Re: [Usability] Programmer needed for minimal and basic design for GNOME HIG experiment

[Alan Horkan <horkana maths tcd ie>]

On Wed, 8 Mar 2006, Matthew Paul Thomas wrote:

> Date: Wed, 8 Mar 2006 02:47:31 +1300
> From: Matthew Paul Thomas <mpt myrealbox com>
> To: William Szilveszter <wszilveszter gmail com>
> Cc: GNOME Usability Mailing List <usability gnome org>
> Subject: Re: [Usability] Programmer needed for minimal and basic design
>     for GNOME HIG experiment
>
> On Mar 5, 2006, at 4:03 PM, William Szilveszter wrote:
> > ...
> > I am currently in the process of designing a true experiment for my
> > undergraduate studies in psychology. I am interested in providing
> > experimental research in the field of ergonomics and computer
> > usability.
>
> Excellent!
>
> > Currently my aim is to show the effectiveness of section 4) Menus. The
> > experiment will focus on menu icons and whether they truly interfere
> > with reaction times. My aim is very simple, however I will be
> > controlling for a number of variables and if successful, anticipate
> > some empirical research that will provide a real base to the GNOME
> > HIG.
> >
> > My experiment however does require a simple program that will run on
> > Win32 and be a self contained .EXE (program would be running on
> > Windows 2000, most likely SP4). The program will require nothing more
> > than 4 to 5 menu entries (content is moot and entries such as file,
> > open, help, etc. are more than worthy). The menu entries will need to
> > be a little more organized and can be based off a simple mock up that
> > is yet to be made. The program needs to simply provide a function
> > where the click of the drop down menu commences a clock and the
> > clicking of the proper menu entry stops it, recording it to a file in
> > order of attempts.
>
> If the timing is going to be done by the program itself, why not add
> the code to a real Gnome program such as Evolution or Gnumeric?

He sent a mail offlist (not intended for me) explaining his test was going
to be quite small and limited in an attempt to analyse only the factors he
was really interested in and to try and exclude other factors.

> would be very likely to cause sample error, either by having a higher

the test is likely to only have a very small sample group.
the results will be interesting but hardly conclusive.

> proportion of illustrated menu items than Gnome will ever have, or by
> choosing menu items with more distinct icons than those in real Gnome
> programs can have.

from the sounds of things he intends to use a few stock icons for his
tests.

- Alan