In
explosiveliberal's diary there's been a lot of discussion of SUSA's new Pennsylvania poll which has Kerry ahead of Bush 49-47. The sample included a lot more people who voted for Bush in 2000 than who voted for Gore, so the poll is skewed in favor of Bush.
How much is it skewed? Well, there's been a lot of argument about that. What I've done is weighted the results to reflect the acutal proportion of Pennsylvania voters who voted for Bush and Gore in 2000. These weighted results indicate how many people in the poll would have chosen Kerry and how many would have chosen Bush if the sample had been balanced in the first place.
Results: Kerry 53%, Bush 43%.