It may be true -- as my NBC colleague Chuck Todd reports -- that the findings of the University of Iowa poll are similar to what private candidate polls show. (And Chuck likes the open-ended question.)
But the sample size is enormous and it comports more with a sample of adult Iowans than it does with hard-core Iowa Democrats. Two different samples, two different sets of results.
A caveat here might well be that Sen. Barack Obama's campaign is predicated on turning out more less frequent Democratic voters and that a broader sample captures his support more accurately than a narrower sample.
The political scientist who conducts the poll is David Redlawsk.
