The Pew Research Center, as it is prone to do from time to time, released a religion oriented poll of Americans today, this time on the level of religious knowledge among Americans. The main takeaway picked up by major news outlets like the New York Times and the Los Angeles Times was that atheists and agnostics scored the best in the survey. While this is interesting and counter intuitive, if not incredibly surprising, I do want to do a better dissection a pull out some nuance in the results.
One thing that’s important to understand when approaching any poll is what the questions were and how they were asked. If you really want to get into the nitty gritty of the poll, I recommend checking out Appendix A for basic methodology and Appendix B (PDF) for the line of questions asked and the results for those. This is especially important for this poll, because as we shall see, the selection of questions for this survey influences the results in no small manner.
Now, the information that drove the news pieces covering this survey is the first summary table, which is basically the aggregate average score broken down by religious group:
As was pointed out by the news outlets, the Atheist/Agnostic category has the highest score. However, it should be noted that it is not much higher than that of the Jewish category or the Mormon category. To get a fuller picture, however, I think it’s good to look at the next table, which breaks the questions into three categories:
Here we can see from whence these advantages derive. Atheists/Agnostics and Jews score decently in the “Bible and Christianity” questions, but take third and fourth place respectively, recouping their deficit to the Mormons and White evangelical Protestants with the other two categories.
Now, even this oversimplifies the picture. Looking at the questions, I think that the three questions concerning Mormonism give Mormons the advantage that gives them such a better score than other Christians. Jewish scripture forms the basis for Christianity, so it would not be surprising that Christians would be familiar with the Jewish portions of the Bible (Old Testament), but as Mormonism is essentially a heretical sect of Christianity, most other Christians wouldn’t need to learn the history and tenets of Mormonism to understand their own faith. Regardless, Mormons scored better than average on the other sections too, so it wouldn’t be inaccurate to assume that they would have better general religious knowledge than the average Protestant or Catholic. [EDIT: looking a little closer, it appears that the Mormon questions were separated out and not included in the second table, as were the two atheism and agnosticism centered questions. I’ll do a second post that gets into some of the deeper details in the second and third sections of Pew’s report on the survey]
Another complication is the separation between Atheists/Agnostics and the “Nothing in particular” category. Certainly, it is justifiable to separate the two as the question allows for a clear separation of the two in the response, and many who would claim to be nonreligious are not atheistic or agnostic, but I suspect that a good portion of that group would qualify for one definition or the other, and as you can see by the above scores, those who claimed “Nothing in particular” didn’t score particularly well, suggesting that in filtering out self-proclaimed atheists and agnostics from the nonreligious bulk artificially raises their scores, as those who would identify as such are much more likely to be aware of religious issues.
Now, Jews do have some advantageous questions as well, particularly concerning when the Sabbath starts (Friday night) and to which religion Maimonides belonged (Judaism). In both questions, the modal answer was one of the incorrect choices (Saturday for the former and Don’t know for the latter, with Hinduism taking a distant second place).
Overall, I’d suggest caution in drawing strong conclusions from these results. They are informative, but the results are strongly contingent on the questions that Pew thought would make a good survey and a bit of tweaking on that end could have yielded rather different results.