Quote:
Originally posted by ltl/fb
If we accept your hypothesis that RR under something greater than 2 is negligible, then it might be irresponsible not to publish because if not published, other researchers won't know that the effect of the variable studied is negligible. If they don't know it's already been studied, they may decide to study it. This is a waste of research resources.
Also, telling people that the increased risk is negligible can debunk myths about stuff.
|
What is the basis for saying RR under 2 is negligible? If the findings have a high degree of statistical certainty, why shouldn't one care even about a 10% increase in risk, if the risk is sufficiently costly? All of these calculations are meaningless without recognition of the costs of both the regulation and the harm. What are the costs of banning smoking in many public places (they do exist, for sure). What are the costs of not banning smoking in public places if we are confident that doing so will reduce the incidence of lung cancer by even 10%?
