By Patrick Appel
A reader writes:
What I found most interesting in this graph were the extremes: that, across the (reported) political spectrum, at least 20% of people pray several times a day. And that, across the spectrum, only a tiny number (looks like less than 5%) never pray. That, and the similarity of all the bars--there really does not appear to be a very big variance, across the spectrum of political positions, in frequency of prayer.
Another adds:
The data provided on the graph labeled "Frequency of Prayer" are interesting but not all that revealing. A good test to apply is asking how well knowing a particular data point allows you to predict behavior. If you "know" that people self-identify as independent and "guess" that they pray daily, you will be wrong approximately seventy-five per cent of the time. If you know they strongly identify as Democrats or Republicans and guess the same, you will still be wrong about sixty-seven per cent of the time. A difference, yes, but how significant?