Slog News & Arts

Line Out

Music & Nightlife

« "Better Transportation Than Yo... | Shakesville No Longer Shakin' »

Monday, June 25, 2007

Polling and Cell Phones

posted by on June 25 at 17:48 PM

Back when I didn’t have a cell phone, I doubted the vaunted cell phone effect had much impact on polls. But these days I’m starting to think you really have to look at a key stat to figure out how representative a poll is of the general population.

Take this survey of Washington state residents, commissioned by the DOT as part of the update of its bicycle and pedestrian plan. (Via Seattlest). We’re talking bikes, right? The average age of respondents to the poll (which was conducted via telephone in April) was 51. 51! Look, I don’t expect 65+ year-olds to tackle Seattle’s notoriously hilly terrain. And the state shouldn’t be able to deprioritize bike amenities because ridership drops off after a certain age. The least a poll could do is try to get the average age of respondents down to around the age of the average Washington adult. Late thirties, I’m guessing? A poll of the older Washington residents who hang around their land lines just isn’t necessarily representative of this state’s commuters. (Voters, maybe.)

RSS icon Comments


It's not just cell phones, but it's caller ID as well. Who knowingly answers calls from "unavailable" or "out of area" anymore? I certainly don't.

Posted by Eric | June 25, 2007 6:30 PM

Without seeing the survey methods (a quick scan of DOT and the polling firm's websites didn't yield anything), I'd be hesitant to criticize. But if they didn't weight the results by age or any other factors, it would seem to be a big flaw. The reason excluding cell phones from political (and other) polling isn't an automatic credibility problem is because the data you *do* get from landlines can be weighted, and people with only cell phones are generally similar to people with landlines *in the same age (or other weighting) category*. As cell phones continue to gain in popularity, at the expense of landlines, those differences may/will become bigger, and cell phones will need to be incorporated into telephone surveying.

So be ready to receive more and more survey calls on your cell phones. And answer the questions from legitimate surveys! It's important for validity of real research that people participate. It sucks that telemarketers have hurt legitimate survey methods so much.

Posted by skinny | June 25, 2007 6:38 PM

I'm one of the few people I know who still have a land line. Mostly because I don't want to "burden" my 84 year old mother with a new phone number (I've had the same one since 1990 or thereabouts, and I'd never hear the end of it if I were to change) but also because I have a turquoise rotary wall phone in the kitchen that makes me feel like a 60's housewife when I use it. It rings, and suddenly I'm Laura Petrie....

Posted by catalina vel-duray | June 25, 2007 7:09 PM

Unfortunately, your last line is right. The average voting age in Seattle municipal elections is probably close to 50. Cell phone or not, they're the ones policy makers want to poll most.

Posted by Trevor | June 25, 2007 8:05 PM

"Look, I don’t expect 65+ year-olds to tackle Seattle’s notoriously hilly terrain."

I'm 58 and cycle up 400+ foot Capitol Hill from Lake Union every day (as does my 60+ year old boss). We manage it because we're not slaves to the single-speed fixed gear bike fashion. We have GEARS! I fully expect to carry on doing so till I retire. And then some.

All this despite the Eugene, Oregon timber executive in his Ford LTD, in 1980.



Posted by Abdrew | June 25, 2007 9:02 PM

I have 2 land lines at home (self employed with a home office). But I refuse to have anything to do with surveys. With push-polls and sales pitches designed to sound like a poll, I'm highly skeptical of anyone claiming to be conducting a survey. I also hate that so many politicians seem to make decisions based on the winds of the latest poll. When someone tells me they are conducting a survey, I just hang up.

Trying to poll cell phones is problematic. By default, land-line phone numbers are publicly available, unless you pay extra for an unlisted number. A polling company (or telemarketer) can easily buy lists of land-line phone numbers. On the other hand, cell phone numbers are by default unlisted. There is no phone book of cell phone numbers. Polling companies (and telemarketers) can't simply buy lists of cell phone numbers. And unless the current business model--where we pay by the minute--changes, they aren't going to be able to get ahold of any kind of complete list or have any ability to effectively poll a cross section of cell phone users.

Posted by SDA in SEA | June 25, 2007 9:34 PM

There's an article about this in Slate today, and the results tell me that no pollster is gonna work that hard to get the opinion of cell only folks like myself. We're younger, poorer, and less likely to vote. Thus, we can be safely ignored.

Posted by Gitai | June 25, 2007 10:11 PM

Political pollsters may or may not start including cell-only voters. But other survey researchers are most definitely going to be including cell phones. And most people doing legit research aren't using lists of numbers in the first place - they use information about which groups of numbers are most likely to include working numbers, but otherwise call randomly.

It will certainly be more expensive to poll cell only users. But if/when it becomes scientifically impossible to not do it, it will happen. And that will likely be sooner than later. Nation-wide, the latest estimate of cell phone only households is pushing 14%, and it's much higher in younger age groups, and shows no sign of slowing or stopping.

Posted by skinny | June 25, 2007 10:21 PM

I think you'd be surprised about the average age of bikers. In injury prevention studies for various urban centers, the average age of bikers goes into the 40s for sure and can easily be skewed higher for certain demographic constraints. I can't find a citation quickly enough, but here's the stats on the riders from Seattle to Portland last year:
Under 18 years old: 4.3%
18-24: 5.2%
25-34: 19.7%
35-44: 26.5%
45-54: 28.7%
55-64: 13.2%
65+: 2.4%

Posted by Otter | June 26, 2007 10:16 AM

@9: No, that makes sense. Look at that distribution--the average would have to be well under 51, probably early to mid-40s. I'm not saying there are no older riders; I see them all the time. I'm saying an average age of 51 is too high for the general population for sure, and almost certainly too high for the cycling population.

To many of the rest of you: it's sad and shortsighted if a state agency like the DOT is only interested in the transportation habits of likely voters. Nobody at the DOT needs to get themselves elected for another term; they're supposed to serve the public.

Posted by annie | June 26, 2007 10:45 AM

This was a poll of all people as to what they wanted done about the bike and PEDESTRIAN plan, right? Then why would/should the bikers be the deciders? Should decisions about skateboard issues be decided by an average age of 18? Want to see skate board only lanes on the freeway or skateboard only sidewalks?

AW - " And the state shouldn’t be able to deprioritize bike amenities because ridership drops off after a certain age."

Of course they should be able to - or to put it another way, the state should balance out projects to service people of all ages.

And since the person doing this for WashDot said "We hope this will continue to build the case that biking and walking are real modes of transportation, real ways to get around," I would say that the work is biased toward a pro biking conclusion.

As for median age thing - yes it is about 38 but do you really think those say under 9 should be considered in that average or maybe even under 16.

AW -"A poll of the older Washington residents who hang around their land lines just isn’t necessarily representative of this state’s commuters. (Voters, maybe.)"

So the average age of a commuter is probably around 50.

I believe government agencies can't poll likely voters, they have to poll the general population as they are not supposed to be doing work for campaigns with public money.

Posted by whatever | June 26, 2007 11:40 AM

I'd be furious beyond belief if some dimwit called me for a survey on my cell phone. Contrary to many people's apparent assumptions, some of us either use prepaid cell phones, or have very limited calling minutes, because we are not heavy phone users. I only use my cell phone for emergencies - to call AAA if something goes haywire with my car, for example. I don't pay for a bazillion minutes a month, and certainly don't want to use up half my allotment taking some dumbass survey that ought to be done by email anyway.

Posted by Geni | June 26, 2007 12:11 PM

Comments Closed

In order to combat spam, we are no longer accepting comments on this post (or any post more than 14 days old).