Originally posted by Symmetric Chaos
Now there's a complicated study to try doing. How to find out the ways people who take surveys are different from people who don't."You didn't respond to our last e-mail survey. Please fill out this form and return it."
generally speaking, people who are more willing to participate in surveys are the type who are the most interested in "how they score", or rather, they are the most likely to fall for things like demand characteristics...
for example, I'm running a study about how eyes and hands work together in a reaching task. To calibrate the equipment, I need to measure people's pupil distance. When I tried to do this to one subject, they replied, "oh don't worry, I'm perfect", which is a nonsensical comment. They kept stopping part way through the experiment to see how they were doing, (to the point where their movement messed up the data.... dumb subject), as if there were some "optimal" or "right" way to reach out and grab stuff. anyways, this is risking a tl:dr rant, so to wrap up, it is probably the people most concerned with supplying the answers that the survey takers want that are most likely to partake in the survey.
However, this almost certainly would be equal between religious groups, so all groups would have that bias equally.
social psych isn't my thing, but they are the field where most survey work is done. I'll save my cynical opinions of surveys for now, and say this, there are extreme steps proper studies go to in an attempt to weed out this kind of bias on survey reporting, that this study did not utilize. It probably doesn't matter because the comparison is between religious groups, but the absolute figures are probably not representative, however, the between groups comparison probably is (I can explain that more to anyone... stupid stats, lol!)