Monday, July 9, 2012

Political polls…good, bad and ugly…. has released yet another poll with this headline:

"Majority feel Obama has changed country for worse."

For one thing, I wish pollsters would stop doing this kind of thing. To ask "Has President Obama changed America for better or for worse?" of course is technically possibly, and there's nothing wrong, grammatically or logically, with the question. And of course, you will get a result, umpty-ump percent say "better" and bippity-boop percent say "worse."

But here's the rub: the responses are mostly meaningless. The question essentially requires that each respondent, instantaneously, without aforethought, must construct a personally relevant timeframe and a comparative frame of reference for her top-of-mind answer. In other words, every respondent is more or less answering a different question.

When says, based on its July 5 poll, that 56% of "likely voters" think that President Obama has "changed America for the worse," we can agree that 56% of voters actually said that, but we can't agree on what they meant when they said it. To figure out what they meant, you'd have to write a book. These highly generalized "tracking questions" make for dandy headlines, but they don't generate good data.

And by the way, it's more or less standard for pollsters to refuse to disclose how they qualified "likely voters" in their survey. You don't know how TheHill defined "likely voters," and maybe you'd agree with their method, and maybe you wouldn't….in any event, "likely voters" don't vote in November, only actual voters will vote in November.

And also by the way, here's a little gotcha for I did consumer marketing and opinion polling for 30 years in my profession. Choosing July 5 to do a one-day sample of 1,000 people is goofy -- too many people not in the places they usually are (e.g., on vacation), doing things they don't usually do, with or without phones nearby….in other words, too many people not accessible in the usual circumstances to answer questions from the telephone interviewer. No matter how Pulse Opinion Research weighted the sample for's survey, they didn't get a representative group of Americans for this poll. They got some dandy headlines, but they didn't get good data.

And finally, by the way, I'm very disheartened to read this result from the same poll by TheHill: younger voters, ages 18-39, are less likely than older voters to believe that the choice between Obama and Romney is "very important" for the future of the country. If this truly represents the current mindset of younger voters, it's seriously a disaster for them and for everyone. I'm not a young voter, but I know that young voters are going to be dealing with the consequences of our current political choices for a long time, and long after I'm gone. They should be stepping up to take a much more active role in shaping the future they're going to have to live in.

No comments:

Post a Comment