Dear Ming Ting
> Darren:Sorry for bother you, I have a basic but important question about SPM99's random effect vs. fixed
> effect model.Thank you for your help in my previous mail about RFX. Please allow me to bother your time
> again. My question is that, if I still want to do "Fixed Effect" analysis for 10 subjects, can I use SPM99 with
> the design input as what I used for SPM96?
>
No bother at all. Yes you can still do fixed effect analyses in spm99
using a similar design to spm96.
> I have this question because of that in recent SPM mail post my question about "equiv. Z", which SPM99 reports as
> "infinite". Now I understand it means it is higher that Z = 7.8. But I am wondering why not many people has the same
> problem as mine?
>
We see infinite Z's all the time. We didn't view it as a problem.
> Is it because SPM99 treat ti the design as random effect by default no matter I input them as "fixed
> effect"?If yes, why kind of medication should I made for SPM99? Or should I go back to use "SPM96"??
>
No SPM99 designs are not random by default. You need to set them up that
way by having 1 scan per subject, and the other specifications that have
been noted on the list in the past.
As far as medication- I have always needed to be heavily medicated
before using spm. I think the authors medicate themselves every friday
night as well. (just kidding).
...and don't go back to spm96. I don't think there are any advantages to
doing that unless you have old data that you don't want to redo in
spm99. In short don't worry about the Inf results, and unfortunately you
still need to design random effects analyses explicitly.
Best regards,
Darren
> ******************************
> * Please also send a copy to *
> * [log in to unmask] *
> ******************************** Sincerely Ming-Ting
> ===================================================================
> Ming-Ting Wu, MD. Department of Radiology
> 386 Ta-chung 1st Rd.
> Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, 813
> Tel: 886-7-3422121 ext 6205, 6235
> Fax: 886-7-3468301
> mobile: 0920040712
> ----- Original Message -----
>
|