Development of a scale, i.e., converting non-numeric attitudes (and other
non-numeric 'stuff') into a number scale, is no easy matter. So you
demonstrated. Some people will treat a 2 point scale as a dichotomy, skipping
gradations in between. Some people will treat a 5 point scale (true, original
Likert, for ex.) as an integer scale, as if there was no gradation of attitudes
finer than what is shown. Perhaps these folks have been well trained to 'follow
instructions' and only mark what are clearly available for marking.
Degree of personal motivation and individual self-understanding might have a lot
to do with the precision of a response, too.
In HS standardized tests I have received reliable reports that some students will
purposely answer B for all of the questions (out of A, B, C, D multiple guess
questions). Even when the tests are 'high stakes,' some students will figure
there is no way they can do well, so they don't spend the emotional energy to try.
Thus, they don't get reminded of their own weaknesses. Plus, it ticks off the
adults, which is a benefit in itself(!)
In the work I reported earlier, I had the advantage that I could 'norm' the scale
with samples. This is a major advantage, which someone wishing to use Likert or
similar scales would do well to explore. In one case where pairs of people
evaluated casting surface finish, with no physical standards, they found that the
sales people, quality assurance people, and production engineers each had a
different average value. Within a group there was good agreement; between groups
there was significant disagreement.
Jay
Rhythmpsyc wrote:
> I tried to use a visual scale in a student undergrad assignment (so they would
> have more continuous-like numbers). It was hopeless. Some of the respondants
> circled anchors, no matter how explicit I tried to make the instructions (I see
> that you (Michelle) left the intermediate anchors off, though, which I didn't).
> In any case, do you tend to get clustering in your distance distributions,
> e.g. at midpoint, quartiles?
--
Jay Warner
Principal Scientist
Warner Consulting, Inc.
4444 North Green Bay Road
Racine, WI 53404-1216
USA
Ph: (262) 634-9100
FAX: (262) 681-1133
email: [EMAIL PROTECTED]
web: http://www.a2q.com
The A2Q Method (tm) -- What do you want to improve today?
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
