AFNI Message Board

History of AFNI updates  

|
January 17, 2003 03:39AM
Dear Andrew,

Thanks for you reply. I'm afraid I cut the question a little too short though: I'm familiar with the 3DANOVA plugins, however ANOVA would not be the best statistical approach for analyzing my data.

I'm studying a population of 30 patients that I scanned 3 times while they were on different medication regimens of a single type of medication. Therefore, a comparable measurement was repeated 3 times for each individual, while varying 'only one' parameter (medication). To analyze the effect of medication on brain activation patterns during task-performance, the best statistical approach would be a test that is called a 'Repeated Measures Design' in SPSS, which allows you to define a within-subject factor 'regimen' with three different levels. This approach takes into account the fact that for each patient, there are 3 measures that are paired in nature, which ANOVA doesn't.

I could use a 3DANOVA 2 or 3 design with medication as a fixed factor, and subjects as a random factor, which would be as close an approximation as one might get for intra-subject effects using ANOVA, however I wondered whether an even more sophisticated test had already been developed for use in AFNI.

Hope I made myself a bit more clear on this (replies are still very much appreciated!),

Kind regards,

Rutger Goekoop
Subject Author Posted

Repeated Measures Design

Rutger Goekoop January 10, 2003 10:38AM

Re: Repeated Measures Design

Andrew Mayer January 16, 2003 10:45PM

Re: Repeated Measures Design

Rutger Goekoop January 17, 2003 03:39AM

Re: Repeated Measures Design (Right, Doug?)

sally durgerian January 17, 2003 11:06AM

Re: Repeated Measures Design (Right, Doug?)

Rutger Goekoop January 20, 2003 10:25AM

Re: Repeated Measures Design (Right, Doug?)

B. Douglas Ward January 20, 2003 12:52PM

Re: Repeated Measures Design (Right, Doug?)

sally durgerian January 20, 2003 04:34PM

Re: Thanks

Rutger Goekoop January 21, 2003 04:35AM



Sorry, only registered users may post in this forum.

Click here to login