Somewhere during the conventions, I got out of the habit of posting these. So for those new to the site, here is a collection of links to "outliers" -- interesting polling related items from the last week:
Mike Murphy thinks televised "dial groups" are madness (a certain blogger agrees).
Mark Mellman explores the disconnect between the way pundits and voters graded the debates.
Michael McDonald analyzes the new voters in North Carolina and Ohio and posts a trove of Census demographic data on the voters of 2004.
Ruy Teixeira posts a trove of his own: analyses of the political geography of OH, Mi & MO, VA & FL, PA, the Intermountain West and a summary presentation.
Nate Silver sees Obama dominating among early voters.
Gary Langer examines the sharpening perception that the McCain campaign is more negative.
Jay Cost ponders the Palin effect and shares his thoughts on Gallup's likely voter model.
David Hill thinks pollsters could do a better job documenting the role that Obama's race plays in this election.
Jon Cohen links to a vintage report on the 1980 election by the late Warren Mitofsky that debunks claims that Reagan trailed Carter by a wide margin in mid October (see also this chart).
Ezra Klein thinks swing voters get too much attention (via Sullivan).
PPP asks you to vote on where they should poll next week, seeks a better list for West Virginia and posts a report on how at how party unity has shifted since the conventions.
And if you haven't yet seen the 3BlueDudes massive mash-up of electoral vote projections, by all means, check it out.