Tuesday, May 22, 2012


Some very smart and cautious people don't pay any attention to polls until the week before election day.

Others of us---political junkies all---cannot leave a poll unread or undiscussed. But I contend that setting limits and boundaries can legitimize even the most warped and biased public opinion sampling revelatory.

For instance, some polls require only that the sample have a warm "adult" body.

The better, more professional surveys limit their queries to voters, and even refine that characterization down to "likely voters". Scott Rasmussen's inquiries like the Mills of the Gpds, grind "exceeding fine".

Rasmussen very rarely accounts for responses from less than 1000 people, always made up of what I call "habitual voters" who have exercised their franchise in the immediately past national, state, local or municipal election--- whichever is nearest past.

No other pollster that I know of cuts it this fine.

But some polls deserve the growing disdain of the consumer.

Here's an example why:

The last national election in the United States (2010) , for the first time in years showed an almost even turnout split between the GOP and the democrat party in participation by party-of-choice--35% and 35% and 18% "independent".

But here in 2012, all of the pollsters except Scott Rasmussen, when forced to reveal their "internals", indicate that they are still "weighting" their sampling percentages on the basis of the 2008 national election, which broke 37%-25% in favor of the democratic party by self-identification.

That is akin to giving an athlete a 12 yard (or meter) head start in the 100 sprint.

At the very least it would make any "margin of error" completely meaningless.

I'm addicted to this sort of information. I'm sure most people don't go this far.

But a great Atlanta homicide detective of the 1950s, Louie Hawse, abjured me to obey one rule in any investigative reporting ---or research--- that I attempted.

"Look under the rock, Bill," the would wheeze.

"Then, after you looked under the rock, turnover the rock underneath the rock and look THERE ! And then....."

Well, you get the idea.

But I'm sure most people will just wag heads at polling results, or maybe even just wait until after the October surprise to begin reading the polls.

There is plenty of entertainment around. You might even buy some FACEBOOK stock.

No comments: