On Date: Wed, 27 Jan 1999 12:30:27 -0500
"Donald H. McBurney" <[EMAIL PROTECTED]> writes:
> I hesitate to chime in here because this subject was discussed at length (great
> length) a couple of years ago on this list. But, I can't resist pointing out
> that the relevance of scales to statistics is controversial, at least. In a
> nutshell, although many psychologists follow my hero S.S. Stevens in
> restricting the use of parametric statistics to interval and ratio data,
> statisticians do not. And they should know.
The last time a related discussion surfaced on TIPS was May 08, 1998, on the
topic of "permissible statistics". I also remember lenghty discussions of examples
of measurement (probably winter 1993?), but mostly on physical entities like
lenght (long discussion whether negative lenght is possible) and technicalities
(a variable with only two values: "male"/"female", "old"/"new" is measured on
an interval scale by definition).
As Donald mentiones, there are statisticians who (wrongly, in my opinion) ignore
this measurement issue, e.g. Lord or Gaito. A very reasonable discussion
can be found in Cliff (1993) or Sarle (1995). Both papers help to reconcile the
heavy demands of measurement theory with the reality of psychological measurement.
- Literature:
Cliff, N. (1993). What is and isn't measurement? In G. Keren & C. Lewis (Eds.),
A handbook for data analysis in the behavioral sciences: Methodological issues
(Chap. 3, pp. 59-94). Hillsdale: Erlbaum.
Lord, F.M. (1953). On the statistical treatment of football numbers.
American Psychologist, 8, 750-751.
Gaito, J. (1980). Measurement scale and statistics: Resurgence of and
old misconception. Psychological Bulletin, 87, 564-567.
Gaito, (1987). ----. Bulletin of the Psychonomic Society, 25, 283-285.
Gaito & Ray (1986). ---. Bulletin of the Psychonomic Society, 24, 449-450.
Sarle, W. (1995). Measurement theory: Frequently asked questions.
(I got this article from the Internet and lost the URL, but kept the file.
Sorry, this was during my first steps into the Web. I'd be happy to provide
the file (LaTeX or postscript) for anyone interested. Warren Sarle gives permission
to reproduce his article for educational purposes.)
Rainer
------------------------------------------------------------------------------
Dr. Rainer Scheuchenpflug
email: [EMAIL PROTECTED]
Office:
-------
Institut f. Psychologie Tel.:(+49) 941/ 943-3820
Universitaet Regensburg Fax: (+49) 941/ 943-1995
93040 Regensburg
Germany
