A question for the blogosphere: I’m hearing rumors that US presidential election polls are way off, since they’re not touching cell phones. I can’t get a good piece on this yet, though. All I can find are opinion columns, a few notes in other places, and descriptions of pollsters being blocked from calling mobile users.
If this is true, and a real problem, how do we integrate mobile phone telephone into building pictures of public opinion? What applications have emerged which let us do that? If we don’t, how far off are poll results?
Comments
@ 09:15
http://www.textually.org/textually/archives/005375.htm
The rise of cell phones has made it harder for pollsters to track down respondents, analysts said, and a rising number of people don’t want to participate at all.
@ 16:42
From my blog, Politics in the Zeros, Sept. 25, 2002
Polls no longer reliable, Dick Morris
Two reasons:
In 28 states, the state legislatures have passed laws giving telephone users the right to opt out of receiving telemarketing phone calls, including public opinion surveys. More and more voters are availing themselves of this right and the pickings for telephone polling firms are getting more and more scarce.
Even beyond the formal opt-out which makes it illegal to call certain voters when taking public opinion polls, the “hang up” factor is looming larger and larger in telephone polling.
Indeed, I have friends who NEVER answer their home phone, letting the machine take the call instead. Others simply aren’t home much during the 6-9 PM calling time of most polls, or their primary phone is their cell, not their landline. And I’ve never gotten a poll or telemarketer on my cell phone, have you?
So, phone polling would appear to be taken on those who 1) are home and answer their landline phone, 2) don’t immediately hang up. This is already a considerable smaller universe than a truly random sample would allow.
(The article by political strategist Dick Morris is no longer on the original site)
@ 10:50
Thanks for the trackback to my blog, Metaforix@.
I just came across the following post by Jeremy Chrysler: The Polls may be dying. It purports to review Richard Ling’s new book, The Mobile Connection, but is mostly a discussion of how excluding cell phone and VoIP users amounts to excluding a disproportionate number of young voters from polling samples.
As cell phone use and VoIP use grow, not only will poll results be further skewed, but young people’s already low level of participation in the political process may decline as they realize that their voices are not being heard.
The poll numbers themselves are less important than how they are spun by the campaigns and the media. For example, suppose that young potential voters disproportionately favor the underdog. Polls that exaggerate the distance between the underdog and the frontrunner may discourage young people from registering and voting, in the mistaken belief that their votes cannot influence the outcome.
This is hardly a trivial problem, as few of us want to encourage unsolicited cell phone calls (for which the recipient pays), even in the worthy cause of civic participation. My guess is that the solution will ultimately lie in reputable online polls, appropriately weighted to reflect the population at large and corrected in some way for the self-selection factor. (Of course, the self-selection factor in telephone polling — as in call-screening, mid-poll hangups, and lifestyle choices that don’t coincide with pollsters’ calling hours — is in most cases conveniently ignored).
@ 07:29
This morning’s Wall Street Journal has a front-page article, “Divergent Opinion Polls Reflect New Challenges to Tracking Vote.” (Paid subscription required for access, sad to say ;-). The cell phone issue is mentioned in passing. There’s more focus on the issues of defining “likely voters” and the impact of identifying party affliation on the outcome.
@ 08:36
There’s a discussion going on at Kuro5hin.org about this, too.
@ 10:10
I work for a newspaper and we are trying SMS voting on cellphones to see if the votes we receive via text messaging will match up to election night results. We want to know if opt in opinion polls using SMS are accurate as they ask for opinion but allow repsonses when it is conveinent for cellphone users. we advertise the poll at movie theaters, in paper, online and on air. We wish to see if this is a good method to get a close view of the younger market’s opinions
@ 01:46
ReciÈn sacado del horno
Dice un titular de el Navegante El tel‚àö√†fono m‚àö√µvil, un ‘enemigo’ para las encuestas electorales en EEUU