Do You — *Achoo* — Support Health Care Reform?

People who see someone sneeze are more likely to tell a survey-taker that they disapprove of the current health care system. Researchers sent a survey-taker into a downtown mall, where she coughed and sneezed before handing her survey to several passers-by. She was careful not to sneeze when administering the same survey to a second group. The sneeze witnesses were more likely to disapprove of American health care. They were also more likely to say they were at risk of having a fatal heart attack or accident or succumbing to a murderer. We know polling results are sensitive to the wording of questions. The delivery of those questions could be a factor, too. We’ll know for sure when we see the first health care push-poll featuring sniffling, sneezing pollsters. [%comments]


I believe it, but it sounds like a pretty small sample set to be drawing conclusions from.

Howard Tayler

If the conclusion is "let's run some more surveys to test this new hypothesis" then the original data-set is about the right size.

Forrest Samuels

It sounds like you just gave FOX News some ideas on how to conduct their next health care survey, with a sneeze.


I can't decide whether I want this to be true or not. On the one hand if this is true, I feel like it would invalidate every survery ever conducted. But on the other hand, it would be really cool if it were true.


I wonder how many more people decline to take the survey and/or run away from a person showing signs of a contagious illness, biasing the sample to begin with.


#3 - I think you mean MSNBC. Fox would make sure their surveyors would actively NOT sneeze, since they are anecdotally anti-health reform.


You wouldn't believe the number of robo-polls I received in the last week where the computer generated voice pretended to have pneumonia.

Bruce Hayden

This is why I love Freakonomics. Scratching the surface and going, this is weird. Does it mean something?


I didn't know about this, but I'm not surprised. I'm even less surprised by your correct observations that both wording and delivery can affect survey results.

Many years ago I worked a year or so for a telephone polling outfit. The wording wasn't my concern; professionals prepared the questions and the rest of the script.

I was startled when my boss told me she was going to assign me a really difficult survey, difficult because the target group was high-powered executives who resisted taking 45 minutes to an hour to do a survey. The survey was giving people assigned to conduct it fits, the higher-ups were getting impatient, and the company that commissioned the survey was pouring on the pressure.

Desperate, the manager in my branch set up a secret survey. That is, he arranged for people to listen to us, the surveyors, conducting surveys and to rate us. We knew nothing about it -- it was eavesdropping, but not in any nefarious sense. Anyway, he identified a subgroup of us who got way more votes than others for "sounding like someone I could talk to an hour, even for a survey." (I have no idea why I got chosen; I think I sound like a bullfrog with a sore throat.)

But we got it done.


maria wang

no. if it is good, why in a hurry to go through without listening to all people to make it really good.