Disecting a Typically Flawed Poll

Notice some of the flawed thinking in this, a typical "General" Congressional election poll.

1. Gerrymandering is not taken into account. The vast majority of districts in America are Gerrymandered, which means there will be no real contest.

2. The poll only asks registered voters, not likely voters. Since about half of all potential voters will not vote, this is an important mistake by these pollsters. When a poll is taken of likely voters, you usually see a much better result.

3. Historically, these generic polls undercounts the actual Republican turnout. Kind of like how the exit polls in 2000, 2002 and 2004 undercounted the Republican turnout.

Thinking ahead to this November’s elections, if the congressional election were held today, would you vote for the Democratic candidate in your district or the Republican candidate in your district?

	Aug. 29	Aug. 9	Jul. 2006
Democrat	48%	48%	42%
Republican	32%	30%	34%
Other / Not sure	20%	22%	25%

Source: Opinion Dynamics / Fox News
Methodology: Telephone interviews with 900 registered American voters, conducted on Aug. 29 and Aug. 30, 2006. Margin of error is 3 per cent.