Manova for repeated measures

Hi all,

I am experiencing some problems when dealing with a repeated measures MANOVA. I have 5 dependent variables (thus the MANOVA) measured at two distinct moments in time (thus the repeated mesures model). To explain my case i will recall an example dataset already present in matlab:

load repeatedmeas

I do not have any between factors so the Age IQ Group and Gender column are not interesting for me. Anyway, the analogy relies on the fact that we can consider the 8 variables here as 2 sets of 4 dependent variables, i.e. 4 dependent variables measured twice, as follows var1 = y1 and y5, var2 = y2 and y6, var3 = y3 and y7, var4 = y4 and y8. The within table seems to already express such arrangement. Anyway, let's do the fitrm model:

rm = fitrm(between,'y1-y8 ~ 1','WithinDesign',within)

and run the manova including also the within factors:

manova(rm)
manova(rm,'WithinModel','w1*w2')

Then I have compared the results, in particular the F statistic, with the SPSS output for a repeated measures MANOVA on the same dataset and the things doesn't match at all!!

Is there somethings I am doing wrong? I cannot not figure out if matlab can really handle a manova for repeated measures.

Since matlab allows to perform on the same fitrm model a ranova or a manova, maybe it interpret the data as follow:

1) In case of ranova(rm) all the columns of the between table are treated as the same dependent variable repeated over time.

2) In case of a manova(rm) all the columns of the between table are treated as different dependent variables.

If this interpretation is true, matlab cannot really perform a manova on repeated measures and moreover and more importantly to me is the following question: why manova function allows to use a within subjects factors model? What is the point? I cannot get it...

Thanks in advance for any answer,

Leo

Answer 1

[Aditya]

Hi Leonardo,

When dealing with repeated measures MANOVA in MATLAB, it's important to ensure that the model specification correctly reflects the structure of your data. The confusion often arises from how MATLAB handles repeated measures designs in the fitrm function, especially when comparing the results to other statistical software like SPSS.

• Verify Within-Subject Design: Double-check that your within table correctly specifies the repeated measures structure.
• Match Model Specifications: Ensure that the models in MATLAB and SPSS are specified in the same way.
• Check Assumptions: Ensure that assumptions such as sphericity are met or appropriately adjusted for in both MATLAB and SPSS.