiased survey.

> Good, so what are you doing to identify those options?

As a member of the documentation team, I regularly run ad-hoc
user tests to determine (1) what aspects of our software need
better explaining, and (2) how well our help actually helps
people. I regularly discuss findings with the design team,
and I file bugs.

--
Shaun