Also, don't look at band statistics: look at birth statistics.
Too small a study group will give invalid results. For example - there's a group here who are getting into organic foods and they all had girl babies.