Mark Liberman tells us, “When you read or hear in the mass media that ‘Only X% of Americans know Y’, don’t believe it without checking the references — it’s probably false even as a report of the survey statistics.” He cites a few sources and gives an incredible example of the answers marked wrong for an open-ended question about former U.S. Supreme Court Chief Justice William Rehnquist. (via Alan Jacobs)

On this point, here’s a survey that appears to be done the right way, asking clear questions about religious practice and affiliation. It concludes that though many say they pray every day and have other spiritual or religious habits or experiences, “One-in-Five Adults Have No Religious Affiliation.” I wonder how much of this is a popular resistance to labeling one’s beliefs either out of a selfish desire to appear smart (like the voter who says he is undecided because he wants to appear to give both sides a fair hearing) or a belief that one cannot claim a label while not accepting absolutely everything ever claimed by someone of that label.