Peter Jones:
> This brings to mind an issue with review systems - often you need to
> know _why_ someone gave something a certain score. It often isn't
> enough to know that the person reviewing is your friend/colleague... (01)
My strong view of the rating issue is that most ratings without comment are of
nearly no value. (02)
There are places for purely numerical ratings, but when discussing factual
information, it is almost never enough. (03)
One example of where numerical ratings work is that I see a need within our
own use for ratings of style/grammar/spelling for which we would likely have
a set of guidelines. The guidelines would specify a set of 'objective
criteria' by which the ratings would be issued. Comments in this type of
case would be only necessary if there were something outside the guidelines. (04)
so, yes, numeric or other scale-based scores for things are only of limited
value to the long term improvement of the dataset in question. But the more
aligned raters are in the criteria used, the more meaningful / useful a
single linear scaled score can be. (05)
blinc (06)
--
This message is archived at: (07)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=tools-yak&i=200307261258.44213.blincoln@grassrootreview.org (08)
|