Hi
On Wed, 27 Jan 1999, David Bennett wrote:
> > From: Gerald L. Peterson <[EMAIL PROTECTED]>
> > Subject: Re: Examples of Interval Scale-
> > useful purpose? I think, either drop this stuff, or devote a chapter to
> > some good discussion of the problem of measurement in psych. BTW, some
>
> I disagree. I teach a course that includes both methodology and statistics.
> The measurement scale used has obvious implications in the type of analysis
> the student performs. They need to learn this.
>
> If you're teaching with a book that doesn't go into the 'problems of
> measurement in psych' you're using a bad book. That is what practically the
> entire course in Methodology is about.
The measurement issue is not at all that clear-cut. There are
respected people arguing both sides of the issue (i.e., Stevens'
classic position that scales matter and other that they do not).
Personally I just distinguish between categorical and numerical
scales and feel relatively comfortable doing standard stats on
the latter.
One reason is that it is virtually impossible to determine, I
believe, whether a psychological scale is ordinal, interval, or
ratio given the contribution of the number system, the particular
instrument being used, and the construct being measured.
A second reason is that psychological scales can be transformed
in various ways without distorting their numerical import (e.g.,
log or arcsin transformation of RTs).
A third reason is that it is not obvious what implications of
different scales are for statistical analysis. I do not think,
for example, that a statistical comparison of two different
cooling systems (e.g., mean temperature after 2 hours) would give
different results if one used absolute or ratio temperature
(Kelvin scale), an interval measure (Fahrenheit or Centigrade),
or some ordinal scale (e.g., 1=Hot, 2=Warm, 3=Cold, 4=Colder,
5=Coldest). So what sense does it make to use different
statistics?
I forget when this big debate last surfaced (it seems to reappear
in the journals every few decades), but there are articles out
there.
Best wishes
Jim
============================================================================
James M. Clark (204) 786-9313
Department of Psychology (204) 774-4134 Fax
University of Winnipeg 4L02A
Winnipeg, Manitoba R3B 2E9 [EMAIL PROTECTED]
CANADA http://www.uwinnipeg.ca/~clark
============================================================================
