Much Polling Data Overstates Church Attendance and Christian Identity

Many have thought this to be the case for decades but now there is a way to measure it. That is the answer people give an actual person doing phone surveys about religion. The suspicion for years is that the person answering questions felt it was his/her obligation not to say or imply he/she is not Christian. This made answers about church attendance, frequency of prayers and belief in the afterlife have bigger numbers than they would have if polling were done differently.

There is now a different way to do polling. It is to text on a cell phone. I text everyday and it has become quite ubiquitous. Polls inquiring about church attendance and praying using texts show lower "Christian" numbers than polls using actual people. In some cases the numbers are significantly different. 

When I mentioned this to my wife, a sociologist, she immediately said young people are more likely to text than older more religious people. Since they are less likely to be church goers this tips data lower than it might otherwise be. This involves two assumptions, that young people text more and they are less likely to be church goers. Even if true is it not encouraging for the future of the faith.

When 3,000 people were asked, "How important is religion in your life?" 48% said "very" when asked by a person on the phone. When asked by text, the number was 39%.  When asked how often they attended church, people talking on the phone with a pollster said "never" 18% of the time. When asked by text the number who said "never" was 28%. This latter number is very close to the percentage of people who, when polled, said they are not affiliated with any religion, "nones."

Any poll that asked about a respondents thoughts is fraught with risk of bias. There may be some reason people who text feel social pressure to deny being religious. I don't know why this would be but it's something hard to nail down. 

An overview of this difference in answers from text rather than human pollsters was provided by one experienced poll company manager. He said differences in answers to machine polls versus human were first noted almost ten years ago. The differences have continued to grow.

Differences in responses to the simple question, "Are you Christian?" was not as large, 60% yes to texts and 63% to human polling. In either case that number slowly but steadily declining.