Media (and Others) Behaving Badly with Stats

Several years ago there was an article in Books and Culture by Christian Smith on “Evangelicals Behaving Badly with Stats.” Well, it seems that Evangelicals aren’t the only ones to do naughty things with stats. It’s been getting around that Evangelicals are to blame for President-elect Trump’s unexpected victory. (See here, here and here.) But is that really the case. Joe Carter has a helpful post at The Gospel Coalition breaking down the 80% number that is being bandied about. I encourage you to read his post.I do have a few thoughts related to Carter’s post. First, we should all be cautious when reading polling data and statistics in general. It was only a week ago that polling data showed a near landslide win for Secretary Clinton. That turned out not to be the case. If that polling data was wrong, why should we have any confidence in polling data that concludes that 81% of white Evangelicals voted for Trump.

Second, it is helpful to remember that not all white Evangelicals voted for Trump. Carter wisely points out that roughly 43% of eligible voters did not even vote. That comes out to over 99 million American citizens that did not get out and vote. Surely, many of these people who chose not to exercise this privilege were Evangelicals, as Carter points out.

Third, the term “Evangelical” is not the easiest term to define these days. Roughly a hundred years ago the terms Evangelical and fundamentalist would have been synonymous. This was not the case after the 1960s though. Today, those who wish to determine who is truly a textbook Evangelical (see “Bebbington quadrilateral”)[1] go out of their way in surveys to ask several pointed questions (church attendance, prayer life, Bible reading, etc.) to find out if someone is actually an Evangelical. The point is that the term Evangelical may not accurately reflect the group you wish to define. (See the article at the Atlantic that notes that depending on one’s definition, Evangelical accounts for as few as 7% of the population to as much as 47% of the population in the States.)

Finally, it is interesting that those conducting the poll decided to section off non-white voters from white Evangelical voters. If one was white, one was given the choice to identify as Evangelical (or born again Christian, which is another ill defined descriptor). If one was non-white, one was not given the choice to identify as Evangelical. Here is the thing about stats, one can get the statistical data one wants by asking the right (or wrong depending on one’s point of view) kind of questions. It is possible to ask questions of one’s survey population in such a way as to affirm what one hopes to find. Now, I’m not saying that the pollsters in question had such an agenda. I’m only pointing out that it is possible to do so.

So, next time you come across stats, particularly stats that reveal surprising results and make bold claims, be sure to pause and consider what was actually being asked. Go to the source of the poll or survey, if possible (sometimes there isn’t even a survey to consider), to see for yourself whether or not what the data revealed truly represents reality.


[1] David Bebbington, Evangelicalism in Modern Britain: A History from the 1730s to the 1980s (New York, NY: Routledge, 1989), 2–3.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s