The “analysis” is the collection of parameters you will use to analyze each session. It includes the signal model, the stimulus schedule, and any preprocessing options such as smoothing and motion correction that you might have run. Data are not actually analyzed at this stage, but with this program you define how you want the raw or pre-processed data for each session to be averaged together later.

You may want to analyze your study in more than one way, e.g. with and without motion correction, or you may wish to define your events based on subject response in one analysis, and on stimulus type in another. The key here is to realize that you have to give a name to all these parameters, and then refer to that name in later processing stages. Each analysis you set up defines a specific set of conditions and the same paradigm file name. IMPORTANT: Setting up your analysis is only done ONCE, and then it is applied to any and all sessions. If you decide you want to change your analysis parameters, you should either create a new analysis (with a new “analysisname”), or create a new analysis from scratch after deleting the original, which can then be renamed with the original name.

Each analysis name will be used in subsequent processing steps, and for multiple analyses, you have to run mkanalysis-sess separately, once for each analysis, giving each a different analysis name.

For this tutorial, you will create an analysis named “sem_assoc” for Bert, including all the following parameters:

To make this analysis, cd to your Study Directory and type:

mkanalysis-sess.new -analysis sem_assoc \
-TR 2 -paradigm sem_assoc.par -designtype event-related \
-timewindow 26 -prestim 4 -inorm -funcstem fmc -nconditions 4


CategoryHomepage

FsFastTutorialV4.5/900_Set_up_session-level_analysis (last edited 2011-05-19 15:29:07 by NickSchmansky)