Email discussion lists for the UK Education and Research communities

## FSL@JISCMAIL.AC.UK

#### View:

 Message: [ First | Previous | Next | Last ] By Topic: [ First | Previous | Next | Last ] By Author: [ First | Previous | Next | Last ] Font: Proportional Font

#### Options

Subject:

From:

Date:

Tue, 22 Sep 2009 14:39:10 -0400

Content-Type:

text/plain

Parts/Attachments:

 text/plain (72 lines)
 ```Hi, I have a question about the design matrix of for a higher level data analysis. The design is different from any of the examples I've seen, and I'm not sure what the appropriate design should be. I want to test the effect of a drug on monkeys. I have three monkeys (S1, S2, S3) and each of them was tested twice, several months apart. (each trial is considered as independent). I'd like to consider the two runs for each monkey as equivalent trials - multiple "runs" that can be combined. I've tried this two ways, and I get different results; I'm not sure which is correct. The first level analysis used a single EV of interest (the drug administration timeourse - a step function at the time of injection, to give a beta for "response to drug") with a number of confound regressors: 6-axis motion, and a variable number of single point regressors to covary out large impulsive motions above a certain threshold. The beta from the drug response is passed up to the next level. 1. The first design combines everything at the second level:       Group EV1 EV2 EV3 S1 1 1 0 0 S1 1 1 0 0 S2 1 0 1 0 S2 1 0 1 0 S3 1 0 0 1 S3 1 0 0 1 C=[1 1 1] 2. The second design consists of three individual second analyses and third level: a. Second Level:       Group EV1 S1 1 1 S1 1 1 C=[1]       Group EV1 S2 1 1 S2 1 1 C=[1]        Group EV1 S3 1 1 S3 1 1 C=[1] b. Third level:             Group EV1 AveS1 1 1 AveS2 1 1 AveS3 1 1 C=[1] I thought I would get similar results from two designs above. However, the results are quite different (the activation maps don't even overlap!). Can someone tell me which is the correct design matrix I should use and why? Thanks in advance. Best, Yunjie```