How to get a statistically significant mean from multiple matrices?

I have ten 4-D matrices which I would like to have a mean of (which is easy enough). However, I would also like the mean to be statistically significant amongst the ten matrices. Is there a way to shut out outliers, or do a form of t-test between more than two matrices? I am not too great with statistics so I think there should be a function that I just don't know about.

Thank you in advance!

Answer 1

[Star Strider]

What you do with respect to statistical analysis depends on how you designed your study. (Ideally, this consideration was part of your study design.) Multiple comparisons are not trivial. Texts on experimental design discuss it in detail.

The most relevant MATLAB function is likely multcompare (link).

See someone in your university statistics department for guidance.

[HA]

Thank you for your reply. Indeed, unfortunately this analysis was not part of the original project plan as the method was based on the method of a recent paper that only used the mean between the data sets, and not the statistically significant mean. It was an after thought that I should test the significance of the mean for my project as one can not be too careful with multiple data sets!

I shall read through what literature I can find and see if I can work with multcompare, failing that I will ask around the office!

Thank you for this starting point.

[Star Strider]

My pleasure.