topic Re: Running Multivariate analysis in Discussions

Running Multivariate analysis

E_Dandekar — Sat, 07 Jul 2018 00:02:57 GMT

Hi,

I'm trying to run a few analysis but can't seem to figure my way around JMP to set-up and produce any kind of results. I'm a graduate student in California.

What I know I need to run is:

1) MANOVA - where I am trying to find differences in 4 variables between 4 groups.

2) Multivariate t-test: I have 2 variables, that are related, and find differences between two groups over time.

A little background: I'm observing body composition changes in athletes. for the MANOVA I am trying to see differences between different tissues (the 4 variables, i.e. fat, muscle, bone) between four sports(this is only at one time-point). The multivariate t-test was suggested to look at two types of fat tissue(2 variables) in the body between two sports over a season (beginning and after).

After searching and reading, the MANOVA set up I try to implement in JMP doesn't seem satisfactory, so I run a "Fit Y by X" running 4 one-way ANOVAs. And sources point that I should try to use a 'hotelling T2 test,' which I have no clue how to do on JMP nor have I yet to find any help on this topic or how to set it up.

Any help would be greatly appreciated and if further details are needed I can provide!

Re: Running Multivariate analysis

Mark_Bailey — Mon, 09 Jul 2018 12:55:28 GMT

This request sounds like a common question about the analysis of a repeated measures design. Please see this technical note for more information.

Re: Running Multivariate analysis

E_Dandekar — Sun, 15 Jul 2018 02:13:12 GMT

@Mark_Bailey Thanks for the reply and suggestion. But to clarify, the variables I am testing are of only one-time point for our four sports. We are looking at body composition, multiple variables (fat mass, bone mass, muscle mass, etc) but not repeated (I wish it were that easy). Unfortunately, I needed this information expedited and in the end, decided to use another program that gave me clearer results/options.

Re: Running Multivariate analysis

Mark_Bailey — Mon, 16 Jul 2018 12:07:59 GMT

Perhaps you could post your solution and the results so that we could all learn from your success, and possibly expedite a JMP solution in the future.