What's in a name? That which we call a rose by any other name would smell as sweet. – Shakespeare, circa 1590s.
Earlier this month a study came out of Illinois and Arizona, two states well-known for being hit by hurricanes (!), about how female-named hurricanes cause more deaths than male-named hurricanes because people just aren’t as afraid of women as they are of men (you can find the actual article
here). The study was conducted by a team from these assorted expert backgrounds: business affairs (including marketing), psychology, women and gender in global perspectives, and statistics. I'm sure everyone is an expert in their arena, but obviously this is not a serious scientific meteorological study!
The data is available on the link to the publication above, so I downloaded it and had a little playtime with numbers (woo hoo! :-)). Here are just a few reasons why I think this study is a bit iffy.
1. GENDER OR STRENGTH?
The naming of storms started in 1950. From 1950-1952, storms were given names from the World War II spelling alphabet (Able, Baker, Charlie, Dog, Easy etc.), so they weren’t male or female names. From 1953-1978, hurricanes were ONLY given female names. Male storm names were not used until 1979, after which they were used alternately with female names.
The study used storm names from 1950-2012. For a fair study, storms should only be included from 1979 onwards.
Also, the authors decided to allocate genders to the non-gender alphabet names from 1950-1952, so, for example, they classify ‘Easy’ as a female name (hmm… interesting choice. Freud would have a field day!).
Here are their number of storms per category from 1950-2012 and, using the same dataset, the numbers from 1979-2012:
|
1950-2012
|
1979-2012
|
Cat 1 (F/M)
|
22/14
|
10/13
|
Cat 2 (F/M)
|
15/6
|
8/6
|
Cat 3 (F/M)
|
21/7
|
9/5
|
Cat 4 (F/M)
|
3/2
|
0/2
|
Cat 5 (F/M)
|
1/1
|
0/1
|
(F = female, M = male)
Outlier storms such as Katrina and Gilbert are not in the data, so kudos to the authors for not including those.
In my study of using only storms from 1979 onwards, we see no cat 4 or 5 female name storms – those big scary ones that people really pay attention to. The average intensity of female storms from 1950-2012 was 2.12 compared to 1.96 from 1979-2012. The average intensity of male storms from 1950-2012 was category 2 compared to 2.11 from 1979-2012.
So from 1979-2012, on average, the weaker and less damaging storms happened to be the female named storms.
Generally, people are more cavalier about weaker storms compared to stronger storms, regardless of name. Which would you evacuate for, a cat 1 or a cat 4? This study could really be about how more people die in weaker storms than stronger storms, because they don’t feel as threatened by the intensity. We can put gender aside.
2. CAUSE OF DEATH?
Although they did remove outliers like Katrina (1,833 deaths) and Gilbert (433 deaths), what we don’t know is how the deaths occurred. Were they because people took the storm seriously and died during evacuation? For example, the day before Hurricane Rita made landfall in 2005, 23 people were evacuated from a nursing home, but died because their bus caught fire!
3. STATISTICS ARE GREAT!
Correlation should not be confused with causation! Although a flashy title, this study does not prove that “Female hurricanes are deadlier than male hurricanes”. If that’s the case, then here are a few more amazing facts (Note: I have not personally checked the data, but full credit for source is given below)…
• Did you know that the number of people who drowned by falling into a swimming-pool could be decreased if only Nicholas Cage stopped appearing in films?
• Think there is a need for more Civil Engineering doctorates in the US? Well obviously Americans need to eat more mozzarella cheese to make that happen!
And if it’s more Computer Science doctorates in the US that you are after, then you really need to head on down to your local video game arcade:
• Here’s a useful one: the per capita consumption of cheese (hmm… cheese :-)) correlates with the number of people who died by becoming tangled in their bedsheets…
but the number of people who died by becoming tangled in their bedsheets is correlated with revenue generated at ski facilities…
… so obviously, if ski lodges are having a bad season (no snow, avalanche, summertime etc), then they should increase their cheese-based meals, thus by-passing the bedsheet/deaths thing entirely and saving everyone a lot of bother!
• And, finally, the real reason why honey producing bee colonies have decreased over the past few years is not because of changes to their environment, but because of the increase in the number of juvenile arrests for possession of marijuana. I guess you can only have honey or marijuana, not both.
These graphs and correlations are just a small selection from the hilarious
Spurious Correlations website (check it out as there are more!! :-)). Thanks to Tyler Vigen, who put it together!
So, in conclusion, the real take-home message is that all scientists should have someone from their Dept. of Marketing as a co-author… the media attention, apparently, is amazing!
Oh, and by the way, if you come across a Hurricane Jyotika, be afraid. Be very afraid. Run for the hills. Or even better, run to Illinois or Arizona.
For those of you in in Florida and on the east coast… I’m watching that blob and may be back tomorrow…
Over and out (for now!),
J.
Blogs archived at http://jyotikastorms.blogspot.com/
Twitter @JyovianStorm
-------------------------------
DISCLAIMER: These remarks are just what I think/see regarding tropical storms - not the opinion of any organization I represent. If you are making an evacuation decision, please heed your local emergency management and the National Hurricane Center's official forecast and the National Weather Service announcements. This is not an official forecast. If I "run away, run away" (Monty Python), I'll let you know.
-----------------------------
Posts
