On Tue, Apr 15, 2014 at 7:12 PM, Steven Walling <[email protected]>wrote:

> On Tue, Apr 15, 2014 at 9:59 AM, Gergo Tisza <[email protected]> wrote:
>
> >    - instead of guessing about user preferences, you could just create a
> >    simple survey which shows them the same text with two different font
> > stacks
> >    side by side, and ask them which is more readable. This is good for
> > making
> >    aesthetic decisions more objective, and also for catching weird issues
> > with
> >    old machines, CJK fonts etc: you can add a comment field to the
> survey,
> > and
> >    if the browser is sufficiently modern to support canvas elements, you
> > can
> >    even save a snapshot if the rendered text; you can skim through the
> > survey
> >    replies which are different from what you have expected, and look for
> >    display problems.
> >
>
> Are you volunteering to build such a survey tool? ;-)
>
> We don't have a powerful/easy to use/not annoying/privacy-respecting survey
> tool that can do side-by-side comparisons. This is why the feature was
> launched using Beta Features for five months first. Putting out in opt-in
> mode and gathering feedback via the channels we have now is the most
> efficient way to make a change that doesn't have a big WMF team assigned to
> like Multimedia or VisualEditor.
>
> When it comes to using a survey to catch problems early and gauging
> preferences, a survey still very much suffers from the self-selection bias
> that all opt-in options have. It's just the name of the game. When you move
> something from opt-in to opt-out you reach a wider audience and encounter
> new complaints/questions/bugs.
>

What would be a good design for such a survey? Would it be a good idea to
ask surveyees which scripts they regularly read, and for each of those
scripts prepare a bit of text, including hard parts (combining characters
and the such), style it with fontx, sans-serif, and ask questions about the
qualities we are looking for?

If so, what would be the questions to ask? When I read the former tests,
base questions seem to be

* How would you rate the readability of this font?
very/completely unreadable - somewhat unreadable - not specifically
readable or unreadable - well readable - very well readable
* How would you rate the neutrality of this font? (I don't really know what
this means exactly, so a different phrasing is probably better, maybe
something like "do you think this font has a specific style", where less is
better?)
Very neutral/not a specific style at all - somewhat neutral/no of a
specific style - not neutral or non neutral/not much of a specific style -
somewhat non-neutral/a somewhat specific style - very non-neutral/a very
specific style/you just showed me papyrus
* Does this font look authoritative?
Very authoritative - somewhat authoritative - neither authoritative nor
non-authoritative - not very authoritative - not authoritative at all/I
just told you you're showing me papyrus
* Does this font seem to render correctly?
yes - no

Is testing like this a road we want to go down at all? If so, is this
specific format a good idea? Can we improve this idea to make it good?

I don't mind making this in the weekend if it is a good idea.

--Martijn



> _______________________________________________
> Wikitech-l mailing list
> [email protected]
> https://lists.wikimedia.org/mailman/listinfo/wikitech-l
>
_______________________________________________
Wikitech-l mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Reply via email to