How to run a 2 factor mixed design in JMP?

Question

Hi everyone,

I must be overlooking something obvious because I'm having trouble doing a very simple analysis in JMP: I'm trying to run a very simple 2 factor mixed design: 2 conditions with 1 repeated measure in them (so 1 between subjects factor and 1 within subjects factor).

Can anyone tell me how to run this in JMP?

Thanks so much for your help!

Cheers,
-Els

unknown · Answer

Hi Els.  I think it was clear, just that I rarely analyse this structure of data. So at the risk of sending you in the wrong direction again ... in the Fit Model dialog you can change the Personality of the analysis to MANOVA.  Also under the Analyze menu is an option for MATCHED PAIRS.

unknown · Answer

Whoops, sorry if I wasn't clear about this, but I want to run a regular repeated measures ANOVA with 2 groups, so 1 between subjects factor (condition) and within subjects factor (the specific test they performed twice). So there are no random effects in this model....

Any idea how I should go about doing this in JMP?

Thanks again for your help!

-Els

unknown · Answer

In the fit model dialog window, there is a red-triangle context menu called "Attributes". You will see listed in there "Random Effect". Once you have added terms to the model you can use these to mark them as random effects. Also in the dialog window are options for personality and emphasis. Once you add a random effect, you will have a 3rd category "method". This will default to REML.

If you look in the sample data under the help, there is a category for mixed models. These tables have scripts attached to them so you can run them to see what the analysis looks like.

Hope this helps.