[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-usability
Subject:    Re: The art of not offering customization
From:       Kristian Koehntopp <kk () netuse ! de>
Date:       2002-05-30 13:49:42
[Download RAW message or body]

On Wed, May 29, 2002 at 03:42:36AM -0700, Michael W. Collette wrote:
> > You can reason about that - it does not help. Observation helps
> > to understand.

I wrote this for a reason. Let me explain:

> I'm finding myself agreeing with Aaron (this is happening far
> too often) on the point about not having user levels.  Even
> with the users that you are describing, the problem isn't too
> many options.  Collectively none of us is smart enough to know
> what are the important options for even the most non-technical
> user.  What is important to one user does not translate into
> all, even if they are at the same technical level.  Humans are
> annoying in that way.

I did the experiment with my father, after reading Nielsen's
Alertbox for several years. Nielsen preaches that you must not
make assumptions in usability, just as in any other science, but
that you should take measurements. Nielsen also says, that he
has observed diminishing returns in usability studies after
studying as many as 5-7 test subjects, so using small numbers of
people for studies is a-ok, if you are on a tight budget.

When my father saw KDE on my system and asked for an
installation to try it, I saw this as an opportunity to apply
these ideas by Nielsen myself, and the results were indeed
enlightening. If I am allowed to generalize from a n=1 sample,
nongeek-people work and percept a system completely differently
that I do, and have different goals and solution approaches.
That emphasizes different things in a GUI design that I would if
I do it myself. 

And it makes reasoning about what is logical somewhat pointless.
Because people do not act logical when doing their work on a
computer system, and have non-logical expectations.

If you have the opportunity to introduce a non-geek person to
KDE, you really should try this for yourself. Drop them in front
of the system, and do not help at all, just observe. Ask them to
provide a spoken comment ("stream of consciousness"), if they
can, but concentrate on their actions, not their words. Have a
cam and mike handy, if you can. The whole thing is extremely
worthwhile, and changes your perspective completely.

> What is really at issue here, and perhaps the entirety of this
> group's mission, is if a user is finding what he/she is
> looking for where they might reasonably expect it.  The
> failure of Control Center in the case you described was that
> there wasn't a clear path to complete what the user wanted to
> do.

Yes, and no. The control center was extremely unorganized, that
was one problem. Part of the problem was, that the controls
address the wrong level of abstraction, though.

I believe my father would have had it easier if there were
controls like "make all fonts larger", "make the system suitable
for <profile group>/behave like <profile name>" - like themes,
only capturing all aspects of the system. 

On the other hand, having controls that control the slide speed
of the kicker, depending on automatic or manual hiding, is
making things actually HARDER for him. Hell, even I find these
sliders to be useless and confusing unless I need them and
specifically search for them, and I have studied CS! There is a
reason for Microsoft Power Toys and the fact that they are NOT
part of the default system distribution.

> Then the user tried to get a handle on what icons are doing. 
> Again, this isn't a case of too many options, but the fact
> that they're spread out into areas based on the underlying
> program, rather than the single concept of "Icons".

Agreed on this issue. Grouping by task or concept, not by
program, would help greatly.

Kristian

-- 
Kristian Köhntopp, NetUSE AG, Dr.-Hell-Straße, D-24107 Kiel
Tel: +49 431 386 435 00, Fax: +49 431 386 435 99
_______________________________________________
kde-usability mailing list
kde-usability@mail.kde.org
http://mail.kde.org/mailman/listinfo/kde-usability
[prev in list] [next in list] [prev in thread] [next in thread]