Two polls show very different results

[pstokely]

http://www.lodinews.com/news/article_7fbaaa31-7b40-5afd-81e7-f66b2afc0137.html

The first e-mail came to the News-Sentinel at 8:21 a.m. announcing a new poll showing Republican David Harmer holding a six-point lead over incumbent Jerry McNerney in the 11th Congressional District race. The second e-mail came in 40 minutes later. It showed McNerney ahead by 10 percent.

What gives? Can there really be a 16-point difference between two polls?

Absolutely, said David Mermin, a pollster for McNerney, D-Pleasanton, who represents the Lodi area.

[Scuba]

1. Most polls are designed to shape opinion, not measure it....

..."My poll shows most Americans love Republicans. Jump on board!"

[Jim__]

2. The first thing we need to know is the margin of error in both polls.

It may be that the results are actually compatible.

Other than that, most election polling counts "likely" voters, and since every polling firm has different criteria for likely voters, poll results are not always compatible. If both polls were using just registered voters, then the results should be compatible, but registered voters are not good predictors of election outcomes because so few registered voteres actually vote.