I have a question about a conjunction analysis.
My experiment includes two semantic tasks for words and pictures (and two
control tasks). I have done a conjunction analyses at the 2nd level by
choosing a factorial design with 1 factor (semantic) and 2 levels (pictures and
words). I use SPM 5.
1) Have I done the analysis correctly? From previous emails I gathered that
some people do conjunction analyses for each subject at the 1st level
analyses and then they do a 2nd level analysis.
2)The results look good at p = 0.001 uncorrected. However, none of the
clusters survive FWE correction. Is the threshold conservative enough? From
previous emails I understood that a conjunction is quite conservative and you
can choose a low threshold. I just want to make sure this is generally
Hope you can help,