By packaging some nonsense in a click-worthy headline, researchers from Illinois and Arizona have proven the viral nature of the web and that no matter how dubious your claim (“Study Shows Hurricanes With Female Names Don’t Get Taken Seriously”), bloggers will just pass it along. The clicks are what count.

Further spreading their story will be those in radio and in television monologues who repeat what they read on the blogs because this particular headline is too good. As a measure of the blogosphere, these marketing experts have a rousing success.

The problem is that their supposed study is total crap.

As some bloggers have pointed out, hurricanes prior to ‘79 (when they started having male names) were deadlier and if you look at only storms since ‘79, the difference isn’t as pronounced. Others have reported that the same is true for weak storms; there is no predictable difference between male and female death tolls.

Because reality could take away from the conclusion they’re selling, the researchers did a something which sounds good, but is completely worthless. They devised six named storms and asked people whether they would evacuate, flipping the names. Some dude in Des Moines, who has never faced a hurricane and has no basis whatsoever, says he wouldn’t evacuate for a girl. This sounds like this could mean something, so the marketers threw it into the mix.

(Compounding the questions and like the bloggers, I’m not going to pay ten dollars for the study, but my math on their free dataset differs from what’s being reported. When I add up the alldeaths column on male and female storms, I get 24 deaths per female and 15 per male, when I divide by 62 and 29 respectively. Maybe somebody could tell me what I’m doing wrong.)

When one is deciding whether to evacuate for a hurricane, they are basing it on a lot of things. They’re looking at its strength, its predicted path, past storms, how big an impact it may have, the strength of their house or shelter, if they have a support system or contingencies in case something fails and the inconvenience of an evacuation. They are also discussing all of this with their friends and other members of the community over days or weeks, taking it all into consideration. Whether or not to evacuate is not decided in a few minutes and there is a lot of non-weather information at play.


Okay, here’s the two Gawker links (edit +1). Within them are links to other blogs.

Also, I’ll go ahead and point out to anyone who has never faced a hurricane or who experienced the after-effects of one once when they were a kid, who thinks that everybody should just get out of the way: thank you for your thoughts and if you’re ever facing the question, after weighing the risks, you’re free to make that choice. I hope the name’s not part of your math.

Update: Since posting and spending an incredible amount of time justifying my conclusion, Annalee worries that this study could undermine “legitimate” inquiries into gender bias.


Update 2: The comment threads on this post were dominated by a few people obsessed with a single point: the idea that I should read the study, though most had not and neither had thirteen out of the other fourteen who blogged about it on Kinja. I feel reading studies is often the best idea, but it is not always required, especially if you’re not questioning its data and are just writing a quick unpaid post for a general interest blog.

Some of those being the most insistent refused to address the issues I raised, such as asking how counting the number of dead could prove that people unable to evacuate in the past or who had little advance warning would, if the storm had a male name. (Correlation is not causation.) Also, though most didn’t want to or couldn’t defend the now largely discredited study — they were only insistent that I should read it, which they too had not done — no one could explain why someone from a family which had never evacuated going back generations would suddenly run from a storm with a masculine name, let alone, how them not dying alone proved they were better prepared, questions it was widely-reported the study’s authors did not address.


These threads remain, though they have little to do with the post and there is a lot of gloating and high-fiving among themselves, because it happened and as a historic record. I may link to this post in the future when discussing hurricane evacuations and if I were to have dismissed the threads, those who had staked a contrary position would have loudly complained.

Though some were insistent on this single point, there is some good discussion in these threads.

With a few exceptions, the majority of the following comments are lecturing me about reading the study or snarking on my responses. While some do, most do not pertain to the post.