Dear Wendi,
I use Richard Gregorys book. It is
comprehensive and well-written. I supplement it with readings from journals and
include various student exercises, including writing a critique of a test and
participation in a class research project in which we gather data on norms,
As long as you discuss validity you should be in good shape.The problem with
texts and measuremeints
is that they spend too much on reliability and much on validity-constructive
and concurrent.Something can be reliable
but not valid;but if something is valid,it will be reliable.
Michael
Whether reliability or validity is more important depends on the purpose of
using a measurement. For applied settings, where the measure may be used to
make decisions affecting a person's life, validity is very important. In basic
research, just measuring something reliably can be most
in the Psychological Sciences
Subject: RE: tests and measurements
Whether reliability or validity is more important depends on the purpose of
using a measurement. For applied settings, where the measure may be used to
make decisions affecting a person's life, validity is very important. In
basic
Its been a while since I've taught it, but Cohen et al. is a good text. It
blends some interesting history in with the dry psychometrics.
A good cross-validation/regression to the mean exercise can be obtained by
assigning students test scores based on some random process (e.g., rolling a