How Serious Is Polling's Cell-Only Problem?

Pew Research ^ | June 20, 2007 | Scott Keeter

[xzins]

Summary of Comparisons Between Landline Samples and Cell-Only Samples

Number of survey questions compared 46

Average (mean) difference between landline and cell-only samples across all 46 questions 7.8%

Range of differences (absolute value) 0% - 29%

Maximum change in final survey estimate when cell-only sample is blended in 2%

Average (mean) change in final survey estimate when cell-only sample is blended in 0.7%

[xzins]

There are other polling problems. I simply won’t respond to unknown phone calls. (See congressman billybob’s recent article in which he discusses what might have been a phony pollster)

Are there allowances for the large number of folks like me who won’t talk to them?

Are we a unique group with characteristics that would skew their results?

[Soliton]

Nother fallacy is that they don’t call cell phone numbers. I changed from landlines to all cell last year, but I was able to keep my long-time phone number. I get the calls on my cell now. (I don’t take them, but I get them). I’m sure that many people do the same.

[cripplecreek]

I won’t talk to them on landline or cell.

[Arkansas Toothpick]

My boss, who is NOT a Republican, thinks that the majority of the cell-only crowd will vote for Obama. However, I don’t think it’s that cut-and-dried. I think a lot of the POTENTIAL Obama voters in the cell-only contingent won’t vote on election day.

[Soliton]

I think a lot of the POTENTIAL Obama voters in the cell-only contingent won’t vote on election day.

Most won't even register.

[j. earl carter]

There are other polling problems

I would say people screeening their phone calls with caller ID is likely to be a bigger problem than cell phones.

[listenhillary]

If I ever get polled, I will mislead them. I want the pollsters to look foolish. Maybe they will stop being considered news.

[palomonte]

a lot of them will only vote when they can text message their vote in.

[AndyJackson]

I don’t think Republicans are inclined to view favorably “folks” [they aren’t really humans]who: invade privacy; use property they didn’t pay for; take up valuable time that could be spent working or thinking or Freeping; and ask loaded questions like do you want your next president to be well educated (and then marking the yes as an Obama vote), all with an obvious political agenda.

[DYngbld]

If I find a poll taker on the other end of my phone I will normally just decline and hang up. I always wonder if the guy on the other end is just putting down his opinion and going on to the next caller. How do those clowns get paid by the number of calls made, or responses?

[cripplecreek]

In the event that I do talk to someone from a campaign I don’t give them much to go on.

When they call and ask what I want to see from their candidate I tell them I want to know the candidate’s opinions, not what I told him I wanted to hear.

[George from New England]

Other polling problems ...

I got an enthusiastic call from an Obama support within the last week. She asked how I’d vote and I said I didn’t know enough about the local candidates. When she push on Obama I said that was a given and convinced her O had my vote.

I love those polls!

[downwdims]

I’ll say this. All the single people I know don’t have land lines. Both sides. The Married ones do. It’s anecdotal at best but it’s what I have. (I’m 30)

[posterchild]

I would say people screeening their phone calls with caller ID is likely to be a bigger problem than cell phones.

That certainly has to eat into their success rate, though they can counter by autodialers which call many numbers and let the autodialer see who answers before placing a human operator on the call.

I've been cell only since 1998 and marketing calls have been extremely rare until about 2.5 years ago when I moved to another state and changed my number and cell provider. Now I get one a few times per year (not counting the obnoxious collection agencies calling for some woman who had this number over 2 years ago.)

I'll often check out numbers I don't recognize at http://whocalled.us

[deport]

However, I don’t think it’s that cut-and-dried. I think a lot of the POTENTIAL Obama voters in the cell-only contingent won’t vote on election day.

---

I tend to agree with your assessment. Also until the dynamics of the voters change to the extent that cell only voters become a larger impact then the current process will still be valid.

In a national election such as President they have this basically broken down to minute details and focus on very specific areas to determine or make projections. Example in PA the County were Scranton is located often determines the outcome of how the state of PA votes.... Thus they will spend more efforts to survey that county in detail to make their projections...

[xzins]

Excellent point. It adds to the likelihood that their numbers are skewed

[Core_Conservative]

There are a couple of misconceptions in the article. They don't seem to realize that a great deal of people only use cell phones (Liberals AND Conservatives). I know a lot of Conservatives that just gave up on the landlines because of the cost of keeping 2 sets of phone bills (I have 5 cell phones in the household and 2 landlines - one for my business). A larger number of people have also opted for the option of Caller ID and WILL NOT answer the phone to numbers they do not know. The only reason I have answered unknown numbers is because (we had a fire in our house this spring) our insurance company refuses to give any information on the line when they call - I tried to get them to call my cell phone - since that is for business too - but they always call my home number - I HATE STATE FARM.

[xzins]

I'll often check out numbers I don't recognize at http://whocalled.us

Wow. You need to post that as its own thread.

[Mr. Brightside]

I have a few telemarketers working for me.

Cell phones are great to call.

BUT... we try to stay away from prefixes that are used for the “prepaid minute” plans.

Those people don’t have money or jobs. Their phones are always getting turned off until they can “bump up” their minutes, etc.

Plus, we are seeing lots of people keeping their phone numbers after moving out of the area.

So if they are surveying people with a Michigan area code, they may be calling people who USED TO LIVE in the state but now live in Alabama.