Pre-Holiday Roundup

With the holiday weekend fast approaching, I want to clear out my in-box of a number of interesting items that have been piling up over this past week:

• OnWednesday, NBC's Chuck Todd reported on a voter focus group conducted in Baltimore County, Maryland this week sponsored by the Annenberg Center of the Univ. of Pennsylvania. The group, which was conducted by veteran Democratic pollster Peter D. Hart, will be aired at some point soon on C-SPAN (according to Todd's column). Check the C-SPAN schedule for details.

Todd's take is that these 12 voters (5 Democrats, 4 Republicans and 3 independents) were more interested in presidential candidates who "provided a vision and leadership rather than one who had real-world experience." And they were also "torn between wanting a candidate who provided hope and a candidate who made them feel safe."

His column is worth reading in full, but when you do, I recommend starting with caveat at the end of page 2 as well as a few of my own: Remember, a focus group is not a true random sample of anything. It can tell you a lot about the 12 people in the room, but projecting thoseattitudes on some wider population is inherently risky. And, as Todd points out, attitudes that seem strongly held by focus group participants can bemisleading. He shares an example of a similar group in 1999 that "indicated how potent of a threat Bill Bradley was to Al Gore."

Consider another example involving a December 2003 focus group, also sponsored by Annenberg and conducted by Hart, that convinced columnist Mark Shields that Howard Dean had "established a beachhead" among blue-collar Democrats and independents in Toledo, Ohio.

• The Peter D. Hart that conducts focus groups for Annenberg (and the NBC/Wall Street Journal poll with Republican Neil Newhouse) is NOT the same person as Peter Hart, the analyst with the media watch group Fairness & Accuracy In Reporting (FAIR). The latter Hart has an op-ed piece out this week calling media reporting on early horse-race polls "a complete waste of time."

• On the American Prospect Tapped blog, University of Maryland political science professor Tom Shaller has published a critique worth reading of a recent study by the group Third Way that took an odd approach to analyzing exit poll data from the 2004 and 2006 exit polls (via Stoller & Cillizza).

• Earlier this week, the Wall Street Journal described how the practitioners of political "microtargeting" are "taking their mastery of sophisticated new campaign techniques into the corporate world" (via Sullivan). The small irony is that I have often seen political microtargeting described as an application of corporate data mining to political targeting.

• CBS News has an interview posted today with Ann Selzer, the pollster who conducts the Iowa Poll for the Des Moines Register, about her methods and interpretations of recent data.

• CBS News polling director Kathy Frankovic has a new column on the CBS.com web site. Her inaugural effort - "Trust But Verify" -- provides thoughts on the value of putting poll results into the proper context.

• Finally, ABC News polling director Gary Langer strives to put just context around a controversial question asked on the recent Pew Center survey of American Muslims.

Enjoy the holiday weekend!