Email discussion lists for the UK Education and Research communities

## SPM@JISCMAIL.AC.UK

#### View:

 Message: [ First | Previous | Next | Last ] By Topic: [ First | Previous | Next | Last ] By Author: [ First | Previous | Next | Last ] Font: Proportional Font
 LISTSERV Archives SPM Home SPM July 2008

#### Options

Subject:

Extracting Peak Maxima in Factorial Design

From:

Date:

Mon, 21 Jul 2008 23:39:30 +0100

Content-Type:

text/plain

Parts/Attachments:

 text/plain (42 lines)
 ```Hi there, I'm trying to plot maxima from ROI coordinates, and am having a difficult time understanding the results. I thought I understood what information the maxima provided me with...but I think I'm not entirely sure of the functions that go into calculating the maximas, because the results aren't exactly as I expected them to be. Any help would be much appreciated. The basics: I'm running a flexible factorial model with two variables: subject (30 subjects) and trialtype (4 trialtypes). So I have 34 columns in my design matrix: 30 subject columns, followed by the 4 trialtypes. I pick a t-contrast (for instance, positive trials > negative trials), plot the maxima, and extract the fitted Ys. Normally I would expect to receive 30 values, one for each participant. But I'm getting 120 values...which I assume is 30 x 4, or one for each trialtype per subject. My real confusion, however, lies in the fact that if I pick a different t-contrast (negative trials > positive trials, for instance), and extract the maxima from this image...I get a *different* 120 values. This makes me think that my initial interpretation of the 120 fitted-Ys as representing the maxima for each trialtype per subject is in error...because if that were the case the values should be constant across all contrasts. And so: can someone help me understand what information is being provided when I ask for the fitted-Ys? What do each of the 120 values represent? Thanks in advance, Matt _____________ Matthew S. Shane The Mind Research Network Albuquerque, NM 87104 [log in to unmask] ```