[Sigia-l] Edward Tufte

[George Olsen]

Karl Fast said:
> However, his work is largely independent of interaction. The idea
> that people will interact with the information systems we architect is
> THE key element that separates our work from his.

That's indeed a critical difference. I'm sure there may be some quibbling
over defining "interacting" -- since people do engage with printed
materials. But it's a different kind of engagement. Back in the days of
"new media" we used to refer to this is "lean back vs. lean forward."

But I think his comments are really more reflective of his academic roots.

Practicing graphic designers normally test their designs as they develop
them -- in part because it's so expensive to screw up when once it's on
the printing press. For every final idea that sees print, there's often
two other ideas that formally presented to the client but discarded --
typical practice to present three potential design directions -- and
numerous other ideas that were explored and discarded even earlier.

This filtering process isn't the sort of "testing" that usability experts
would necessarily recognize, but it does involve bouncing ideas of people,
having peers critique your work, etc. It's not the same as talking to real
users, but skilled designers are paid in part for their ability to put
themselves in the user's mindset.

That said, skill and experience can substitute for testing -- in large
part because they are internalizations of the lessons typically learned
through usability testing.

For example, typographic theory pretty accurately predicts the legibility
of typefaces when typefaces have been tested in the lab.

It's one reason (in graphic design and other fields) that person and
companies that specialize in verticals (i.e. a particular industry) tend
to get paid more than horizontal generalists. Because they know that
particular field intimately, they're more apt to design things right the
first time.

The question of course is how do you know that got a good design. It's not
an either/or question, but rather one of probablilities. Skilled designers
are more likely to get it right the first time, but there's a risk that
they won't, which is why iterative prototyping and feedback (whether or
not that's formal "testing") are good insurance policies.

Of course that points to the flip side of testing -- research. Needless to
say having a better understanding of the issues and users upfront,
improves the odds of better insights during the design process, which in
turn improves the odds of getting it right the first time.