Hi Peter,

Thanks for your guidance. It worked! But one thing I want to ask is, when I plot beta against time bin, should I add a data point for time bin=0? Because my stimulus is ON at first 2 time bins, and I don't want to create a false impression that beta is already max right at the beginning of each trial, which doesn't look like the classical HRF curve.

Second question is, if I have 2 sessions for each subject, is it okay just to average both sessions to create 1 plot?

Thanks a lot for your advice!
Andy


On Fri, Apr 25, 2014 at 9:46 PM, Andy Yeung <[log in to unmask]> wrote:
Hi Peter,
Thanks very much for your clarification. I think what you said is correct.
Let me try it out and come back here to report my progress. Thanks!

Best,
Andy


On Fri, Apr 25, 2014 at 7:52 PM, Zeidman, Peter <[log in to unmask]> wrote:

Hi Andy,

I think the confusion here is due to terminology.

 

When you say you had two stimuli, I think you mean you had two conditions, let’s call them A and B. Some blocks were of condition A and some blocks were of condition B. The blocks were then randomly intermixed, e.g. ABBABAA. Is that correct?

 

If that’s correct, your condition A blocks and condition B blocks should be entered as separate conditions when you specify the GLM. Then each will be given a set of FIR columns.

 

Please forgive me if I’ve not understood your question!

 

Best,

Peter.

 

From: Andy Yeung [mailto:[log in to unmask]]
Sent: 25 April 2014 10:57
To: Zeidman, Peter
Cc: [log in to unmask]
Subject: Re: [SPM] How to use FIR for multiple contrasts in single session

 

Thank you Peter and To all,

 

Yes I had 2 stimuli in a session, they were in pseudo-randomized blocks of 6 seconds separated with 15 seconds of baseline conditions. TR was 3 seconds.

So, (6+15)/3 = 7 time bins.

Design Matrix gave me 7 columns corresponding to 7 time bins. But 2 stimuli are mixed in these 7 columns, I suppose. Can you help?

 

Best,

Andy

 

On Fri, Apr 25, 2014 at 3:58 PM, Zeidman, Peter <[log in to unmask]> wrote:

Dear Andy,

First, can I confirm you really mean 2 stimuli in a session? Are these blocks or events?

 

The FIR model will create a series of columns for each of your conditions. So if one condition (as defined in your onsets)  is stimulus1, and the other condition is stimulus2, then you’ll get what I think you want.

 

Best,

Peter.

 

 

From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Andy Yeung
Sent: 25 April 2014 04:54
To: [log in to unmask]
Subject: [SPM] How to use FIR for multiple contrasts in single session

 

Dear all,

 

I would like to investigate the change in beta at different time bins, for 2 stimuli in a session.

To Specify 1st level, the FIR model allows me to input contrasts scan images and define IOI according to TR.

 

But I don't know how to separate 2 stimuli at defining contrast in Results. In design matrix, each column is for each time bin. So, eg time bin 1 contains con images from both stimuli 1 and 2.

 

Can anyone help me on separating the stimuli into separate series of time bins?

 

Yours,

Andy