Distorted Polls

I saw a quick blurb on CNN this morning, while I was eating my breakfast, that indicated that their "poll of polls" showed McCain ahead in the polling for President. I was really upset. What I didn't think about, and I should know better, was the way in which polls can be distorted. I took a lot of statistics in college - both under-grad and grad levels. I've even taught statistics to middle-school students.

An "outlier", or one result in a survey that is grossly distorted from the rest, can have a profound influence over any summary. Outliers are often removed from a study when a final summary is done - and sometimes they're not.

Then (thanks Air America for the tip) I ran across an article that actually lays out all of the most recent polls. This doesn't mean we're out of the woods yet - the next round of polls may be as disappointing as I had feared - but for the moment things aren't so dark.

Read the post on outliers at First Draft. And if you don't want to read the whole thing (some of the statistical stuff can be heavy), at least scroll down to the actual poll results all done up in a fine little chart.