DON'T HAVE THEM RANK THINGS ... use some sort of an "absolute" rating scale
like
here are three ways to manage your intranet ... (these might not make any
sense)
never would use ................................ would always use
1 2 3 4 5
(or some variant of this)
1. outsource this to a third party provider
2. have separate servers for each account
3. dedicate ONE person to handle all the intranet work
------
here, you COULD if you wanted to after the fact ... rank things in order
but ... the main problem if you do this from the git go is that ... they
might rank the things 1,2, and 3 ... and like them all ... or hate them all
... at least with RATINGS ... you get some feel for DEGREE of favor or disfavor
At 04:27 PM 5/13/02 +0100, [EMAIL PROTECTED] wrote:
>Hello
>
>I was advised to contact you by Kimberly Rex who works for Dr. Knodt.
>
> Unfortunately I am a bit stuck due to my lack of up to date knowledge
> of statisitcs (last studied statistics
>when I did my forestry degree some 18 years ago!!). I have been trying to
>get advice on an appropriate
>method of statistical analysis for my proposed MSc research and the
>statistics department in our university
>(Robert Gordon University, Aberdeen, Scotland) is very small and the staff
>are not aware of some of the
>techniques I am hoping to use.
>
>I am proposing to survey intranet managers and ask them to rank the issues
>that impact most on the way
>they manage their intranet. I want to test to see if a certain type of
>intranet (e.g. local government intranet
>set up 3 years ago) has a particular pattern of ranking issues compared to
>another type of intranet.
>
>I plan to put the issues into appropriate categories so the ranking could
>be based on the categories (say 3
>categories). The allocation of ranks would use pair-wise comparison and
>geometric means applied to
>produce normalised ranking (similar to the Analytical Hierarchical Process).
>
>I was wondering if you have any suggestions as to appropriate statistical
>analysis for comparing the ranking
>against intranet characteristics. The test which springs to mind is the
>Spearman's Correlation or cluster
>analysis. Could this be applied in this situation?
>
>Cheers
>
>TIA
>
>Nick Ananin
>
>
>.
>.
>=================================================================
>Instructions for joining and leaving this list, remarks about the
>problem of INAPPROPRIATE MESSAGES, and archives are available at:
>. http://jse.stat.ncsu.edu/ .
>=================================================================
Dennis Roberts, 208 Cedar Bldg., University Park PA 16802
<Emailto: [EMAIL PROTECTED]>
WWW: http://roberts.ed.psu.edu/users/droberts/drober~1.htm
AC 8148632401
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
