Without additional data it's difficult to be sure from the sparse information presented, but I'm thinking that the study's inclusion of people who are not from supportive / religious backgrounds 'may' have reduced the final success figures to a point below what they might have been if the study had included only people from supportive and religious families...in other words, I'm thinking that the success rates may have been significantly higher if only people from a preexisting supportive and religious background had been studied.
Despite the sparse and incomplete nature of the available information, I thought that this might still be of interest to FReepers in that it reinforces what many of us have felt for some time.
FYI
The media responds to the survey.
OK, but this is still based on self-reported data with a few bells & whistles from the statistical end. Regression analysis has an error rate. There is no alternative verification for the self reports the subjects made.
Cynic & statistician.