Thank you for your quick response. Initially I had the same idea. Unfortunately, this easy solution does not work. All trials are successor of the previous and predecessor of the following trial. Therefore we can't just randomly distribute pairs of trials. The whole chain of trials has to be in proper order. ([A-B-A-A-B-B-B-A-...] = [AB BA AA AB BB BA ...]) My thinking was to somehow add these restrictions to the initial part of optseq2 algorithm and fix for null event interuptions afterwards...
Tilman
--------------------------------------------------------------------------------- Tilman J. Gaber, Dipl.Psych. Translational Neuroscience in Psychiatry and Neurology Dpt. of Child and Adolescent Psychiatry and Psychotherapy RWTH Aachen University Neuenhofer Weg 21 52074 Aachen Germany
PHONE: +49-(0)241-80.85347 FAX: +49-(0)241-80.3335070 Email: tgaber@ukaachen.de ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Freitag, 26. Februar 2010 18:56 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
You can specify that there are 4 trial types (AA, AB, BA, BB), and give the duration twice that of the individual.
doug
Gaber, Tilman wrote:
Dear mailinglist recipients,
I am currently planning an optimized experiment using optseq2 and a 2x2 factors behavioral response task (Simon-task). The task consists of a pseudo-randomized series of two kinds of trials (congruent and incongruent). In addition there is a sequential factor, i.e. it is of interest whether a trial follows a trial of the same or the opposite kind. In other words there are (2x2) four different conditions where trials relate to precedent trials (AA,AB,BB,BA). Would it be possible to incorporate this sequential factor into the optseq2? I am aware that null events will fall between trials and thus interrupt trial sequences. Trials following null events would therefore be treated separately (i.e. for data analysis). In order to obtain an equal number of trials per condition, this would require increasing the number of trials depending on the number and order of null events. Currently I am able to produce a randomized order of equally balanced trials regarding trial type and sequential order (AA,AB,BB,BA). Any thoughts or suggestions are more than welcome!
Best, Tilman _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Optseq2 does not have the ability to change the identity of the stimuli during the run.
doug
Gaber, Tilman wrote:
Thank you for your quick response. Initially I had the same idea. Unfortunately, this easy solution does not work. All trials are successor of the previous and predecessor of the following trial. Therefore we can't just randomly distribute pairs of trials. The whole chain of trials has to be in proper order. ([A-B-A-A-B-B-B-A-...] = [AB BA AA AB BB BA ...]) My thinking was to somehow add these restrictions to the initial part of optseq2 algorithm and fix for null event interuptions afterwards...
Tilman
Tilman J. Gaber, Dipl.Psych. Translational Neuroscience in Psychiatry and Neurology Dpt. of Child and Adolescent Psychiatry and Psychotherapy RWTH Aachen University Neuenhofer Weg 21 52074 Aachen Germany
PHONE: +49-(0)241-80.85347 FAX: +49-(0)241-80.3335070 Email: tgaber@ukaachen.de ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Freitag, 26. Februar 2010 18:56 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
You can specify that there are 4 trial types (AA, AB, BA, BB), and give the duration twice that of the individual.
doug
Gaber, Tilman wrote:
Dear mailinglist recipients,
I am currently planning an optimized experiment using optseq2 and a 2x2 factors behavioral response task (Simon-task). The task consists of a pseudo-randomized series of two kinds of trials (congruent and incongruent). In addition there is a sequential factor, i.e. it is of interest whether a trial follows a trial of the same or the opposite kind. In other words there are (2x2) four different conditions where trials relate to precedent trials (AA,AB,BB,BA). Would it be possible to incorporate this sequential factor into the optseq2? I am aware that null events will fall between trials and thus interrupt trial sequences. Trials following null events would therefore be treated separately (i.e. for data analysis). In order to obtain an equal number of trials per condition, this would require increasing the number of trials depending on the number and order of null events. Currently I am able to produce a randomized order of equally balanced trials regarding trial type and sequential order (AA,AB,BB,BA). Any thoughts or suggestions are more than welcome!
Best, Tilman _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Mmmh... So, would it be possible to create a series of trials beforehand (which I actually have) and then let the optseq2 algorithm allocate the adequate amount of null events at optimal time points (iteratively of course)? Tilman ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Montag, 1. März 2010 19:08 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
Optseq2 does not have the ability to change the identity of the stimuli during the run.
doug
Gaber, Tilman wrote:
Thank you for your quick response. Initially I had the same idea. Unfortunately, this easy solution does not work. All trials are successor of the previous and predecessor of the following trial. Therefore we can't just randomly distribute pairs of trials. The whole chain of trials has to be in proper order. ([A-B-A-A-B-B-B-A-...] = [AB BA AA AB BB BA ...]) My thinking was to somehow add these restrictions to the initial part of optseq2 algorithm and fix for null event interuptions afterwards...
Tilman
Tilman J. Gaber, Dipl.Psych. Translational Neuroscience in Psychiatry and Neurology Dpt. of Child and Adolescent Psychiatry and Psychotherapy RWTH Aachen University Neuenhofer Weg 21 52074 Aachen Germany
PHONE: +49-(0)241-80.85347 FAX: +49-(0)241-80.3335070 Email: tgaber@ukaachen.de ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Freitag, 26. Februar 2010 18:56 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
You can specify that there are 4 trial types (AA, AB, BA, BB), and give the duration twice that of the individual.
doug
Gaber, Tilman wrote:
Dear mailinglist recipients,
I am currently planning an optimized experiment using optseq2 and a 2x2 factors behavioral response task (Simon-task). The task consists of a pseudo-randomized series of two kinds of trials (congruent and incongruent). In addition there is a sequential factor, i.e. it is of interest whether a trial follows a trial of the same or the opposite kind. In other words there are (2x2) four different conditions where trials relate to precedent trials (AA,AB,BB,BA). Would it be possible to incorporate this sequential factor into the optseq2? I am aware that null events will fall between trials and thus interrupt trial sequences. Trials following null events would therefore be treated separately (i.e. for data analysis). In order to obtain an equal number of trials per condition, this would require increasing the number of trials depending on the number and order of null events. Currently I am able to produce a randomized order of equally balanced trials regarding trial type and sequential order (AA,AB,BB,BA). Any thoughts or suggestions are more than welcome!
Best, Tilman _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
No, but that's an interesting idea, and might not be too hard to implement. So you already have a sequence, you just want nulls to go in them?
doug
Gaber, Tilman wrote:
Mmmh... So, would it be possible to create a series of trials beforehand (which I actually have) and then let the optseq2 algorithm allocate the adequate amount of null events at optimal time points (iteratively of course)? Tilman ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Montag, 1. März 2010 19:08 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
Optseq2 does not have the ability to change the identity of the stimuli during the run.
doug
Gaber, Tilman wrote:
Thank you for your quick response. Initially I had the same idea. Unfortunately, this easy solution does not work. All trials are successor of the previous and predecessor of the following trial. Therefore we can't just randomly distribute pairs of trials. The whole chain of trials has to be in proper order. ([A-B-A-A-B-B-B-A-...] = [AB BA AA AB BB BA ...]) My thinking was to somehow add these restrictions to the initial part of optseq2 algorithm and fix for null event interuptions afterwards...
Tilman
Tilman J. Gaber, Dipl.Psych. Translational Neuroscience in Psychiatry and Neurology Dpt. of Child and Adolescent Psychiatry and Psychotherapy RWTH Aachen University Neuenhofer Weg 21 52074 Aachen Germany
PHONE: +49-(0)241-80.85347 FAX: +49-(0)241-80.3335070 Email: tgaber@ukaachen.de ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Freitag, 26. Februar 2010 18:56 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
You can specify that there are 4 trial types (AA, AB, BA, BB), and give the duration twice that of the individual.
doug
Gaber, Tilman wrote:
Dear mailinglist recipients,
I am currently planning an optimized experiment using optseq2 and a 2x2 factors behavioral response task (Simon-task). The task consists of a pseudo-randomized series of two kinds of trials (congruent and incongruent). In addition there is a sequential factor, i.e. it is of interest whether a trial follows a trial of the same or the opposite kind. In other words there are (2x2) four different conditions where trials relate to precedent trials (AA,AB,BB,BA). Would it be possible to incorporate this sequential factor into the optseq2? I am aware that null events will fall between trials and thus interrupt trial sequences. Trials following null events would therefore be treated separately (i.e. for data analysis). In order to obtain an equal number of trials per condition, this would require increasing the number of trials depending on the number and order of null events. Currently I am able to produce a randomized order of equally balanced trials regarding trial type and sequential order (AA,AB,BB,BA). Any thoughts or suggestions are more than welcome!
Best, Tilman _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Exactly! I created a snippet of python code to create a series of trials of two kinds ( a ; b ). I restricted the output in terms of the maximum number of immediate trial repetitions and the number of each sequential pair of trials (i.e. AA AB BA BB) - which have to be balanced (i.e. 1/4 of total n). I attached an exemplary sequence of trials to play with. For a start I would not worry about null event interruptions, since I would like to limit the duration of null events (to approx. 2x trial length) and have a short ITI anyway. In other words: I would be happy to have the nulls put into the sequence (optimized by optseq2). :)
Tilman
________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Dienstag, 2. März 2010 19:04 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
No, but that's an interesting idea, and might not be too hard to implement. So you already have a sequence, you just want nulls to go in them?
doug
Gaber, Tilman wrote:
Mmmh... So, would it be possible to create a series of trials beforehand (which I actually have) and then let the optseq2 algorithm allocate the adequate amount of null events at optimal time points (iteratively of course)? Tilman ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Montag, 1. März 2010 19:08 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
Optseq2 does not have the ability to change the identity of the stimuli during the run.
doug
Gaber, Tilman wrote:
Thank you for your quick response. Initially I had the same idea. Unfortunately, this easy solution does not work. All trials are successor of the previous and predecessor of the following trial. Therefore we can't just randomly distribute pairs of trials. The whole chain of trials has to be in proper order. ([A-B-A-A-B-B-B-A-...] = [AB BA AA AB BB BA ...]) My thinking was to somehow add these restrictions to the initial part of optseq2 algorithm and fix for null event interuptions afterwards...
Tilman
Tilman J. Gaber, Dipl.Psych. Translational Neuroscience in Psychiatry and Neurology Dpt. of Child and Adolescent Psychiatry and Psychotherapy RWTH Aachen University Neuenhofer Weg 21 52074 Aachen Germany
PHONE: +49-(0)241-80.85347 FAX: +49-(0)241-80.3335070 Email: tgaber@ukaachen.de ________________________________________ Von: Douglas N Greve [greve@nmr.mgh.harvard.edu] Gesendet: Freitag, 26. Februar 2010 18:56 An: Gaber, Tilman Cc: freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Optseq2 - Incorporate sequential order of trials as a factor
You can specify that there are 4 trial types (AA, AB, BA, BB), and give the duration twice that of the individual.
doug
Gaber, Tilman wrote:
Dear mailinglist recipients,
I am currently planning an optimized experiment using optseq2 and a 2x2 factors behavioral response task (Simon-task). The task consists of a pseudo-randomized series of two kinds of trials (congruent and incongruent). In addition there is a sequential factor, i.e. it is of interest whether a trial follows a trial of the same or the opposite kind. In other words there are (2x2) four different conditions where trials relate to precedent trials (AA,AB,BB,BA). Would it be possible to incorporate this sequential factor into the optseq2? I am aware that null events will fall between trials and thus interrupt trial sequences. Trials following null events would therefore be treated separately (i.e. for data analysis). In order to obtain an equal number of trials per condition, this would require increasing the number of trials depending on the number and order of null events. Currently I am able to produce a randomized order of equally balanced trials regarding trial type and sequential order (AA,AB,BB,BA). Any thoughts or suggestions are more than welcome!
Best, Tilman _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Hi, I am still curious whether and how the above mentioned feature could be implemented into optseq2. I would be delighted and very greatful to hear your ideas.
Cheers, Tilman
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
Gaber, Tilman