Reality check: all trackings except Zogby showed movement towards Obama today

[thesubstanceofdreams]

Obama improved in Rasmussen, Tipp, Diageo (same difference but much fewer undecideds, Obama over 50%), Research2000, and Gallup.
Keep in mind Rasmussen and Gallup have by far the largest sample size, so they are less prone to large random variations (see Zomby).

[NJmaverick]

1. I think this is the result of he Obama infomercial

[GreenPoet64]

2. Zogby's poll . . .

I participated in Zogby's poll. The question about how likely you would be to vote was worded strangely. I had already voted, so I wasn't sure if I should participate in the survey. I almost opted out.

Then after a moment of pause, I decided that my likelihood to vote would be 100% because I had already voted. If you've never had a statistics class, you might not answer the question that way.

In other words: the poll is probably skewed to those who have not voted.