That Pie Chart Sure is Purty

On Monday, the El Paso Times ran a story with the headline "Richard Wiles preferred heavily in race for sheriff." The story led off with this graphic:


Isn't it a pretty picture?! Riddle me this - what does the fine print say? Click here if you can't read it. Here's a summary: 350 likely voters were polled with a +/- 5.5%. Here's what I have to say about that -

1. The 2006 population (estimated) of El Paso county is 736,310. Simple math shows the sample size for this poll was only 0.05% of the entire county.

2. Plus/Minus 5.5% - You've gotta be kidding me!

3. The Reuel Group (the firm responsible for administering the poll) allows users the option of creating customized voter lists, totally negating any type of "random" sample that may occur. Or at least throwing the concept of "random sampling" into a realm open to criticism. And, while I cannot confirm, it is rumored that the group themselves have publicly endorsed Wiles. That stench you smell is what I like to call "conflict of interest."

4. Does anyone else see a problem in polling likely voters as opposed to registered voters? Here's why they have to print it that way - you see, these polls are conducted via automated telephone calls (you know, the type that say "To elect a douchebag, press 1. To come to your senses, press 2.") Yeah, my 2 year old cousin can pick up a phone and press numbers. The methodology also causes one to ask, "Which candidate did they chose to be option 1?"

5. In the body of the story, the Times reports that if you removed the percentage of voters who chose "Don't Know" Wiles would have the support of 56% of the remaining people polled. Do you want to know what that math equals? 145 people which equals 0.02% of the county population. I wouldn't necessarily brag about that.

Let this be a lesson to you folks - pretty pictures don't necessarily make the best things on which to base your opinions. Unless you want to trust a second rate news source. And then you have to remember - you get what you pay for.