Re: Usability testing

Subject: Re: Usability testing
From: "Tim Altom" <taltom -at- simplywritten -dot- com>
To: "Tom Johnson" <johnsont -at- starcutter -dot- com>, "TechDoc List" <techwr-l -at- lists -dot- raycomm -dot- com>
Date: Mon, 3 Jul 2000 15:11:34 -0500

Usability testing isn't designed to perfect a document, although that's the
unattainable goal we all share. Rather, it's to detect obvious problems in
layout, word choice, or organization. Accept variation as a fact; some user
or another will always have problems. The job of usability testing is to
find the outliers, those problems that cause a number of users to have
caniptions. What you can do is collect data, then cast an x-bar chart or
something similar to find average and variation. If you can get your
documentation into the "control" range consistently, consider your job done.
If testing shows that comprehension is "out of control", then you have work
to do.

The calculations are pretty straightforward, and are right out of any decent
book on statistical process control. For example, let's say that for a dozen
test subjects (more than enough, actually), you time how long it takes each
to find the text telling the subject how to do perform a given task. If the
time per varies outside of three standard deviations plus and minus, things
are "out of control" and need to be looked at. If, however, the times all
together fall within + or - 3, then you can be pretty confident that users
will generally find what they need quickly. The same measurement can be
applied to the number of mistakes made, the number of false pages located,
or number of bad choices made from the TOC.

Tim Altom
Simply Written, Inc.
Featuring FrameMaker and the Clustar(TM) System
"Better communication is a service to mankind."
Check our Web site for the upcoming Clustar class info

> So what does this say about usability testing? Some people will find
> something wrong even if hundreds of people use it with no problem. Can you
> come up with a document that will make everyone happy? Maybe, maybe not.
> Can an employee single-handedly make a document (or a company) look
> foolish? Most definitely.

Previous by Author: Re: New TECHWR-L Poll Question
Next by Author: Re: Usability Testing--Interim (I hope) summary
Previous by Thread: Usability testing
Next by Thread: Re: Usability testing

What this post helpful? Share it with friends and colleagues:

Sponsored Ads