When I taught stats, I would go over these ideas and emphasize
their relevance to what numbers mean. However, since all the usual
examples are far removed from psychology, I stuck to assessing whether our
numbers could ever have meaningful zero points, whether and when intervals
for popular scales could have equal units, and what having "more"
self-esteem could mean. Is the coverage of the scales of use for anything
more than leading in to the problems of measurement, and debates about
stat assumptions? Not that these aren't important issues--they are
central in my coverage. It is just that the text coverage of such things
seldom gets into these measurement issues, usually confuses students with
examples that are not tied to psych work, and I wonder if they serve a
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
texts consider frequency counts as ratio---discrete measurement but with a
true zero? How would you clarify that to a class that is wondering why
scales matter, since they only get a page in the text and then are ignored
for the next 3-4 years? Gary Peterson
