I haven't offered them because 1) we've discussed them in other threads before and 2) no one has asked.
I'll offer my methods now - they are plain statistics and have been peer reviewed. Two methods were primarily employed, simple arithmetic and linear regression (on my regressions, I saw adjusted R^2 consistently around .74 for each of the model runs). Controlled for a number of factors that are accessible via the full survey data (available to each American via a FOAA request - I have hard and electronic copies). Surprisingly, a larger portion of the inmates said they converted to atheism (if that is really possible) after their incarceration than those converting to Christianity. There are notable increases in post-incarceration religious participation, however, meaning that people who were non-practicing Christians become more interested in religion after they are jailed.
Anything else?
To me, this pretty much sums you up, jochhe. You consider "feasible" only those things which are consistent with what you already believe.Not good results LG, feasible results.
If I told you that the the name for white boys that has the highest correlation to maternal education is "Dov" you may or not find that feasible - but it doens't matter, because it is true (Levitt, 2005).
The statistics are what they are.

Reply With Quote
