MANOVA - 3 Effect variables, 3 dependant variables

[#1]

Hi!

not sure if i am in the correct spot. 

I am trying to go thru in JMP and create a Univariate tests (Between-subject Effects) . i have 3 effect variables, names X1, X2 and X1 x X2 and three dependant variables X19, X20 and X21. 

here is a screen shot i am trying to duplicate in JMP.

However, in making JMP and reading thru Multivariate-Methods_v15.pdf, i cannot seem to locate the procedure syntax for creating the above representation.

Any advise or links would be useful.

thank you

Nick

[#2]

I assume that you X1, X2, X19, X20, and X21 stored in separate data columns with those given names and correct data type and modeling type.

1. Select Analyze > Fit Model
2. Select Personality > Manova
3. Select X19, X20, and X21 > click Y
4. Select X1 and X2, click Macros and select Full Factorial
5. Click Run
6. Select Univariate Tests Also
7. Click Choose Response and select the appropriate kind

The documentation you cited provides a description of each kind of response to help you decide.

By "syntax," I assume you are referring to the JMP Scripting Language. A script is not necessary. If a script is desired, perform the interactive steps given above, click the red triangle and select Save Script > To Script Window.

[#3]

Mark,

i did that but i had to use the identify. The  repeated measures did not provide me the multivariate answers i needed. And the univariate did not function.

I used the least squares to determine the effect tests for the above. 

The problem is that JMP has not documentation on how to use MANOVA. Unless you can show me the procedures that function for the various parts of the MANOVA.

[#4]

Hi Nick,

I'm not sure if the video in this link will add clarity to your original questions but I wanted to share it in case it does: https://www.youtube.com/watch?v=vGT3NEYrOg8 

Also, I've passed your feedback to our documentation team so it can be addressed. Please let us know how else we can help.

Best regards,

~Laura

[#5]

One more thing to clarify, the MANOVA documentation is under the "Multivariate Response Models" chapter. Here's the link to the MANOVA online documentation that describes the fitting options, output, and provides examples:

https://www.jmp.com/support/help/en/16.0/#page/jmp/multivariate-response-models.shtml# 

The equivalent PDF version is here:

https://www.jmp.com/content/dam/jmp/documents/en/support/jmp16/fitting-linear-models.pdf

HTH,

~Laura

[#6]

Laura,

thank you for the links!

Nick 

[#7]

laura,

That helps. I found it a while ago. Performing MANOVA in JMP can be a bit testy. You need to provide understanding of your groups. Further explanation in the MANOVA 15th would be helpful for the choose response for repeated measures, identity ,etc, would be useful for new JMP folks. As for univariate you need to create an effect output in  fit group. there are several things i need to be able to do in MANOVA for univariate with groups and MANOVA. but these are separate exercises to do in JMP using MANOVA, Multivariate or  least squares. But more disussion on how to perform MANOVA for both would be added value to JMP users.