Very few of my same-age friends have a real telephone with a listed number in their home. (I would love to see actual statistics for cellphone-only households by age group.)
I was thinking about this tonight when I participated in a quick political phone survey gauging the prospects of the candidates for Mayor of Boston, seeking vague [meaningless?] opinions on the “direction” of the city, and asking for whom I would vote for US Senate if the election were held today. Surely my age group is severely under-represented in public opinion polls. How well can you actually control these studies to represent the true population? The only personal question they asked was for my highest level of education.
This is one of Nate’s hobby horses over at 538.com. Few pollsters do cell phone polling either correctly or at all, and tend to under-represent the cellophone demographic (who increasingly have zero land lines, as you mention).
But I think they’re starting to do more cellophone polling to correct for that these days.
This complaint would be a sound one, and MR’s confirmation would be valid, were it not for another painful truth: that the snake in the grass is offering up the fruit of polls that concentrate on that demographic anomaly in order to pump up the stats for the side deemed most likely to benefit from it…admittedly, just as their opponents do the opposite. This has been going on for some time and the Voice of the Young has been magnified as a result (keep in mind where the numbers are friends: were the numbers reversed, the justice for the young and for their progeny that your [you might say the more staid among them while I would say prudent] elders have been f ighting for might have been achieved had the young fought along side of them (instead of fighting for an unattainable dream, the result of which has been, for you, the greater likelihood of walking around for the rest of your lives with nothing but ‘change’ in your pockets.
Funny you should bring this up… I’m currently temping at Abt Associates, and they just published a white paper on a similar subject, titled “Compensating for Noncoverage of Nontelephone Households in Random-Digit-Dialing Surveys: A Comparison of Adjustments Based on Propensity Scores and Interruptions in Telephone Service” and I was reading the abstract on my first day. The full text is here: http://www.jos.nu/Articles/article.asp. Though they say right up front that “coverage problems with cellular-only households are briefly discussed in this article,” I assume that finding a solution to either issue would solve the other.