Is Cell Phone Bias Skewing Polls?

[babylonsister]

http://politicalwire.com/archives/2010/10/13/is_cell_phone_bias_skewing_polls.html


Is Cell Phone Bias Skewing Polls?


A new Pew Research study finds that not only can landline-only surveys exaggerate the Republican enthusiasm because cell-only users tend to vote more Democratic, but they also may be even more biased than in 2008.

In three of four congressional election polls conducted since spring, estimates from landline-only samples had slightly more Republican support and slightly less Democratic support. One of the polls showed no difference. In the most recent congressional poll, Republicans led Democrats 46% to 45%, among landline-only users who were registered to vote. But when cell-only registered voters were included, Democrats took the lead, 47% to 44%. Among likely voters, Republicans held a 12 point lead among those with landlines only, 53% to 41% but the lead shrinks to seven points, 50% to 43%, when cell-only voters are added.

Jon Cohen: 'While Pew's update to their long-running research on cellphones and surveys isn't a broad rebuke to pollsters who don't interview on cellphones, it raises fresh doubts about the precision of the reams of polling data fueling estimates of what may happen on Nov. 2."