Can you eliminate testing?

[Tim Altom <taltom -at- IQUEST -dot- NET>]

The threads on writing tests intrigue me. At Simply Written we don't give
tests, for two reasons. First, we look at samples, even dummy samples, and
closely quiz the writer about all the text. Why did you phrase this so? Why
this organizational pattern? Who wrote what? I want to know the mental
processes, not the fact of writing per se.

Second, we've devised methods of writing that we train people to use,
reducing the need to have people who can invent and reinvent. Our Clustar
Method has proven to be a marvelous boon for us, because it gives us highly
useable consistency. Writers still have to solve problems, of course, but
the problems are more limited and the answers more reachable in less time.
That's let us develop newer writers faster too. It's easier to teach our
method than it is to find that one perfect gem of a writer.

I have to wonder if writing tests aren't meant to either scare off the
utterly raw recruit, to, as one commentator put it, avoid legal problems, or
to fill in the hiring person's ignorance gap. A truly experienced techdoc'er
ought to be able to tell pretty quickly if someone can write without having
them spend 5 hours in a test. In my experience, like knows like.


Tim Altom
Simply Written, Inc.
317.899.5882
www.simplywritten.com
Creators of the Clustar Method for task-based documentation