 |
 |
In FSL, there is a separate model for each run. In SPM, each run is analyzed separately, but within a single GLM. The regressors associated with each run are 0s in all other runs and thus do not influence the other runs. In SPM, you use contrasts to average across sessions, which is what FSL does as a second level modeling step. Thus, you don't need to have all conditions in all runs. You just need to average the conditions of interest - and I stress the word average as if you don't average, then you could get a scaling difference between subjects. Best Regards, Donald McLaren ================= D.G. McLaren, Ph.D. Research Fellow, Department of Neurology, Massachusetts General Hospital and Harvard Medical School Postdoctoral Research Fellow, GRECC, Bedford VA Website: http://www.martinos.org/~mclaren Office: (773) 406-2464 ===================== This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is intended only for the use of the individual or entity named above. If the reader of the e-mail is not the intended recipient or the employee or agent responsible for delivering it to the intended recipient, you are hereby notified that you are in possession of confidential and privileged information. Any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited and may be unlawful. If you have received this e-mail unintentionally, please immediately notify the sender via telephone at (773) 406-2464 or email. On Fri, Mar 20, 2015 at 2:36 AM, Johannes Klene <[log in to unmask]> wrote: > Dear Donald and others, > thank you for your time and suggestion. The only reason I'm trying to > model this trial type for all runs is because I have multiple runs for each > subject, which I'm analyzing simultaneously (as in > http://mypage.iu.edu/~ajahn/docs/Run_SPM_FirstLevel.m ). So, within > subject, I have some runs with error trials and some without, and this is > what is causing an error with that script. > In FSL, I just analyzed runs individually and then combined the cope and > varcope images later. Is this then what I should also try to do in SPM? I > guess the real problem I'm having is not yet being familiar with SPM, and > therefore trying to follow generic scripts. How do you combine runs within > subject in this case? Can you point me towards some relevant documentation? > Thanks. Kind regards, Johannes > > > 2015-03-19 18:53 GMT+01:00 MCLAREN, Donald <[log in to unmask]>: > >> If an EV has no events, then you don't need to model it. The only reason >> to leave a blank column in the model is to make it easier to generate >> contrasts. However, if you download the gPPI toolbox, then you could look >> at createVec.m for how to create contrast vectors for each subject even if >> the subjects have different columns in each session. >> >> Hope this helps. >> >> Best Regards, Donald McLaren >> ================= >> D.G. McLaren, Ph.D. >> Research Fellow, Department of Neurology, Massachusetts General Hospital >> and >> Harvard Medical School >> Postdoctoral Research Fellow, GRECC, Bedford VA >> Website: http://www.martinos.org/~mclaren >> Office: (773) 406-2464 >> ===================== >> This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED >> HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is >> intended only for the use of the individual or entity named above. If the >> reader of the e-mail is not the intended recipient or the employee or >> agent >> responsible for delivering it to the intended recipient, you are hereby >> notified that you are in possession of confidential and privileged >> information. Any unauthorized use, disclosure, copying or the taking of >> any >> action in reliance on the contents of this information is strictly >> prohibited and may be unlawful. If you have received this e-mail >> unintentionally, please immediately notify the sender via telephone at >> (773) >> 406-2464 or email. >> >> On Thu, Mar 19, 2015 at 12:06 PM, Johannes Klene <[log in to unmask]> >> wrote: >> >>> Dear SPM experts, >>> I'm new to SPM... I am trying to run a first-level analysis on a task >>> which may, in some runs, lead to an empty EV (modeling incorrect trials, >>> while sometimes no mistakes were made). Something very similar has been >>> discussed before on Jiscmail: >>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=spm;da558d85.05 , but to >>> be honest this discussion went a bit over my head and it was also 10 years >>> ago. >>> Is anyone aware of a solution for my case? If it changes anything btw, >>> I'm not at all interested in this EV, it is simply present to better model >>> the data and I don't plan to involve it in any contrast. >>> Thank you. Kind regards, Johannes. >>> >>> p.s. my first-level script is largely based on this example: >>> http://mypage.iu.edu/~ajahn/docs/Run_SPM_FirstLevel.m >>> >> >> >