Bart's "Median Rating" method?

[Steve Eppley]

Demorep wrote:
> Mr. Eppley wrote--
>> I think the definition of Bart's Median rating method needs 
>> clarification, 
-snip-
> This is deja vu all over again from late 1998.
> I suggested that if medians are to be used, then (1) only the
> choices getting above 50 medians (on a 0 to 100 scale) should 
> (2) go head to head 
-snip-

No, Bart's median method is totally different from Demorep's.
Demorep has simply grabbed the buzzword "median" and used it 
in an inappropriate way.  To Demorep, 50 is the median.  To 
Bart and to me, the median refers to the middle of the actual 
votes cast.  Demorep should have read more carefully.

Perhaps instead of "median" Demorep should use the term 
"mid-scale."


---Steve     (Steve Eppley    seppley at alumni.caltech.edu)