tag:blogger.com,1999:blog-7085195159661027971.post1038782398151315592..comments2024-03-15T16:14:36.387-04:00Comments on Just the social facts, ma'am: The Story of the HurricaneDavid Weakliemhttp://www.blogger.com/profile/02336229317604663975noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-7085195159661027971.post-88533934886021836442014-06-15T15:43:13.760-04:002014-06-15T15:43:13.760-04:00The data set doesn't include that information,...The data set doesn't include that information, so fortunately I can't spend time checking that. However, given the nature of their model, there's a good chance that the interaction of anything with storm damage would have a "significant" effect. <br /><br /><br />As far as the general idea of the study, there is a lot of experimental evidence that factors that should be irrelevant if people were rational actually can affect perceptions. Then it's reasonable to suspect that the irrelevant factors might make a difference in the real world. The reason I was skeptical is that there are so many influences on hurricane deaths that if there's any effect it should be small--maybe a fraction of a percent. Any effect that would show up in 92 cases would have to be really big. David Weakliemhttps://www.blogger.com/profile/02336229317604663975noreply@blogger.comtag:blogger.com,1999:blog-7085195159661027971.post-85486196544510125882014-06-13T13:46:57.576-04:002014-06-13T13:46:57.576-04:00As Krugman (sorry, Kristof; sometimes they are dif...As Krugman (sorry, Kristof; sometimes they are difficult to distinguish in their obtuseness) writes, the takeaway from the study was supposed to be that female-named hurricanes caused more damage than male-named hurricanes because our patriarchal and sexist society doesn't take women as seriously as men and so underestimates the likely impact of a hurricane with a female name. You're generous in having assumed that there was something wrong with the study; when I read the initial flurry of news reports about it my reaction was that the theory was so insane that it was not even wrong. <br /><br />Since you have the data downloaded and ready for analysis, I'm curious. How does, say, the day of the week on which the hurricane made landfall (or, better yet, the date of the month, since everyone in hurricane country can be reliably counted on to be in church on Sunday) come out as a predictor compared to the femaleness of the name?Earle Millernoreply@blogger.com