Re: Rating Systems
- To: i*@rt66.com
- Subject: Re: Rating Systems
- From: t*@lanl.gov (Tom Tadfor Little)
- Date: Tue, 24 Sep 1996 09:04:07 -0600
John writes
:Rating systems based on point scoring are subject to periodic "grade"
:inflation. A better system would be to require evaluators to assign all
:irises reviewed (including those that died?) as follows:
:
:Excellent (top 10%)
:Good (top 25%)
:Average (middle 50%)
:Below Average (bottom 50%)
:
:**And** require that the ballot cannot allow 10% of the entries evaluated to
:be placed in the top category, 25% of the entries to be in the top two
:categories, etc.
Yes, that's an excellent suggestion. There needs to be some way to "tie
down" the ratings so they don't wander off. The downside is that a
relative system doesn't allow for the fact that a judge may actually
see more excellent irises in one year than another. I prefer absolute
ratings myself, but it is very hard to get everyone to apply them the
same way.
===============================================================
Tom Tadfor Little tlittle@lanl.gov -or- telp@Rt66.com
technical writer/editor Los Alamos National Laboratory
---------------------------------------------------------------
Telperion Productions http://www.rt66.com/~telp/
===============================================================
