Thursday, February 05, 2009

Do data matter? The case of Christian divorce statistics.

One of the first things that I learned in college was about Thomas Kuhn and paradigm shifts. Here's an extended discussion of it, but, basically, once people get a theory or idea in their heads, it will stay there even in the face of countering evidence. There is inertia to our worldview, and it takes a lot to change it.

Two years ago, when I started blogging, I did a series of divorce rates among Christians. The prevailing view on this was that Christians had high divorce rates, equal to people not affiliated with religion. Turns out this isn't the case. Both Christians and members of other religions have lower divorce rates than people unaffiliated with religion, and people who are more religious have lower rates than those less religious. I did what I thought was a nice deconstruction of Barna's statistic, explaining their what-I-considered-misleading assumptions.

Well, perhaps in my naivete, I figured that Barna's statistic would run out of steam and people who talk about this topic would do so more accurately.

After Googling the topic, it seems that Barna's statistics on this topic are as widely used as ever. Sigh, no paradigm shift yet.

As a bit of an experiment, I sent a letter to, who is probably the most consulted site on this topic (and one that puts forth Barna's findings). I briefly explained the problem with Barna's work, and sent along a manuscript that I'm working on that summarizes the issues.

Here's the question: Do you think religioustolerance will change their entry? If it's receiving a lot of hits, which it is, and it tells a catchy story--Christians gone bad--why would they change it? Now I realize they are very busy and probably understaffed and facing budget crisis, like everyone these days, but you'd figure they might want to get around to correcting what can be argued is a clear mistake.

Any predictions?


Jeff L said...

Someone (it may have been Kuhn) once said that sometimes the only way for a shift to occur is for all the supporters of the current paradigm to die out, allowing the next generation of scientists (or whoever) to be more free to pursue the new paradigm.

There's also the Quine-Duhem thesis: anomalous data essentially don't matter because auxiliary hypotheses can always be modified in order to save the theory or paradigm.

But we can hope...

kent said...

My assumption is that they will not because (I have such a jaded soul) they get too much mileage out of the erroneous data. If they are quote with increased frequency the odds of their making a significant alteration is pretty low.

Fajita said...


Authority, econommics, rhetorical power, leverage, and theory can all be obstacles to a legitimate search for truth. Not to mention pride and political agendas.

I confess my vulnerability to these forces as well.