In yesterday's post, I broke several cardinal rules of quoting statistics. I didn't quote the sampling error, I didn't provide a link to the full report, I didn't double-check the figures. In short, I quoted a quotation of the statistics, which means who knows how the data should be interpreted?
And that, of course, is key. We can grant that the statistics are perfectly sound, but all that means is that if you asked everyone in the population, you'd get about the same distribution you got by asking only a tiny sample of the population. Statistics tell us nothing about what statistics actually mean.
I don't know if I used too many words in my "essay or multiple choice" bit, or what. Maybe my main point can be summed up, and understood, in these words: Giving one response rather than another to a question in a telephone poll does not constitute an act of Christian love.