Neal
I think "parametric modulations" are the answer to your question. Please
search the archives, and you should find much discussion on this (or see
example use on "data" part of SPM website).
Rik
--------------------------------------------------------
DR RICHARD HENSON
MRC Cognition & Brain Sciences Unit
15 Chaucer Road, Cambridge,
CB2 7EF England
EMAIL: [log in to unmask]
URL: http://www.mrc-cbu.cam.ac.uk/~rik.henson
TEL +44 (0)1223 355 294 x522
FAX +44 (0)1223 359 062
MOB +44 (0)794 1377 345
--------------------------------------------------------
>-----Original Message-----
>From: SPM (Statistical Parametric Mapping)
>[mailto:[log in to unmask]] On Behalf Of Neal Fitzpatrick
>Sent: 05 September 2007 16:56
>To: [log in to unmask]
>Subject: [SPM] Reaction Time Regression
>
>
>Hi List,
>
>A question from a young SPMer. We have results in SPM2 from an event-
>related working memory paradigm, and we want to analyze data
>using reaction
>time as a regressor in order to isolate performance based effects. In
>other words, the goal is to identify the specific components
>within the
>neural network that vary depending upon task performance. To
>do so, we
>have been inputting the reaction times as a regressor at TRs
>of interest,
>and a 0 for any TRs of no interest. We then have two questions:
>
>1) Will processing the data in this manner isolate those areas
>within the
>neural network that respond to change in RT?
>2) When considering traditional ER designs, it appears that
>the vectors of
>interest for the data are time shifted to accommodate the
>delayed onset of
>the HRF, but that this is not the case for vectors of onset
>where other
>numbers (such as reaction times) have been entered. Does this
>delay need to
>be entered manually when conducting regression analysis? That
>is, should
>we shift our reference function by 2-3 TRs in order to accommodate the
>latency in the HRF or is SPM still doing this automatically?
>
>Thanks so much for your help.
>
>Sincerely,
>Neal Fitzpatrick
>
|