Sunday, October 09, 2011

Evangelicals and cheating

Are married born-agains less likely to cheat than others?  Looking at GSS data (sample size = 6,423), the answer is no. Among whites, the percent unfaithful is 16.5 for evangelicals and 17.7 for everyone else. For blacks, it is 24.7 and 26.6 percent. The differences are not statistically significant. I checked to see if controlling for education would change the results. It doesn't.

1. I'd thought most people knew that although the percentages match in people who claim the label, when you analyze the data based on people going to church three or more times per week you find a substantial difference. (I think Barnas measured this.)

The problem with this measure is that although it's possibly true, it's also quite likely to be tainted, because people embroiled in a tumultuous marriage are unlikely to have the time and social freedom to go to church three or more times per week. The axes are not perpendicular (sorry, my math is geometric algebra, not stats).

Can we disentangle these variables? I can't think of a way.

-Wm

2. Anonymous4:48 PM

Unless Christians lie less o.O

Meta-analysis of clinical trials: Eat walnuts

I am always looking for easy eating choices that are good for you. This new meta-analysis of 26 clinical trials looked to see if walnuts ma...