Re: Writing Test

["Sarah Bane" <Sarah -dot- Bane -at- ProphetLine -dot- Com>]

David Strausfeld asked for suggestions: <<We're recruiting some new
technical writers and I wanted to give a short
writing test to each applicant.  I was hoping to find a really bad piece of
technical writing on the web and ask them to edit it.   A multiple choice
test of some kind might be okay too.  Believe or not, I can't find what I'm
looking for despite spending like an hour on the web.>>

Most textbooks for use in college technical writing courses include editing
exercises, and the ancillary materials (e.g., instructor manual) usually
include tests. Some also include passages for revision. Textbook materials
have been reviewed by the publisher and thus *should* be free of the
potential problems Bill Keeley mentions regarding samples of "bad writing"
gleaned from the Web. I wish I could be more specific about what's in which
current texts from which publishers, but as an adjunct faculty member I
seldom receive review copies of anything except the text the committee has
chosen for my course. (My generalization about content is based on a
six-year former life in which I was a tenure-track university faculty member
and was a member of the textbook committee.) Publishers don't give away
books lightly, but they happily take anyone's money; you might visit the
bookstore at the nearest college and buy a copy of whatever is being used in
their TW courses.

Bill K., recovering Catbert, also states that <<A pre-employment test can be
a highly effective
selection tool provided that it is valid, reliable, and non-discriminatory.
Pre-employment tests should directly relate to the knowledge, skills, and
ability (KSAs) necessary to perform the job to the greatest extent possible.
KSAs should be determined through job/task analysis and defined in the job
description.  If a individual is expected to edit multiple-choice test items
on the job, it is reasonable to ask the person to edit multiple-choice test
items in a pre-employment test.  If the person will not be editing
multiple-choice items on the job, the case can (and probably will) be made
that the test is invalid.  Is the test reliable?  Does it reasonably predict
on-the-job performance?  Is it standardized?  Does it discriminate on the
basis of race, creed, color, gender, national origin, or physical
impairment?>>

I do understand the concerns here (and textbook materials would help with
this objection as well, because publishers carefully screen for bias and so
forth). But would an applicant who was given a multiple-choice editing test
really assume that his/her work if hired would entail editing
multiple-choice tests? Seems to me that any uncertainty of that nature could
be avoided by having the applicant initial a disclaimer statement on the
test: "In taking this test, I acknowledge that it is intended as a
demonstration of my editing skills and does not represent the actual duties
of any employment I might accept at XYZ Corporation."

However, Bill's suggestion that passages for revision be produced in-house
makes a lot of sense. Such passages could be archived drafts, for example.
It sounds like an excellent way to find out whether this person's skills
would be a good fit with your organization.

Sarah Bane
Technical Writer, ProphetLine, Inc.
and Adjunct Instructor, Westark College
sarah -dot- bane -at- prophetline -dot- com
sbane -at- systema -dot- westark -dot- edu

Opinions expressed are my own and not endorsed by ProphetLine or by Westark.