cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar
OldGugga
Level II

how to carry out multivariate analysis of variance

I have several hundred Ss who fall into 1 of 6 groups based on residence (city, suburb, rural) and disability (yes, no). All Ss took a personality test of about 50 items, which are scored to yield 6 traits. Items are intermixed on the test to avoid order effects, but the traits may be intercorrelated. I want to analyze the 6 traits as dependent variables in a 3 (residence location) x 2 (disability status) analysis, with follow-up tests of significant effects as needed. In some stat programs this might be called multivariate analysis of variance, but how does JMP do such an analysis? Do I use the Fit Model platform or the Multivariate Methods platform, or something else?

2 REPLIES 2
txnelson
Super User

Re: how to carry out multivariate analysis of variance

Go to 

     Help==>Statistics Index

Click on the list of statistics, and then start typing Multivariate or MANOVA, and the list of statistics will move down to where you will see multivariate analysis of variance

multi.PNG

This will give you the start you need.

You can read up on this in the Fitting Linear Model's documentation

     Help==>JMP Documentation Library

Jim
statman
Super User

Re: how to carry out multivariate analysis of variance

To look at the correlation (and scatter plots) of the dependent variables(y's):  Multivariate Methods>Multivariate add all of the y's.  Also test for outliers (test (Mahalanobis, Jackknife, et. al.) depends on the type of data the y's are). (You can also do the correlations by region or disability). You may then run fit model on the Y's of interest and add the two independent variables (residence and disability) as a Macro> factorial to degree.  This will give you analysis of variance and effects tables for the two factors and their interactions.  Make sure the two factors are type nominal when you run the fit model.

"All models are wrong, some are useful" G.E.P. Box