i think most would agree that the inclusion of something akin to ? on an
item scale creates problems ... but, the lack of having that ? does not
mean that all problems go away ... because, even in that case ... what does
a S do if they truly have not particular leaning one way or the other?
unless then leave it blank ... they will put down some response which, of
course ... is not accurate for them
just like anything else ... the saving grace is to have a reasonable number
of items AND hope that "on average" ... then tend to be fairly + or fairly
- for, if the average is close to some midpoint value (whether this is an
actually spot on the scale or not) ... the interpretation of the score can
have its greatest confusion ...
one could be quite - for about 1/2 of the items ... and quite + for about
1/2 of the items ... averaging sort of "neutral" across the entire scale
... or, one could put the middle kind of response on almost ALL items ...
thus getting a "neutral" kind of average across the entire scale .... and
we all know that these two different response patterns DON'T have the same
meaning at all ...
the overall problem is the average ... and what IT tells us and what it
does not tell us ...
At 08:46 AM 1/5/01 -0500, Bob Hayden wrote:
>An example might shed some light on one point involved here. Recently
>Plymouth State College considered the possibility of arming -- well,
>part of the question was WHOM. Many of us refered to them as Campus
>Security while they insisted they were Campus Police.
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
