Friday, August 1, 2008

Polls apart

There has been some discussion the last couple of days about the reluctance of mainstream media outlets to publish the results of automated polls, despite their fairly solid track record (in particular, SurveyUSA seems to do a pretty good job). The Numbers Guy at the Wall Street Journal has a post and a print column about this, on the reluctance of existing pollsters and the MSM to embrace results generated by interactive voice response (IVR) systems. For example, a CEO for an existing face-to-face polling company said: “While other firms do use IVR polling and claim great success, Strategic Vision believes that nothing can replace a live caller. The live caller can detect nuances in answers, can better screen for age, and have interaction with the interviewee.”

Now, in general, I'm not a big fan of IVR systems, given their comprehension problems and sometimes laborious navigation, I'm not certain they have a huge disadvantage in the world of polling. My story: When I was about 20, a pollster came to the door of the family manse and started asking questions of my mom. After about five minutes, Mother lost interest in the whole thing, called me down to the living room, and had me finish the 30 minutes of questioning. So, please, don't try to tell me that the live caller or visitor can necessarily do anything better (and are these callers really trained to properly "detect nuances in answers"? Sounds like an invitation to inaccuracy to me.).

No comments:

Clicky Web Analytics