As the elections near we are frequently watching new polls come out, which often affects the way we vote. How accurate are all these polls and how can we tell if they are biased or not?
Thanks,
Jennifer Menzer

Dear Jennifer:

As you might expect, not all polls are equally valid. They vary tremendously as to how many people they survey, how they define a “likely” voter, the way they word questions, and the order in which the questions are asked.  Regretfully, because surveys always make big news, especially “horse-race” polls reporting who’s ahead, all are reported to the public as if they were equally valid.  Reputable polls should make information on their sampling techniques, sponsors, and survey instruments available to the public. (Just a note: Florida is a hard place in which to poll because of its age and language diversity.)

For an excellent overview of “How To Read Political Polls Like A Pro,” go to: 

How to Read Political Polls Like a Pro

If you would like to compare the results generated by different polls in Florida to see how much “swing” there is in the results, there is no better place to go than to the Real Clear Politics website. It tracks all polls done on key races in each state. Currently, this site is tracking polls done in three key races in Florida:

2006 Florida Governor Poll

2006 Florida Congressional District 13 Poll

2006 Florida Congressional District 16 Poll

The biggest problem with polls?  None can accurately predict turnout which is why the polls taken before an election can be wrong.  In Florida, Republican turnout has exceeded Democratic turnout for the past couple of election cycles, especially in non-presidential election years. Republican turnout in the 2006 primary exceeded Democratic turnout.  Year in and year out, It is hardest to predict the turnout rates for independent voters, who now make up over 20% of all Florida registrants.

Get Weekly Deals | Write a letter to the editor | Subscribe and get two weeks free | Place an Ad