At 06:58 PM 8/24/03 -0400, Donald Burrill wrote:
Hi, Dennis.
 You did not address the question, so far as I can see.  You stated an
answer, which I take to be your personal opinion on the point, but you
supplied no supporting arguments.
 I had asked "are those properly called 'z' scores?", because z scores
are commonly defined in terms of population values (aka parameters),
while the values I was questioning were calculated from sample values.
 Possibly a minor and not-very-interesting point, but I'm unwilling to
take an unsupported assertion as a reasoned reply ;-).

I guess I don't see the quibble point. By definition ... a z score ... which has mean = 0 and sd = 1 ... is a scale that tells how many sd units a score is from the mean


Now, if you had population data ... then the z would refer to THE position of a score around the mu value ... however, if you have sample data and are interested in what the z MIGHT be in that population for the given X value ... then I would assume that the z is an estimate of what THAT score would have as a z ...

I picked up 3 random books on my shelf and they said:

Ferguson (old book) ... A standard score is a deviation from the mean divided by the standard deviation .... (note: z is just one example of that)

Glass and Stanley ... show z = (X - X BAR) / S(x) ... and we know that their terminology is to use S as an estimate of sigma

Moore and McCabe ... in their section on a ND( with mu and sigma) say that if you subtract mu from X and divide by sigma ... this is "often" called a z score ... I see nothing in their discussion (there or later) that restricts z to a population case ... having mu and sigma

I have never seen such a distinction that you are suggesting ... doesn't mean you are incorrect but ... books seem NOT to distinguish ...

Do you have any explicit example from a text ... that DOES make this distinction?


On Sun, 3 Aug 2003, Dennis Roberts wrote in part:

> At 02:27 AM 8/3/03 -0400, Donald Burrill wrote:
> >
> >Interesting semantic question, though:  Are those properly called "z"
> >scores?
>
> yes they are

 -----------------------------------------------------------------------
 Donald F. Burrill                                         [EMAIL PROTECTED]
 56 Sebbins Pond Drive, Bedford, NH 03110                 (603) 626-0816

---------------------------------------------------------- Dennis Roberts Email: [EMAIL PROTECTED] Web: http://roberts.ed.psu.edu/users/droberts/drober~1.htm

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to