Monday, October 22, 2007

The Proper Role of Polls

American media has a love affair with polls and surveys. The results are reported as facts with equal weighting to football scores, guilty verdicts, and press conference quotes.

It's not that simple.

Polls are difficult to design well. How you ask questions shades the type of answers you'll receive. It's a challenge to ask the right number and types of people. The results are usually not interpretable as sound bites. Even done with utmost care, polls represent indicators of opinions and mindsets.

The good folks at Stand to Reason have analyzed the recent poll "demonstrating" that "72% of Americans believe that the decision to have an abortion should be left up to a woman, her family, and her doctor" and "69% believe that abortion is 'the taking of human life.'" They have an excellent deconstruction of the results.

And then they conclude with these excellent points about polls:
"Once again, the real value of a poll is to get people talking. It raises more
questions than it answers. ... There's a big difference between polls and
dialogue. Polls ask sound bite questions and get sound bite answers. There's no
chance for respondents to explain or nuance their answers. That's why you should
use poll results to start the conversation, rather than to tell you what people
think. When you ask people lots of "What did you mean by...?" and "Why do you
believe that...?" questions, that's how you really get to know them."

Let's aim for dialogue, men, and put polls in their proper place.

No comments: