Saturday, April 28, 2012

CDC Study of April 27, 2011 Tornado Deaths

Here is the headline:


ATLANTA -- Most of the victims of last year's epic tornado outbreak in Alabama had at least one thing in common: They knew the storm was coming.

With yesterday being the first anniversary of the giant tornado outbreak, many media outlets are reporting a study by the Centers for Disease Control that purports to show that "most" tornado victims were warned.

But, does it?

You never know whether this is faulty media reporting or a flawed study. But, since this is being reported in so many places, lets look at the numbers as reported by MSNBC.

The CDC has been examining reports of 255 deaths, including a few for which no Alabama death certificate has been found yet. It's possible a few people were injured in Alabama but died in hospitals in nearby states, Chiu said.
For 120 of those 255, the CDC determined whether the victims knew of the coming tornadoes ahead of time. And 105 were warned.

If 105 were warned out of 255 deaths, that is 41%, hardly a "majority." Put in context, that is a very low number.

In the case of the Greensburg, Kansas, tornado in 2007 99% were warned.

I suspect the explanation for the low percentage of people knowing the storm was coming was the power failures caused by the two lines of thunderstorms that moved through the region in the hours before the killer tornadoes formed.

2 comments:

  1. "I suspect the explanation for the low percentage of people knowing the storm was coming was the power failures"

    I thought of that when I stumbled across a re-tweet posted by James Spann yesterday. A young woman who lost a friend in one of the 4/27 Alabama tornadoes told Spann that "her last tweet was about you". The friend's last tweet said: "I hate bad weather. Hope someone warns me to go to my safe place if I need to cause the power is out so I can't watch James Spann. :("

    It's only one example but I wouldn't be surprised if there were many, many others in the same boat as this unfortunate young woman...

    Elaine

    ReplyDelete
  2. Mike, I believe they start with a basis of 120, whose awareness could be determined, rather than complete data set of 255. Of the remaining 135, they were not able to determine whether or not they had received the warning. Of course, the headline would have to be much more complicated (i.e., complete). So, while they are correct saying "most" with 105 of 120 (87.5%), it certainly is not the complete story, if not misleading.

    ReplyDelete