Dear SPMers,
I was wondering if the SPM conjunctions at second
level could be better done with correlations.
At a group level, by a conjunction between conditions
A and B, we wish to claim that an activation in
condition A implies and is implied by activation in
condition B. At random effects, this issue can be
dealt with greater sensitivity as we can test if
subjects having low/high activation in condition A
have a correlated low/high activation in condition B.
In contrast, the Null conjunction tests if overall as
a group activation in condition A is significant "AND"
overall group activation in condition B is
significant. I am concerned about the Null conjunction
does not account for the fact that the activation in
one of the conditions could have just missed the
threshold. Global conjunctions could be biased by
strong activations in one of the two conditions.
Secondly, i feel that correlation tests could better
tease out the relationship between two conditions
rather than combination of independent t-tests.
Thanks in advance for your comments
-krishna
Krishna P. Miyapuram
Work : Graduate Student, Dept of Anatomy, Univ of Cambridge
Home : 25, Parkside, Room 10 (Phone +44 - 1223 - 476858
Address : Downing College, Regent Street, Cambridge, CB21DQ, England
URL : http://people.pwf.cam.ac.uk/kpm23/index.html
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
|