I'd need to see the breakdown of how AP conducted the poll before I put much stock in this, given that, in at least one incident in the past, AP has put forth polls that were skewed

First, only 81% of respondents were even eligible to vote, and there's no indication of how many of them actually went to the polls in 2004..



1. Party Leanings - The poll is slanted 52-40% towards Democrats, even though the voters in the 2004 election were split evenly at 37% between Republicans and Democrats.



2. Religion - Next, a whopping 19% of respondents had "no" religion, while in 2004 only 10% of voters had "no" religion, and they voted overwhelmingly for Kerry (+36%).



3. Age of Respondents In this poll 31% of the respondents were between 18-34, even though the 18-29 year olds (a slightly smaller demo) only made up 17% of the electorate in the 2004 election. I think it's pretty safe to say that by including 30-34 year olds that number would still not have come close to the IPSOS sample.



4. Income Level of Respondents - This one is amazing. In this poll 15% of respondents made under $15,000 per year. In 2004, only 8% of voters were in this income bracket, and voted 63-36% for Kerry.



5. Marital Status - In this poll, only 56% of respondents are married. In 2004, 63% of voters were married, and voted 57-42% for Bush.



6. Geography - In this poll, only 17% of respondents were from "rural" areas. In 2004, 25% of voters were from rural areas, and voted 57-42% for Bush.



7. Race - In this poll, there were 71% white respondents and 12% Hispanic respondents. In 2004, 77% of voters were white, and only 8% Hispanic. Bush won the white vote 58-41% and Kerry the Hispanic vote 53-44%.



Given AP's history of slanted one really has to wonder if the purpose of conducting it was to get good results, or push an agenda.