Teaching multivariate statistics: text & software recommendations

#1
Hello all,

I'll be leading an introductory seminar to multivariate statistics in the spring and was looking for recommendations for texts that focus on applications of multivariate statistics, without delving too much into the math behind everything. The students are in a variety of psychology master's programs, but most will be counseling or school psych, and will have, by all accounts, weak math and relatively introductory statistical backgrounds (ANOVA, simple linear regression, etc.).

I'm looking for recommendations for a text with an applied focus (preferably for psychology students) that walks them through some basic procedures for using statistical software to check assumptions and screen data and to perform some or all of the following statistical tests:

- factorial ANOVA
- repeated measures analysis
- ANCOVA
- MANOVA
- MANCOVA
- multiple regression
- factor analysis
- reliability
- discriminant analysis
- logistic regression

I'm debating whether to have the students use either

a) the university-licensed SPSS, which I could demonstrate in class, but which they'd have to work with on their own in a university computer lab or

b) Minitab, which the students would have to obtain licenses for (about $30 to rent or $100 to buy), but which they'd be able to use on their own laptops during class and at home.

Any preferences on SPSS or Minitab for teaching multivariate stats? Any recommendations for a text to help students with the software (and the topic generally)?

Thank you for your time.

Regards,
Jaso
 
#2
Using Multivariate Statistics 5th. by Tabachnick and Fidell. Covers all those topics, with examples in SPSS, SAS and others. The book is oriented towards a learn by example approach. Little emphasis on theory. Should be perfect.
 
#3
I've taken a required stats course within an arts program (very basic) and we used the free program "R"

It was easy to use. All students (with computers) were able to download the program from the internet.

Not sure how it stands against the programs you mention, but it is really accessible. Check it out.
 
#4
From my experience, SPSS and Minitab aren't too terribly different in terms of how you do things. While it is nice for them to do it on their laptops, I would stick to SPSS, just because it appears to be much more the standard. Besides, some of the best times I've had with statistics involved being stuck in the basement of the psychology building working on stats homework because those computers were the only ones that had the software.