WI: Romney Tied with Obama 46 - 46

[Andrew1]

What was the likely voter screen on this poll?

Looking at the "Marquette Law School Poll, May Instrument" on their website, they seem to have used a separate screen for general election likely voters (Question Q2) and recall election likely voters (Question Q2b).

However 87% of their sample said they were "Absolutely certain" to vote in the general election, and 85% of their sample said they were "Absolutely certain" to vote in the recall election.

It seems highly improbable that turnout in the recall will be almost the same as for the general.

Have these figures been altered or weighted in some way?

If turnout will be much lower in the recall, using a recall likely voter screen on a general election poll question is obviously misleading.

(I decided to start posting as I was confused about this poll).

[Andrew1]

Maybe it is just an effect of the political climate in Wisconsin at the moment.

The poll shows a very small enthusiasm gap. Obama leads 46-44 with registered voters and is tied 46-46 with likely voters (apparently that is general election likely voters).

I would expect Obama to be doing best with registered voters, not so well with general election likely voters, and worst of all with recall likely voters.

However this poll implies there would be very little difference between all three, given that it suggests turnout for the recall will be almost as high as the general.