Wikipedia suggests this example: “The use of fertilizer improves crop production on farms and in gardens; but at some point, adding more and more fertilizer improves the yield less per unit of fertilizer, and excessive quantities can even reduce the yield.” Too much manure can ruin the entire garden. I am beginning to think that too many public opinion polls are smothering the political process with mountains of manure. Can’t you smell it?
I think polls are becoming less predictive of voter behavior. I don’t think it is because the science of public opinion polls has gotten worse. You would think that technology has improved the statistical analysis. I think it is because the sheer volume of polls and the infinite number of variables at work have created a diminishing return. The more frequently they poll, the more companies that conduct the polls, the more segmented the results of the polls, the less likely they are to have any objective meaning. On top of that, I am sick of hearing about the results knowing that their ability to predict future events becomes a self-fulfilling prophecy at some point. Enough!
This particular poll put me over the edge. Here’s a poll that I know in my heart must be flawed. I know it is flawed because to believe the results is to accept that ignorance is an American Value:
According to an automated survey by Public Policy Polling (PPP) taken this weekend, 52% of Mississippi Republicans and 45% of Alabama Republicans believe that Barack Obama is a Muslim. Only 12% and 14% respectively believe that he is a Christian.
First of all, who cares? The Congress shall set no religious test for office.
Second of all, I would wager a guess that many of these same voters who believe that Obama is a Muslim are equally upset that he spent 20 years listening to Rev. Wright in a Christian church. He’s one open-minded Muslim if he went to a Christian church for 20 years just to fool some folks in the future if he happened to beat the odds and become President.
In the same poll, 29% of Mississippi Republicans believed that interracial marriage should be illegal, and Alabama, that bastion of liberalism, had only 21% of the GOP electorate that believed it should be illegal.
I know that my roots are in the Union, but can this be possible in the year 2012? Maybe the Mayans were right. Maybe the world is ending this year.
The good news is that automated polls are less reliable than live interview polls, and the margin of error is plus/minus 4%. I would sleep better at night if I knew that the margin of error was plus/minus 29%, but perhaps it is. The poll didn’t ask registered Democrats the same questions and perhaps they are also against interracial marriage and believe Obama is a Muslim. Maybe these results are a net improvement since 2008, and the trend would show people becoming more tolerate and aware. Maybe Keith Olbermann fixed the results to embarrass the GOP. Maybe...
These poll results can be explained. People lie to pollsters. People respond differently to questions depending on who asks the question and how the question is phrased.
I bombed out of Probability in college so maybe I am wrong. Maybe these surveys accurately reflect modern America in Mississippi and Alabama. If so, that smell isn’t manure coming from down South. It’s ignorance.
Polls in moderation can provide meaningful benchmarks and help us determine what matters to Americans. Polls in excess exist to reach a singular goal – sensationalism, shock, and profit for the polling organization.
It’s the law of diminishing returns. The more I hear these and similar polls, the less I believe them.
No comments:
Post a Comment