How to use Mauchly's test results in matlab2014a

Hi, I wanted to conduct 2-way repeated measures ANOVA by matlab2014a.I had two with-subjects factors,Factor1,Factor2.Before I did the ANOVA,I did Mauchly's test firstly and found the output was:

But when I did the same analysis by SPSS,the output was:

So I was very confused with the two different kinds of output. I need to know:

1) What's the corresponding relationship between the data in SPSS' output and the matlab' output on Mauchly's test ?

2) How to use the input argument 'c' in mauchly function of matlab? If I wanted to get the same output of SPSS in matlab,how could I set the argument 'c' ?

3) From the output table of SPSS,I get 3 p value from Mauchly's test and know exactly which source(Factor1,Factor2,or Factor1*Factor2) don't meet the sphericity assumption and find the corresponding P value(P value of main effect or interaction effect) after adjustment in the table "Tests of Within-Subjects Effects".However,I get just one p value from Mauchly's test in matlab.How could I know which P value was adjusted in the table output by the function 'ranova'?That is to say,I need to know which data(p value of main effect or interaction effect) was adjusted when the result of Mauchly's test was significant in matlab?

I will be very appreciated for your answers. Thank your!

Answer 1

[Tom Lane]

You can ask for the C matrices from the ranova function, for example:

[tbl,a,c] = ranova(R,'within','w1*w2')

Then you can supply that to mauchly:

mauchly(R,c)

[William]

Hi , Thanks for your answer very much! I have done "mauchly(R,c)" as you said.But I got a 4X4matrix

I know the third row is the test results for Factor1 main effect and the fourth row is the test results for interaction effect .But I still don‘t know what's the meaning of the first row and the second row？ I will be very appreciated for your answers. Thank your!

[Tom Lane]

Look at the output from the ranova command. I suspect you will find four within-subject tests included. It may be that the first two are just univariate tests so there is no issue of sphericity for them.

[William]

Hi,

This is my codes for ANOVA:

When I opened the output argument 'c' of the ranova as you said, I saw a cell:

The elements of this cell were:

c{1,1}:

c{1,2}:

c{1,3}:

c{1,4}:

I still don't know why there were 4 elements.

So I still didn't know why there were 4 rows in the output argument of 'mauchly(R,c)', which was different from the output argument of SPSS.

Could you explain in detail? Thank you !