On Feb 14, 2015, at 3:17 PM, Leonid Gavrilov wrote:
> Dear Colleagues,
> We are pleased to alert you that our new peer-reviewed study on Predictors of Exceptional Longevity is now publicly available at PubMed Central® (PMC) – a free full-text archive of biomedical and life sciences journal literature at the U.S. National Institutes of Health’s National Library of Medicine (NIH/NLM):
> [ available in PMC 2015 Feb 5 ].
> Comments and suggestions are most welcome!
The abstract says:
> Analysis with multivariate logistic regression found that parental longevity and some midlife characteristics proved to be significant predictors of longevity while the role of childhood conditions was less important. More centenarians were born in the second half of the year compared to controls, suggesting early origins of longevity.
This is the standard “correlation suggests causation” fallacy.
What happens in general is: is people find a correlation and want to claim a causation, but they know they’ll receive criticism if they do because “correlation isn’t causation” (a mantra they find annoying). So they write a claim along the lines of something “probably causes” the thing it’s correlated with.
This doesn’t in any way address the problem with assuming correlation is causation, it just adds some wiggle room if criticized and in practice gets most critics to leave you alone. But it’s arguably actually worse, because it’s more of a meaningless claim – it’s a retreat away from bold claims, saying less to try to reduce criticism.
The fact is, the data is logically compatible with many many other conclusions. It can be explained many ways. Why does it “suggest” this conclusion over the others? And even if there is some good reason, why state a bad one and not mention the good one?
This is all well known.