Correlation

But causation?
Venker goes on to explain that of CNN’s list of the “27 Deadliest Mass Shootings In U.S. History, only one was raised by his biological father since childhood.

“Indeed, there is a direct correlation between boys who grow up with absent fathers and boys who drop out of school, who drink, who do drugs, who become delinquent and who wind up in prison,” she writes. “And who kill their classmates.”
It's bad news if so. We've been talking about fixing failing families since I've been alive, and the problem has not improved outside of those wealthy and stable elements who were in the least danger to begin. Our culture has turned aside from family, even though family is the source of much of -- and much of the best -- human meaning.

6 comments:

Wickedstock said...

Thanks for all the information..

E Hines said...

Meckler--and Venker--both carefully elide a couple of salient dimensions. (I'm eliding the fact that Meckler's bit is a year-and-a-half old.) One is the number of sons with fathers present who go on to commit this sort of crime--or any other. The other is the number of sons with no fathers present who do not go on to commit this sort of crime--or any other.

And a bonus third: the journalists carefully cherry-picked only the most extreme cases--which by definition are not close to the typical case. Having carefully excluded the bulk of the population of cases and deliberately done a non-random selection, there's no way to tell whether these cases are outliers (except "27 deadliest" strongly suggests that), or if outliers how statistically influential they are.

With those data we could say whether, for instance, the number of murderous fatherless sons is small or large, which would more intelligently inform solution proposals. With those data, we also could better assess whether the presence or absence of a father matters to major criminality in any significant way--and that would better inform solution proposals, too.

But facts might interfere with the pre-written narrative. It's an example and extension of what a stat professor said once (he was joking; I doubt the extension is nonserious): if your theory requires a linear relationship, collect two data points.

I said "carefully" above because surely these journalists' interns pointed these things out to them when those interns did the journalists' research for those pieces.

Eric Hines

Assistant Village Idiot said...

I think fixing families is a good thing for its own sake.

The difficulty with quick gatherings of information like this is that people who leave their children or get kicked out have on average worse genetics along many measures. Plus the person who chose to marry them and the provided reasons they were left or kick the other out also have worse genetics on average. There are plenty of those who have decent genetics, of course, and plenty of the parents who stayed together who provided bad genetics anyway. But the numbers are going to trend toward pathology.

Grim said...

Perhaps, but there's no reason to be fatalistic. At one time families held together in communities of various genetics, or else we wouldn't be noticing the dissolution. Daniel Patrick Moynihan wrote about this dissolution in the black community a generation ago, which shows that he was noticing a change. That change has spread to the community at large. The genetics have not changed very much, but the pathology (as you put it) has spread.

That is cause for concern, but also for hope: a door that swings one way often swings the other way, too.

Ymar Sakar said...

Look at lds and amish.

douglas said...

" the number of murderous fatherless sons is small or large, which would more intelligently inform solution proposals."
I don't think that's whay they're attempting to find. IT seems to me we know that as these events are extremely small and rare to begin with, there isn't *any* category of people likely to have a significant number within it's ranks, so that's not what you should look for. We then have no choice but to look for common attributes and see if we can determine if they are salient or not, or have some nexus that creates a condition necessary to the development of one of these personalities.