At 05:56 AM 1/5/01 +0000, James Aldren wrote:
> The paper doesn't actually mention ?s at
>all, which suggests to me that this probably isn't a very common, and
>certainly
>isn't a recommended thing to do.
i think by common convention ... the midpoint could be seen as ? or (some
middle number) or ... neutral ... or undecided ... and these are used
approximately equivalently ... technically, they clearly are not the same
but practically, i think these are seen as equal symbols
>We're now in the position of having to conduct focus groups to find out what
>people thought the ? meant. But is it not the point of a survey to find out
>what people think? We shouldn't have to have another survey to find out what
>the first survey meant. Focus groups are useful but they should focus on
>discussing why results emerged and what to do about them, not WHAT the
>results
>were.
careful survey design could have avoided this mess ... you could have had
to the side ... some OTHER option like ... "i don't know anything about
this item" ... or ... "i really don't CARE about this item" ... at least
give someone an out for one of several reasons ... but, in the typical form
... they have to use the midpoint or ? ... or, just leave it blank
do you think it is really worth the trouble to do a focus group for JUST
the above? what possible info will you get that you could not speculate on
... and, how many of these responses were there?
much of this depends on how critical this information is ... if you are
trying to get a "rough" idea of what people feel .... then the data you
have is probably good enough ... if you really need critically accurate
information ... then the scale probably should have been longer ... piloted
first and then redone before the main testing ... and some of these more
"delicate" scaling issues addressed before data were collected
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
