tag:blogger.com,1999:blog-8237846547805652402.post5317018360596894698..comments2014-02-03T14:46:43.219-05:00Comments on EP-ology by Carl V. Phillips: Unhealthful News 198 - Reporters think science is Magic: the case of the Iowa voteCarl V Phillipshttp://www.blogger.com/profile/01919902852457771666noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-8237846547805652402.post-27665754244748859422012-01-22T14:19:48.147-05:002012-01-22T14:19:48.147-05:00I am not sure I entirely understand your point or ...I am not sure I entirely understand your point or points. I think you are concerned that the shorthand phrasing tends to obscure the possibility -- that I note in the full version of the statement -- that we are not sure there is any risk at all. I would agree, though the idea is to provide an unbiased (as that term is defined in statistics) summary, so it it also obscures the possibility that the risk might be 2% or 3% of that from smoking. <br /><br />Of course, there is a well-documented tendency for people to homogenize risks and treat zero risk as fundamentally different, so that they consider the difference between 1% and 0% to be much greater than 1% versus 3%, even though the latter is obviously twice as great. So by the scale of common psychology, though not an actual scientific scale, 99% is biased, and even 99.9% would be, because differences among small risks are ignored, while the difference between 0 and anything is exaggerated. <br /><br />On the other hand, since everything poses some risk, the chance of the risk being zero is vanishingly small. It could really only happen if they benefits exactly outweighed the costs (if we are talking about just mortality risk, those benefits would be things like preventing traffic and other accidents, treating deadly depression, and the indirect effects on longevity of greater productivity). Of course, part of the goal is to communicate something that people will believe and that does not play into the liars' propaganda, so it is easiest to just avoid noting that there might be zero risk.<br /><br />That said, when I am talking to people that I know are honestly seeking the truth and are sufficiently numerate, I have no problem just saying "there is vanishingly small risk" or "the risk from switching is the same as from quitting cold turkey" or perhaps even the shorthand "safe". But for way too many people, hearing something like that, even though it is scientifically more defensible than saying the seemingly precise 99% (let alone the seemingly more precise 98% that some people say), will be confusing or will be used dishonestly. So basically I have to reserve frank scientific information for personal conversations, and cannot use it when speaking to groups or writing.Carl V Phillipshttps://www.blogger.com/profile/01919902852457771666noreply@blogger.comtag:blogger.com,1999:blog-8237846547805652402.post-21119290367549602262012-01-21T22:14:47.522-05:002012-01-21T22:14:47.522-05:00But there is a linguistic trick here, 99% less har...But there is a linguistic trick here, 99% less harmful implies harm. Interpretation is subjective. And of course safe is a taboo FDA word. So when FOX says the majority of polls show the majority of people see Newt in a more negative view than other candidates- what a great sea of no information that makes Newt negative. Verbage is everything. p < 0.05.Anonymoushttps://www.blogger.com/profile/00218447592942640268noreply@blogger.com