Claus -
This email might help (a bit):
http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0608&L=SPM&P=R44072&I=-3
Does it actually make any difference to the results whether or not you
choose unequal variances (or nonindependency) in the GUI?
Rik
--------------------------------------------------------
DR RICHARD HENSON
MRC Cognition & Brain Sciences Unit
15 Chaucer Road, Cambridge,
CB2 2EF England
EMAIL: [log in to unmask]
URL: http://www.mrc-cbu.cam.ac.uk/~rik.henson
TEL +44 (0)1223 355 294 x522
FAX +44 (0)1223 359 062
MOB +44 (0)794 1377 345
--------------------------------------------------------
>-----Original Message-----
>From: SPM (Statistical Parametric Mapping)
>[mailto:[log in to unmask]] On Behalf Of Claus Lamm
>Sent: 01 September 2006 21:36
>To: [log in to unmask]
>Subject: [SPM] paired t-test/sphericity correction
>
>
>I was wondering why there is an option for non-sphericity
>correction for the
>paired t-test.
>
>The way I always understood it, the sphericity assumption can only be
>violated when more than 2 factor levels are entered into a
>repeated measures
>analysis. in the case of the paired t-test, this is not the
>case (k=2 levels).
>
>Thx for any comments, and have a good weekend
>
>Claus
>
>P.S.: to clarify things: I use the paired t-test in a rfx
>analysis, entering
>the contrast images of two conditions for 18 subjects as
>measurement pairs.
>the sphericity assumption states that the variance of the
>difference scores
>in a within-subjects design (Var(diff) in a paired t-test) are
>equal across
>all the groups; as there is only one such Variance in a paired
>t-test, this
>assumoption can"t be violated, no? or am I missing sth. here?
>
|