Date:         Sat, 12 Sep 1998 21:57:25 GMT
Reply-To:     Mike Hewitt <mmhewitt@FLASH.NET>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Mike Hewitt <mmhewitt@FLASH.NET>
Organization: Flashnet Communications, http://www.flash.net
Subject:      Re: z-scores

Thanks for your input.  My intention was exactly that, to calculate
z-scores for each state and then put the info back together.
 
Michael Mason <M.Mason@MCD.UNIMELB.EDU.AU> wrote in article
<199809121724.DAA28752@ariel.ucs.unimelb.EDU.AU>...
> From: Mike Mason
> Date: 09/13/98
>
> Tim and Mike: Surely transforming the scores into Z-scores within each
state
> (which is what I understand Mike to mean by 'z-scores for each state')
will
> be quite valid; since all the scores are on the same scale, it won't
really
> have any other effect than to set the mean to zero and the standard
> deviation to 1.  The differences between states will remain.  If one
> transformed them all into z-scores across the whole range of states, then
> the differences would be obliterated as Tim says.  What perplexes me is
> that Mike finds the existence of the differences unexpected.  Although
the
> states all score out of 100, is it possible that the same judges function
> across a whole state, and/or have different interpretations of the
scoring
> standards as a group, compared with other groups of state judges? (Pardon
> me, I'm in Australia, and don't have much idea about how marching bands
are
> judged.)  Alternatively, as Tim suggests, even if the judging is all on
the
> same objective standard, some states where marching is a bigger thing and
> standards are high, could well have a legitimately higher state mean.
 
>