Personally I've never seen mean % bold signal change values (from a region)
as high as 2%. They typically range from -1 to 1%, in my experience,
although there are larger values reported in the literature. So an effect
of 0.7% would be a strong effect, in my experience.
Obviously if you combine the featquery and effect.sh techniques, you'll get
some kind of hybrid. Perhaps Steve can correct me if I'm wrong, but the
output of featquery needs to be divided by 100 to turn it into a
percentage, even when you check the "% signal" option. Even so, that would
imply an effect of 1.45% which still isn't 2.0% nor is it the 0.7% from the
I don't know why featquery and effect.sh give different results -- they
shouldn't, particularly not by a factor of 2!