Setanta wrote:The main problem, Ruach, with such a study, would be how to quantify and qualify belief.
Asking them whether they believe in god and testing for membership in a religious community are two quantifiable and objective tests that come to mind. They are not perfect -- but
if they are heritable and
if the study is statistcally sound, that would be good enough to make a decent case. I'm still skeptical though.
UPDATE:
I searched the Medline database for scientific medical publications and found
the abstract of Bouchard's original paper. Two points:
1) In statistical tests, 'significant' doesn't mean what it means in everyday language. It means "the chances of the observation occuring just by coincidence are less then 5 percent. An effect doesn't have to be large to be significant
2) The study examined 37 monozygotic twins, tested against 35 heterozygotic twins in the control group. It's not junk science, but still, the groups are rather small
3) As I read the abstract, the correlation in religiousness reflects a correlation of how submissive people are to authority. It makes sense to me that this submissiveness is inherited. For what it's worth, my everyday experience tells me that heterozygotic dog five-lings (or whatever you call that in English) show very different patterns of submission from too early on to be explained by training.