Sure a 10 % risk–that is in an observational study–some basic rules for evaluating the pile of crap.
Besides, I thought that the melanoma experts were convinced it was UV ray exposure. Jon Samet and the ACS probably have some survey study association of melanoma with tobacco use, even second hand smoke exposure. Research is so much more rewarding when you can cheat for a chance at publication and ignore the rules.
1. Statistical Significance is only a test of statistics and chance of error–not plausibility or causation.
2. A 10 % risk is a relative risk of 1.1–nothing close to the robustness required for observational studies to claim any evidence of an association that supports an assertion of causation.
In other words, if the editors had any sense they would say–your study shows nothing, a small association could be the product of randomness, methodology. selection bias, recall bias in an intake survey–all would extinguish a 10% increase in rates.
I would also question whether this study is the product of a multiple inquiry survey and if the researchers did the Bonferroni Claw, but I know they would spurn my question as just a trouble maker’s remark.
This study makes me ask questions, like most of the Harvard Nurse Health Study cohort studies.
The Nurses Health Study, tended by Harvard and the notorious Ascherio and Willett junk science team–should pull down their shingle–they are cargo cult scientists. Phony baloney but they claim their methods mean something.