Hi All,
I have some questions about masking contrasts. Our study is designed such that English-speaking subjects listen to blocks of nonsensical stories, sensical stories, and then Korean stories as a control. What we would like to do is isolate the activation present during sensical stories while excluding the pure listening activation (Korean stories). I have questions about the .con files that are generated in 1st level analysis in SPM8. I am planning on using the Results manager to define two contrasts: contrast 1 is the masking contrast (Korean) and will be used as an explicit mask for contrast 2 (nonsensical-sensical stories).
Question 1: Can we use the activation pattern during Korean stories as a mask to isolate the activation of interpreting the English stories?
If so,
Question 2. Do the .con files contain information about the p-values that I input at this stage? For example, if I select a p=0.05 does the resulting .con file have information about my contrast at p=0.05? And if so, does this mean that a different .con file will be generate for each different p-value I try for the same contrast?
Question 3. If I define contrast 2 and then use contrast 1 as an explicit mask, what will be generated in the resulting .con file (con_0002)? Will it only contain information about the unmasked contrast 2, or will it contain information about contrast 2 masked by contrast 1? If con_0002 only contains information about the unmasked contrast 2, can I save the resulting masked image (contrast 2 masked by contrast 1) for use in 2nd level analysis?
Any help would be greatly appreciated. Thanks!
Clarice Wang
|