Allow me a brief analogy. Imagine if you were driving your car, and
someone pointed out that if you shifted out of first gear, you could
travel much faster than 10 MPH. Sure, before you knew this, you were
driving, and "achieving your goals" (e.g. getting to work), but you
were hardly using the tool (your car) at its most effective level.
Let's forget about power users and keyboard shortcuts - I'm not
interested in GOMS or Fitt's Law level analysis.
My real concern is that software systems often have features that
people don't use, or don't use properly. I'll give a practical example:
an experienced colleague, whose work I often edit, can't properly
outline in Microsoft Word. This creates frustrations for this person
and everyone who works with this person. This is neither a trivial
issue (I've spent about 1 hour fixing a document today alone), nor is
this an inexperienced or unsophisticated user.
I probably can't be convinced this isn't a problem, but I can agree
that this isn't any easy problem to resolve. In fact, I'm not convinced
that it can be solved, but if not, it's likely that we'll always have
tiers of users; the "computer sophisticates" who are starting to take
over the lower ranks of the professional world will see their skills
degrade (or at least stay static relative to a moving target) as the
next generation of more sophisticated users comes of age, and the cycle
On Jul 20, 2006, at 6:33 PM, Peter Bagnall wrote:
> Why bother teaching users more stuff? > > If they are achieving their goals, great, the design works. Why do we > feel a need to force them to become power users when their current > level of knowledge is meeting their needs? I'd argue maybe we don't. > > Users will take on as much complexity as they are comfortable with. > I'm all for providing alternative command vectors for power