In teaching journalism classes the Religion Guy has often used the little 1954 classic “How to Lie with Statistics,” a great primer for any reporter, especially one like this writer who is mathematically challenged. The following has nothing to do with “lies,” but reminds us that though numbers appear to be hard facts they’re always subject to some spin.
That theme is raised as the media report on the new second installment of data from the Pew Research Center’s 2014 survey about religion with 35,071 respondents.
Such a massive sample allows a small margin of error. And unlike most pollsters the Pew team is very sophisticated about religion. For instance, if a person identifies as “Presbyterian,” is that the moderate to liberal Presbyterian Church (USA) or staunchly conservative Presbyterian Church in America, or some other body?
One caution: Princeton sociologist Robert Wuthnow would want us to note down on page 126 that the “response rate” among attempted phone calls was only 11.1 percent for landlines and 10.2 percent for cell phones. As the Religion Guy noted previously, this is a nagging problem in 21st Century polling.
Pew’s first installment last May grabbed many a headline with the news that Americans with no religious affiliation – those headline-grabbing "nones" – increased from 16 percent in a comparable survey in 2007 to the current 23 percent. (Hurrah to Pew for replicating its prior poll to show us such trending.)