Mickey Kaus notices that one of the participants in Frank Luntz's post-debate focus groups on Fox seems to have appeared twice. Scandal? Time will tell. In the meantime, it's an excuse to link to an 11-year-old Andrew Ferguson essay entitled "The Focus Group Fraud." It's a classic:
Focus groups have been widely used in politics for only fifteen years or so. Once an instrument of national campaigns exclusively, they have become popular at every level of electioneering in the 1990s. Consultants will field a focus group for a variety of purposes. They can show campaign ads to test their effect. Pollsters, before conducting a survey, might use a focus group to test ideas about what to ask and how to ask it. After the survey, a pollster can use a focus-group discussion to probe confusing or contradictory results. So what's the problem? Focus groups, wrote the political analyst Stuart Rothenberg in a recent Roll Call, 'are the most misused and fraudulent political technique of the decade.' Rothenberg's complaint is that in all this flurry of activity candidates and consultants forget that focus-group results are not 'projectable' onto the larger population. Polls are scientifically designed to apprehend public opinion; the results, within a margin of error, do tell you something about the thinking of voters at large. Not so with focus groups. The groups are too small. There's no way of knowing that they represent anything more than the opinions of twelve people sitting in a room talking to a moderator in anticipation of making forty bucks.
I know what you're thinking: You can make $40 bucks watching a debate? Sign me up!