"Did you go to church this week? That's the question that Gallup pollsters have been asking Americans for more than 75 years. And each year since 1939, about 40 percent of those polled have said yes. (The actual question: "Did you yourself happen to attend church or synagogue in the last seven days?")I am not a statistician, so any comments I make could be totally naive.
That doesn't mean that, on any given Sunday, 118 million Americans (40 percent of the population) will actually be in church. According to sociologists who study religion, the actual number of people in church each week in the United States is significantly lower than the Gallup Poll indicates. Just how low is a matter of some debate...."
Kirk Hadaway and Penny Marler, researchers for the Episcopal Church and Samford University respectively, reasoned that if you knew the number of churches in the United States, and also knew the average attendance at each church, then you could compute the number of Americans in church on any given Sunday.
Their figures indicate 20.4 % -- half the Gallup figure of 40%.
Even such a statistically-challenged person as me can see that this does not fall within a simple sampling error.
A quick answer might be that people don't answer the pollster accurately, even though their anonymity is preserved. I'm not sure that one needs to assume dishonesty. As Frank Newport, Editor-in-Chief of the Gallup Poll points out, "I would say that we are right in saying that 4 out of every 10 Americans represent themselves as being regular churchgoers, but that does not mean that they are in church 52 weeks a year."
The researchers quoted in this article seem to agree that the Gallup figures, consistent as they seem to be, are high. At least one person interviewed, though, feels that the Hadaway and Marler figures are too low. Most of the others agree that the figure is somewhere between 30% and 40%.
Technorati tags: religion, statistics