Conduct MANCOVA in JMP

[#1]

Hi everyone,
I would like to conduct a MANCOVA model, where:
DV: (Posttest scores) and (completion time for posttest).
IV: (Groups), (Gender), (Degree), and (Major).
CV: (Pretest scores) and (completion time for pretest).

Shall I conduct MANCOVA the same way we conduct ANCOVA, where the DVs are plugged in the role variables box , and IV + CV in the effects box?


Thank you

[#2]

Welcome to the community.  Sorry, I can't give specific advice as I don't understand your question or your situation.  Here are my thoughts:

1. You are interested in understanding covariates?  This is usually in association with a designed experiment. I'm guessing these are the Pretest variables?

2. It appears you have 2 response variables: Posttest scores and Completion time

3. It appears you have 6 potential factors you think might be associated with those responses:(Groups), (Gender), (Degree), (Major), (Pretest scores) and (completion time for pretest).  Did you explicitly manipulate The first 4?  How? What is a group?

4. Did you run an experiment or do you have a set of data collected with all of these variables identified? 

Any chance you can post your JMP data table (anonymized is OK)

[#3]

Thank you for your reply. 

The experiment design consists of three phases; (1) pre-test (2) learning, (3) post-test. As for the groups, it consists of only two levels; (1) control, and (2) experimental. We first used ANCOVA to compare between the groups using the post-test scores, while controlling for the covariates (pret-test scores), and using (gender, major, and degree) as additional independent variables.

Now, we are interested in comparing between the groups using the post-test scores and post-test completion times, while controlling for the covariates (pret-test scores, pre-test completion times, and training time) and using (gender, major, and degree) as additional independent variables. We want to conduct MANCOVA to accommodate for two dependent variables, three covariates, and four independent variables. When conducting MANCOVA on Fit Model window (the same way we do ANCOVA), it generates to separate least square fits ( one for posttest scores, and one for posttest completion time).