14.03.2014 Views

Modeling and Multivariate Methods - SAS

Modeling and Multivariate Methods - SAS

Modeling and Multivariate Methods - SAS

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Chapter 5<br />

Fitting Multiple Response Models<br />

Using the Fit Model Platform<br />

<strong>Multivariate</strong> models fit several responses (Y variables) to a set of effects. The functions across the Y variables<br />

can be tested with appropriate response designs.<br />

In addition to creating st<strong>and</strong>ard MANOVA (<strong>Multivariate</strong> Analysis of Variance) models, you can use the<br />

following techniques:<br />

• Repeated measures analysis when repeated measurements are taken on each subject <strong>and</strong> you want to<br />

analyze effects both between subjects <strong>and</strong> within subjects across the measurements. This multivariate<br />

approach is especially important when the correlation structure across the measurements is arbitrary.<br />

• Canonical correlation to find the linear combination of the X <strong>and</strong> Y variables that has the highest<br />

correlation.<br />

• Discriminant analysis to find distance formulas between points <strong>and</strong> the multivariate means of various<br />

groups so that points can be classified into the groups that they are most likely to be in. A more<br />

complete implementation of discriminant analysis is in the Discriminant platform.<br />

<strong>Multivariate</strong> fitting has a different flavor from univariate fitting. The multivariate fit begins with a<br />

rudimentary preliminary analysis that shows parameter estimates <strong>and</strong> least squares means. You can then<br />

specify a response design across the Y variables <strong>and</strong> multivariate tests are performed.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!