Response rates nosedive for polls
Jim Gerraghty of NRO's TKS blog highlights the declining rate of response to telephone polling on political races. If you're only completing a shade less than 21% of your calls — way, way down from the traditional 30% response rate threshold pollsters traditionally enjoyed — and then only a little more than a third of those who did actually answer the phone then agreed to participate, just how safe is it to assume the other 90% of the population who have opted out of participation holds the same opinions? And this doesn't even get into the issue of whether the small segment reached were all truthful in their responses and the every growing number of people who are totally unreachable because they rely 100% on cellular services.
Rosslyn Smith 10 15 06