The Problem With Telephone Polls
Wednesday - October 13, 2010
I’m getting those evening phone calls again - pollsters, people who want to know whom I intend to vote for, or how I feel about certain hot-button issues or claiming they’re just doing some general survey.
My standard response is, “Sorry, I don’t do phone polls. Goodbye.” Click.
What occurs to me is the possibility that many people in my age group, education background and general dislike of interrupting phone calls hang up. That could mean that 500 folks who enjoy talking to strangers on the phone create the results.
I’m not trying to put the polling firms out of business, but I’d like to plant the seed of doubt that telephone polls produce anything more than results from folks who say “sure, I ain’t got nothing better to do this evening than chat with you, so go ahead, fire away with your questions.”
A sample size of at least 500 responses is a general requirement for a political poll - responses, not call-outs. And as a national poll-ster explains it, “Since some people do not answer calls from strangers, or refuse to answer the poll, poll samples may not be representative samples from a population due to a non-response bias.”
Another pollster admits that “we simply don’t answer our phones like we used to. We have caller ID and caller-specific-rings.
When pollsters call they are treated as any telemarketer or unknown caller would be, thus the people who pollsters actually get to talk to are becoming increasingly less representative of the general public.”
The Harris Black International polling agency says refusal rates routinely exceed 40 percent of all households. Unreachable respondents (due to traveling, working, answering machines, the absence of telephones in college dorms, etc.) can regularly run another 30 percent of the sample.
But let me go back to my starting thesis - the kind of people who do participate in telephone polls. I’ll be skewered for elitism or worse, but I’m betting the sample here runs highly low-education, Caucasian and Hawaiian, Republican (always hoping for a revival) and ultra-religious (hoping for a Duke Aiona or a Cam Cavasso). Older people of Japanese ancestry have tended not to share opinions on the phone or even in face-to-face surveys. The 18-30 group often will, but also may not vote.
I’ll keep hanging up.
I scored zip in my Primary predictions. Mufi got crushed and my only saving observation was: “If Hannemann should lose this one, blame that silly brochure ...”
I blew the Carlisle-Caldwell call, too. Should have asked my more-politically akamai wife about that one.
I overestimated Norman Sakamoto, who came in third in the LG race. I figured a hearty second at least.
So I’ll not predict the General, except for two sure things: Dan Inouye and Mazie Hirono will be reelected. Beyond that, this chastened columnist sayeth zilch.
I erred in my Sept. 22 column saying former Army Maj. Gen. Edwin Walker committed suicide. A lifelong heavy smoker, he died of lung cancer.
E-mail this story | Print this page | Comments (0) | Archive | RSS Comments (0) |
Most Recent Comment(s):