Sanders: of polls, pollsters and their critics
VS 
Arkansas Poll - Zogby Poll
It has been the biggest story of Arkansas' early political season, sans the candidate shuffle: Which poll had which gubernatorial candidate up and by what margin? The University of Arkansas' Arkansas Poll had Democrat Mike Beebe up and the Wall Street Journal/Zogby International Poll had Republican Hutchinson up. Both Democrat- and Republican-leaning blogs erupted with criticism of the polls that didn't reflect well on their preferred candidates. Last week, this column devoted space to problems it saw with the Arkansas Poll's methodology related to the oversampling of Democratic-dominated congressional districts. This cast doubt on the validity of the poll's sample, which risks the accuracy of its results. The statewide daily followed with a story where the respective polls' authors - Janine Parry for the Arkansas Poll and John Zogby for the WSJ/Zogby Poll - explained their polls and openly criticized the other's methodology. That threw more fuel on the fire, leading the blogs to move from criticizing the poll results to criticizing the pollsters.Were there any winners or losers in the poll fight? The winner from this writer's perspective was John Zogby, the internationally renowned pollster, who despite the local criticism of his online sampling method, predicted the Democratic wins in New Jersey and Virginia governors' races. The loser was UA, which arguably lost its status from as the definitive measurement of Arkansas public opinion. The UA likely will be ready next time with a poll that will withstand critical analysis. If not, we'll be treated to another merry-go-round of criticism.
It has been the biggest story of Arkansas' early political season, sans the candidate shuffle: Which poll had which gubernatorial candidate up and by what margin? The University of Arkansas' Arkansas Poll had Democrat Mike Beebe up and the Wall Street Journal/Zogby International Poll had Republican Hutchinson up. Both Democrat- and Republican-leaning blogs erupted with criticism of the polls that didn't reflect well on their preferred candidates. Last week, this column devoted space to problems it saw with the Arkansas Poll's methodology related to the oversampling of Democratic-dominated congressional districts. This cast doubt on the validity of the poll's sample, which risks the accuracy of its results. The statewide daily followed with a story where the respective polls' authors - Janine Parry for the Arkansas Poll and John Zogby for the WSJ/Zogby Poll - explained their polls and openly criticized the other's methodology. That threw more fuel on the fire, leading the blogs to move from criticizing the poll results to criticizing the pollsters.Were there any winners or losers in the poll fight? The winner from this writer's perspective was John Zogby, the internationally renowned pollster, who despite the local criticism of his online sampling method, predicted the Democratic wins in New Jersey and Virginia governors' races. The loser was UA, which arguably lost its status from as the definitive measurement of Arkansas public opinion. The UA likely will be ready next time with a poll that will withstand critical analysis. If not, we'll be treated to another merry-go-round of criticism.
