Rasmussen Reports, a highly reputable polling organization released a report this Easter weekend, 64% Believe Jesus Christ Rose From the Dead. They did the same survey a year ago Easter and had this result back then: 77% Believe Jesus Rose From the Dead. Hands up if you believe that 13% of Christians lost their faith in the foundational precept of their religion in a single year. That out of 225 million Christians, nearly 30 million lost their faith in one year?
If you don't believe that, and I don't, then what is going on with the numbers? Same sample size (1,000), same question, same margin of error.
I can't see any obvious reason to explain this other than that these surveys are dramatically less reliable than we are led to believe and that they are much more sensitive to timing or context or weather or sampling error, etc. With a margin of error of +/- 3%, a statistical fluke of sampling high one year and low the next (i.e. a maximum error range of 6%) would still leave a dramatic drop in faith of 7%. Still well outside the realm of believability.
I am usually deeply skeptical of most surveys anyway but this is a quite remarkable piece of evidence if you wished to make the argument that most surveys are not just worthless but dramatically misleading.