Thursday, September 08, 2005

The Problem with Polls

CNN published a poll in which it says that,

"Fifty-six percent of 609 adults polled by telephone September 5-6 said they believe the hurricane devastated the city beyond repair. And 93 percent of poll respondents said they believe Katrina is the worst natural disaster to strike the United States in their lifetime.

But a majority of respondents -- 63 percent -- said they believe the city should rebuild. And 66 percent said they believe all New Orleans residents should evacuate the city.

Opinions varied widely, however, on the response of federal, state and local officials regarding Katrina. Forty-two percent of respondents characterized President Bush's response to the disaster as "bad" or "terrible," while 35 percent said it was "good" or "great."

Federal government agencies' response was described as "bad" or "terrible" by 42 percent, and "good" or "great" by 35 percent. State and local officials' response was described as "bad" or "terrible" by 35 percent and "good" or "great" by 37 percent.

Respondents also disagreed widely on who is to blame for the problems in the city following the hurricane -- 13 percent said Bush, 18 percent said federal agencies, 25 percent blamed state or local officials and 38 percent said no one is to blame. And 63 percent said they do not believe anyone at federal agencies responsible for handling emergencies should be fired as a result."

These are some of the most inconsistent statements I've seen. If 56% believe that New Orleans is "beyond repair," meaning that it cannot be repaired, how can 63% be in favor of rebuilding it? Perhaps the definition of "beyond repair" is a little vague. Suppose that "beyond repair" means that it would cost more than it is worth to rebuild. In that definition we have also negated reason to rebuild, so we should not expect more than 44% to be in favor of rebuilding. How can we explain the 19% difference? Or by "beyond repair" are we simply differentiating between repairing and replacing? If that were the case, it would seem fairly obvious that many homes would be more costly to repair than to replace; were that the universally accepted definition of "beyond repair" I would have expected more than 56% of respondents to have answered yes.

Maybe more than 63% are in favor of New Orleans being rebuilt -- but which parts of the city and by whom? The article said that the question was phrased, "should the city rebuild?" By "city" do they mean the government? individuals who resided in New Orleans? private companies? charities? A different set of answers might have been obtained had the question been more specific.

The conclusions are virtually meaningless without background on why the respondents chose the responses they did and what they meant by their responses. We can neither explain nor understand the responses to polls like this one without also knowing the reasoning that holds the responses together.

No comments: