help
help > RE: pair t-test for 1 group
Feb 13, 2017 11:02 PM | Andrew Zalesky
RE: pair t-test for 1 group
Hi John,
details about exchange blocks are provided in the manual. The exchange block vector should be the same length as the number of rows in your design matrix.I provided an example in my previous post.
Adding an exchange block obviously won't guarantee that your results suddenly become significant.
Andrew
Originally posted by John Zang:
details about exchange blocks are provided in the manual. The exchange block vector should be the same length as the number of rows in your design matrix.I provided an example in my previous post.
Adding an exchange block obviously won't guarantee that your results suddenly become significant.
Andrew
Originally posted by John Zang:
Dear Prof. Zalesky,
Thanks for the kind reply.
if I get it correctly, my design matrix and contrast are fine. However I didn't add exchange blocks. I guess this is probably the reason (in help icon, it is optional)?
Another question: the exchange blocks vector should be a n*1 input where n is the length of correlation matrix?
Thanks for the kind reply.
if I get it correctly, my design matrix and contrast are fine. However I didn't add exchange blocks. I guess this is probably the reason (in help icon, it is optional)?
Another question: the exchange blocks vector should be a n*1 input where n is the length of correlation matrix?
Threaded View
Title | Author | Date |
---|---|---|
John Zang | Feb 10, 2017 | |
Andrew Zalesky | Feb 10, 2017 | |
John Zang | Feb 13, 2017 | |
Andrew Zalesky | Feb 13, 2017 | |