Hi Amna,
How one would constrain the intensity threshold to certian limit so the results are comparable between sessions?
The intensity threshold slider is set based on the maximum value for each file. If the maxima are different, the threshold slider will not behave in the same way. There is no solution for making the display homogeneous across files. This issue has already been identified. A complete re-coding of the connectivity graphs is needed, but we lack of resources for this.
One way of representing the difference between the two conditions is simply to compute the difference with the Process2 tab, and to display it.
Note that this "intensity threshold" slider should not be used as a proper way of thresholding to the data. It hides everything that is lower than a fraction of the maximum, and its effects on a connectivity graph based on Granger values might not be very meaningful. What you need to do is a proper test across subjects, as you mentioned.
I would like to run a statistical analysis (across subjects) for showing alterations in connectivity (because of training) in right temporal region in session1 comparing to session2, is there any option avaiable in Brianstorm to run such analysis.
You should be able to use non-parametric tests on these connectivity results.
http://neuroimage.usc.edu/brainstorm/Tutorials/Statistics#Nonparametric_permutation_tests
Francois