Comments on: Do new studies portend litigation rationality on vaccines?

By: Ted Frank

Ted Frank — Thu, 07 Sep 2006 11:19:03 +0000

PN: I am guilty of using shorthand for a well-documented plaintiffs’ attorney fraud, but in the case of sudden acceleration, there was no evidence supporting the cruise control failure hypothesis (which still has never been replicated), while all evidence pointed to pedal misapplication.

SC: My first thought upon seeing the study was the same thing: could it perhaps be a function of social ineptitude, rather than aging sperm? (I believe Slate made the same point.) Either way for purposes of the vaccine controversy, however, that points to a genetic answer.

By: Supremacy Claus

Supremacy Claus — Thu, 07 Sep 2006 08:37:39 +0000

The study failed to control for paternal social skills deficit as a mild version of the disorder. It could account for having a child late in life.

One easy way to do so would be to provide the rate of autism in siblings born when the father was younger, especially older half-siblings from another mother. Maternal depression is another potential risk factor. The half sibling study would attenuate that factor. Use older siblings as a control group.

Was it the father’s wierdness accounting for both the late birth and autism, or just his age?

By: Peter Nordberg

Peter Nordberg — Thu, 07 Sep 2006 08:03:01 +0000

I am not sure I understand you, Ted, on the sudden acceleration point. If I can show a statistical association between some factor F and some outcome O, I have “demonstrated conclusively” that every instance of O is caused entirely by F?

I ask because I have some briefs I’m writing where this inferential principle, if valid, could come in awfully handy.