A day later, I’ve been pointed to some nice writings but scientists and statisticians. See here: http://pb204.blogspot.co.uk/2015/01/science-by-press-release.html

A few things become clear: 1) the coverage is of the press release, not the actual paper but 2) there are still major problems in the assumptions of the paper. The “luck” appears to be mostly an artifact of the press release and the abstract, though the analysis remains flawed. The lack of social and environmental analysis is definitely a property of the paper. And of course most of the journalists reporting it are innumerate to boot.

Statsguy has a wonderful critique of the article, but winds up writing about lifestyle as a cause for cancer, and I’m sorry, without environment and social analysis, that’s just bullshit. Yes, HPV causes cervical cancer and smoking causes cancer, but what causes smoking? And neither of those things are good explanations for breast or thyroid cancer, whose increasing prevalence appear to result from a combination of changes in diagnostic technology and practice and environmental causes.

So it would be wrong to say cancer in general is “behaviourally” caused. Some cancers are behaviourally caused, though even there, how you could say smoking causes cancer and not also look at policies that promote the tobacco industry around the world, I don’t know. Ditto for the HPV vaccine. If HPV causes cervical cancer and the HPV vaccine prevents the cancer, then it suggests to me the important behaviour is vaccination, not sexual activity.

There’s a whole layer of moralizing that goes on top of the behaviour talk–which is typical of American (and I’m assuming some other) medical culture. But that will have to be for another post.