Tuesday, November 24, 2015

Bearing the Crown of Scorn

In "Who's the Most Humble? We Are!" Kevin Drum draws my attention to something Carly Fiorina said to which People for the American Way drew his attention.

"I do think it's worth saying," Fiorina declared, "that people of faith make better leaders because faith gives us humility"

Heh indeed. They also have a better sense of irony and self awareness.

Drum is not pleased.

He did write

"I've got nothing against organized religion. It provides an important part of life for a lot of people and does a lot of good charitable work. "

But then "" and wrote a devastating critique.

My rant volume goes up to 11

(his rant quality is much higher -- "always click the link" -- Kevin Drum).

My comment.

But even those of of us who don't believe that He technically Exists must be hypothetically eternally grateful to Jesus Christ for His willingness to suffer. I'm not talking about a bit of scourging and 3 hours on the cross. For such a benevolent and omniscient being the pain due to the humiliation of being associated with people like Fiorina must make crucifixion seem like a walk in the park.

OK enough snark now how about some science. I tried to check if Luntz was telling the truth. I know that socially undesired outcomes including teenage pregnancy are more likely in highly religious states. I also know what "ecological fallacy" means.

Of course I googled. It turns out that it is hard to get to other than obviously biased by googling [religious faith and teenage pregnancy]. The first few hits were all to heritage foundation sites. I chose to go to google scholar and found

http://www.sciencedirect.com/science/article/pii/S1054139X04004124

"Conclusion

Frequency of attendance and religious affiliation have little impact on sexual behaviors once intercourse occurs."

Which contradicts Luntz's claim because of the qualifiers "every single" and "all".

I also recall Case and Katz (1991) "The Company You Keep"

http://www.nber.org/papers/w3705

Which makes me feel old. The relevant bit of the abstract is "We find that family adult behaviors are strongly related to *analogous* youth behaviors. The links between the behavior of older family members and youths are important for criminal activity, drug and alcohol use, childbearing out of wedlock, schooling, and church attendance."

** mine.

The point is that, in regressions including the other adult behaviors, parents' church attendance is significantly related to youths' church attendance but not statistically significantly related to youths' criminal activity, drug and alcohol use, childbearing out of wedlock, or schooling.

Now this doesn't mean that organized religion has no effect -- to conclude that would be to reject the alternative hypothesis. It doesn't even mean that Luntz's claim is false -- he said "correlated" not "partially correlated," so he referred to simple regressions not multiple regressions.

I am sure Luntz knows about omitted variable bias. I am sure he is deliberately appealing to omitted variable bias to strengthen (and flatter) the faith of the audience.

I am as atheist as Kevin Drum is and as confident that organized religion is, on balance, a good thing*, but I am sure that Luntz's claim about the current level of empirical support for this view is based on the deliberately misleading use of the word "correlated".

* Yes I know that, according the the Gospels as written decades after his death, Jesus said "the truth shall set you free" and I just said I believe religious beliefs are are for the most part good false beliefs. I disagree with Jesus on the value of believing the truth.

No comments: