[Sigia-l] The planetary classification conundrum... mentalmodelsoutside of the Web

[Gent, Andrew]

[Fred said] 
That's exactly my question... at what point do we sacrifice logic to an
existing mental model, aka "common belief?" 

Thanks Fred. Excellent question. In fact, I believe from an IA
perspective, which is what you were using this as an analogy for, there
are two questions: 1.) "when" do you choose one over the other and 2.)
"how" do you. In other words, how do you most effectively counter a
familiar mental model with an improved reality?

I'm happy to stick my neck out and take a stab at answering the first
question. You argue for logic when there is a significant, measurable
improvement that will result from the change. There are plenty of
examples of this principle in practice within IA and related
disciplines, usually arguing to stick with existing mental models
because there is little benefit to contradicting them. (e.g. the
inverted "L", seven plus or minus two, etc.) 

However, once you get beyond the obvious, rules become less definitive.
Take color for example. When did we (designers) move comfortably away
from the stricture of maintaining blue as the link color? Some did it to
be edgy and ahead of the curve. Many did it once enough browsers had
enhanced graphics support to be able to control text fonts and
background colors and images reliably so we *knew* what impact the
change would have and could effectively control the user interaction
experience.

The second question, "how", stretches the boundaries of IA to some
extent, but it is something I find myself spending a significant amount
of my time on. (Which is why I raised it.) Having decided a better
logical model will bring improvements, how do you convince the
stakeholders and design the rollout to minimize negative feedback?

Years ago I was doing a redesign of a web site where the home page had a
large graphic of a castle and four links (text on the image). Lots of
people were confused by what was behind the links (lack of clarity,
confused navigation taxonomy). However, an equal number of people liked
the web site because (we suspect) they liked the graphic.

We knew we had a better design. We also knew we would get significant
pushback from the advocates of the current web site (and its designer).
How do you make the change and not get into a subjective mudslinging
contest in front of the sponsor? Our approach goes back to the answer to
question #1. Demonstrate the improvement. The improvement in this case
was largely based on accessibility and we had the usability test results
to prove it. With that in hand, we had meetings with the sponsors and
key user groups to review the design -- and the test results -- before
rollout. The other thing we did was include a narrow slice of the
original graphic in the banner to retain a sense of continuity between
the old and the new site. 

That example is an easy one. For many of my other projects the benefits,
although recognizable from an IA perspective, are harder to demonstrate.
To give another example which wasn't so easy -- years ago I was leading
a project to document a new software "product" created by putting 14
existing products together under a single SKU with a common installation
process. The information strategy was to write and print one 100 page
book describing the product components and provide the other
documentation (doc sets for all 14 products) on CDROMs. (The internet
wasn't in full bloom yet.) For several months I collected data and
published memos explaining the benefits in terms of cost to both the
company and customer (which were significant!). But the product team was
adamant that "our users want printed documentation". Finally, in
desparation we had the team order one copy of each of the 14 doc sets.
The turning point was when the documentation arrived -- on two wood
pallets six feet high. No one argued after that. 

I apologize for rambling on. But I would like to make one final point,
coming back to the original analogy. What is the benefit from
classifiying Pluto as a planet or a "pluton"? I suspect (again I am
guessing here since it is not my field) the goal of the new
classification is not so arbitrary as it seems. Now that we have the
ability to see farther into the galaxy, we are probably finding more and
more celestial objects that could be arguably additional planets. How do
we decide? We decide by having clear criteria for classification. So the
improvement from changing the classification is not a philosophical
debate over Pluto; it is a way to avoid even more contentious arguments
over other celestial bodies in the future. (Again, there are lots of
analogies here to taxonomy changes I've been involved in recently where
the benefit is future-based and hard to demonstrate.) The question from
an IA perspective might be: "what can the scientists do in the
transition to the new classification to make it more palatable to those
of us raised in a 9-planet galaxy?"

Andrew