[MNJohnnie]

Sample was of 1033 registered voters of which 319 were registered GOP primary voters

Then it is pretty useless polling since Republican voters, not the public at large will chooses the nominee. So what they should run is head to head match ups and poll only Republican voters.

[finnman69]

They did break out the republican primary results in the poll.

[kevkrom]

Sample was of 1033 registered voters of which 319 were registered GOP primary voters

Registered voters instead of primary voters, and a small sample size (319, which yields a 5% margin of error) makes this poll pretty iffy to start with. Since no candidate has shown significant movement since the previous poll and the fairly large MOE, it's pretty safe to say that this poll tells us next to nothing.

I also don't buy that conservatives are split evenly between Giuliani and Thompson -- this flies in the face of every other poll I've seen.