[151] in I/T Delivery
Usability Workshop Report & Summary
daemon@ATHENA.MIT.EDU (Tim McGovern)
Wed Apr 8 11:36:21 1998
Date: Wed, 8 Apr 1998 11:37:41 -0400
To: fortoul@MIT.EDU, tjm@MIT.EDU, brendakg@MIT.EDU, mbarker@MIT.EDU,
jjv@MIT.EDU, pbh@MIT.EDU, azary@MIT.EDU, crockett@MIT.EDU,
jlittell@MIT.EDU, eyanow@MIT.EDU, lisanti@MIT.EDU, ninadm@MIT.EDU,
wade@MIT.EDU, hogue@MIT.EDU, rar@MIT.EDU, delivery@MIT.EDU,
ccount@MIT.EDU, ganderso@MIT.EDU
From: Tim McGovern <tjm@MIT.EDU>
Cc: usability@MIT.EDU
At last Friday's Discovery Team Leaders meeting, Greg Anderson asked me
to send my little report on the Usability workshop around. Here it
is...with apologies to the plain-text viewers out there, this wasn't
intended to be any more than notes for myself. Most of this reflects
my own opinions & action items, and doesn't attempt to be a
comprehensive summary of the workshop. I hope some of you find it
useful. I'd ask that if any of you have done similar reports, please
share them.
Tim
--------
I attended the Usability Testing workshop held here at MIT on March
25-27, 1998.
The instructor was Dr. Judith Ramey <<jramey@u.washington.edu>, from
the University of Washington. I found the approach taken by Dr. Ramey
to be quite provocative. While there are some hurdles to overcome in
establishing usability as bona fide work here at MIT, I expect it to be
valuable in my own work.
The first change in my thinking came as a result of Dr. Ramey's
redefinition of the whole sphere of usability. She prefers to call it
Usability <bold>Research</bold>, to make explicit that it isn't a
one-time task attached to the end of some product development cycle,
but rather an on-going effort <underline>throughout</underline> the
cycle. I came away convinced that almost anything could benefit from a
usability test, at least anything that a <underline>user</underline>
comes into contact with.
In a year, if I remember nothing else, I'd like to remember this
"<underline>poem</underline>:" ;-)
<underline>Usability is a <bold>B</bold>usiness <bold>P</bold>rocess
</underline>
Usability is a <underline>structured</underline> business process
that gathers information on <underline>specific</underline> issues
from people <underline>like</underline> the intended users
to produce product <underline>changes</underline>
within a political/cultural <underline>context</underline>
which may surface divergent <underline>value</underline> systems.
It is not adhoc, anecdotal, or comprehensive.
It is often not quantitatively measured.
<underline>Some Other Highlights from the Workshop</underline>
We heard and talked a lot about <underline>high realism</underline>.
This is the goal of usability testing procedures: to allow the
participants to feel, think and act as if they were using the product
when and where they normally would.
We talked a lot about <underline>user classification</underline>. This
turns out to be a key ingredient in good usability work, and doesn't
always map 1-to-1 with our more convenient handles for users at MIT.
For example, for many systems, classfication such as undergrads, grads,
faculty and staff may <bold><underline>not</underline> </bold>meet our
usability testing needs, even though that may be the right way to
segment the market for other purposes.
We talked a lot about the importance of users' <underline>mental
models</underline>. Dr. Ramey emphasized that fifty percent of a
product is in the user's head. There are limits therefore to how
usable a given product can be, and there are limits to how malleable
any given product can be. As part of this, she emphasized the need to
remember that there are <underline>competing intelligences</underline>
at work, and that making assumptions about how users think and feel
about products without any experimental results is dangerous at best.
We talked at great length about the <underline>steps</underline> of
usability work. We were pushed to get past fuzzy, buzzword-laden
<underline>issue</underline> statements, to the real heart of the
problem. We practiced how to <underline>unpack</underline> our issues,
translating them into clear, crip and testable problems defintions.
We talked about getting a clear idea about the <underline>object of
study</underline> in a particular situation, as it affects greatly how
to state, unpack and test the issues. Conventional wisdom holds that
as complexity increases, testability declines. But if you focus on
only those issues within the product that produce the greatest risk,
you can learn a great deal via usability research. Focus on the
<underline>most common, most critical uses</underline> of the product.
Usability work is great at identifying <underline>big
effects</underline>, not really at detecting minor, or subtle, ones.
In some of the workshops, it was very tricky to separate usability
issues, e.g., access, navigation, etc. from <underline>content
issues</underline>. And to some degree it is artificial, and not
always possible. Usability, like many disciplines, takes some
practice, and like any habit, it takes practice to acquire it, and to
use it facilely.
People can <underline>show</underline> you more than they can tell
you.
How can we imbed the <underline>learning</underline> into
<underline>use</underline>?
Use strategically!!!
The paradox of the active user.
Satisficing.
Let your problem statement dictate your procedure.
<underline>MIT Steps that Need to Happen</underline>
* We need to validate capturing and showing user experiences
* We need to validate that no matter where the problem shows up,
it's still a bug.
* We need to help MIT make a leap of faith that usability will improve
products even though it is not an exhaustive, comprehensive, or
quantitative.
* We need to support/encourage project/product managers who are
inclined to
define, schedule and allocate resources for usability work.
* We need to increase the visibility of usability work via various
forms
of communications, written, show-and-tell, personal.
* We need to start small, do a usability test: see one, do one, teach
one.
Seek small steps of gratification
* We need to form alliances with like-minded souls at MIT, and
elsewhere.
<underline>References</underline>
Jakob Nielsen's web site: http://www.useit.com/
Good web usability web site: http://usableweb.com/
Usability Professionals Assn.: http://www.upassoc.org/
Usability Listserv: listproc@hubcap.clemson.edu
subscribe utest FN LN
