Originally posted by Andrew Zalesky:

Hi Cristobal,

yes - this looks correct. 

You could also run two separate one-sided t-tests using a contrast of [-1 1] and [1 -1]. This may yield findings that are more interpretable. 

Best wishes, 

Andrew

Originally posted by Cristobal Mendoza:

Hi All,

I have two groups (control and patients). My design matrix looks like this (with more than 2 subjects per column):

1 0

1 0

0 1

0 1

Is the contrast [1,1] correct for performing an F-test to assess two-sided differences between these groups?  

Best regards,

Cristóbal

 

Hello Dr. Zalesky,

Thank you very much for your help

Best regards

Cristóbal