Lies, damned lies, and...
According to a Gallup analysis of 676,000 interviews conducted as part of their Gallup-Healthways Well-Being Index, the very religious portion of the populace in the USA have higher well being than do the less religious and the non-believers. Not only that, those who profess a faith, no matter who religious they are do better on the well-being than do we non-believers. We are the unhealthiest of the lot, according to Gallup's statistics.
Kind of scary isn't it - if you believe it.
To be honest, I had my doubts right off the bat. It just does not seem credible based on personal observation. But of course, I could be wrong. And with 676,000 interviews, it seems to be a pretty solid sample.
And I will admit that if the survey had shown Seventh Day Adventists to be significantly healthier, I would have believed it. A non-smoking vegetarian diet is part of that denomination's teachings, and those who follow it strictly generally do show up as healthy. But surely, there are not enough Seventh Day Adventists to skew the overall statistics.
Nothing to do, but look at the study details.
They split the population into three categories:
- Very religious -- Religion is an important part of daily life and church/synagogue/mosque attendance occurs at least every week or almost every week. This group comprises 41% of the adult population.
- Moderately religious -- All others who do not fall into the very religious or nonreligious groups but who gave valid responses on both religion questions. This group comprises 28.3% of the adult population.
- Nonreligious -- Religion is not an important part of daily life and church/synagogue/mosque attendance occurs seldom or never. This group comprises 30.7% of the adult population.
I find it interesting that they managed to find those in the overall No religion / atheist / agnostic category in each of these three divisions - even though the idea of a very religious atheist seems questionable.
Now the other detail we should look at is survey methodology, as described by Gallup:
Results are based on telephone interviews conducted as part of the Gallup-Healthways Well-Being Index survey Jan. 2, 2010-Dec. 30, 2011, with a random sample of 676,080 adults, aged 18 and older, living in all 50 U.S. states and the District of Columbia, selected using random-digit-dial sampling.
Telephone interviews! What the survey is completely based on is self-reporting. People stating whether or not they are religious, people saying whether or not they smoke, people describing their diet, their health, their emotional well-being, with no confirmation that any of the facts were correct. The survey covers what people say they do, not what they actually do.
This is quite important in assessing the validity of the conclusions.
Religious Tolerance has noted that over a period of many years, people in North America have consistently overstated their religious practices. Consistently, since the 1930s, over 40% of people have self-reported weekly church attendance, as does this latest Gallup survey. And consistently, studies of actual church attendance have shown that only about 20% attend church weekly.
In other words, half of those who say they attend church weekly are giving misinformation. Wikipedia generously suggests that they are actually lying to themselves to keep up their self-image.
If that is the case - whether it be knowingly lying to impress the interviewer, or unconsciously lying to maintain a self-image, is it not highly probable that these individuals would use the same rational to put a positive spin on their diets, smoking habits, and emotional state? If they exaggerate their religiosity, then it is reasonably to suspect they exaggerate how healthy their lifestyle choices are.
If we assume - and I emphasize assume - that if just half of those that exaggerate their church attendance similarly exaggerate their healthy lifestyle, then it is quite possible that the very religious would drop from the healthiest to the unhealthiest portion of the population.
It seems clear that the study results can be ignored. With the degree of error in self-reporting, no legitimate conclusions can be drawn.