Folks,

A week or two ago there was a question about a multivariate text. It was
rightly said that the APA paperback, while hardly an appropriate text, was
a valuable resource if one were teaching such a course.  Well, here's
another.

It is old, it is out of date in many ways, but if one really wants to
understand multivariate data analysis, there is nothing better than 
Cooley & Lohnes, 1971. I never used it as a text because its use of matrix
algebra would have scared my undergrads to death. Shouldn't have been a
problem since C & L do an excellent job of teaching matrix algebra on
their way to teaching multivariate data analysis. They also include
Fortran and programs in Fortran for doing all the neat matrix
manipulations on multivariate data sets. That ought to tell you how dated
it is.

My point is that for someone who really wants to get a thorough 
understanding of multivariate data analysis -- Cooley & Lohnes eschew being
statisticians -- this is a marvelous book.  I sometimes today struggle
with the issue of providing software that does the number crunching and
leaves the conceptual learning which came so naturally when the students
had to use, at best, a Monroe IQ 213! (or whatever it was) to chance.
Today they get answers, but they generally don't have the foggiest
what they mean. Guess I'm being an antiquarian again.

Al

Al L. Cone, Ph.D.
Professor & Chair
Department of Psychology                701.252.3467  X 2604
Jamestown College
6019 College Lane                       [EMAIL PROTECTED]
Jamestown, ND 58405                     

Reply via email to