PPP's North Carolina poll: 80% were landline users

[molova]

At a time when 40% of the US population still has landlines (as of 2015), PPP conducts a poll where 80% of respondents are landline users. Not surprisingly, Trump came out ahead by 2%:

Methodology: Public Policy Polling surveyed 1,024 likely voters from September 18th to 20th. The margin of error is +/-
3.1%. 80% of participants, selected through a list based sample, responded via the phone, while20% of
respondents who did not have landlines conducted the survey over the internet through an opt-in internet
panel.
http://www.publicpolicypolling.com/pdf/2015/PPP_Release_NC_92116.pdf

[MANative]

1. As with all data, garbage in, garbage out.

[democrattotheend]

2. PPP is usually a good pollster

They were one of the most accurate in 2012. So I am hesitant to throw them under the bus just because we don't like the result. They probably weight the poll to compensate for overrepresentation of landlines.

[molova]

5. So then...the best pollsters should save money instead of calling cell phones

I mean, if all they have to do is "probably weight" the data, why not save money and do that?

Right?

[MyNameIsKhan]

11. Any IVR based polling is not accurate, live interviews are needed

[democrattotheend]

14. I disagree

IVR can actually be better because
1) It's faster, and thus can do more interviews over a shorter number of days
2) With a candidate like Trump, IVR may get more honest answers because people who would be embarrassed to tell a live interviewer they are voting for him might not have the same reservation with a robot.

I used to work for a labor union and we had pretty good results using IVR to poll congressional districts.

[BlueCaliDem]

3. Ach. I've been watching PPP lose credibility as a good poll over the past months...eom

[bmstee01]

4. Why Bother

Using a landline? My god I have one friend in their 30s that has a landline and it's a running joke! Also, when I get calls on my phone that I don't know the number I don't answer.

[triron]

6. I do

that as well (i.e. don't answer)

[bigbrother05]

9. 60+ and haven't had a land line in over 10 years

[ProudToBeBlueInRhody]

10. I have one

First, I don't want a lot of people knowing my cell number. I can screen a lot easier. My landline is like a junk email box.

Second, for longer conversations a landline won't drain my cell battery and is clearer.

[Iliyah]

7. Thats their structure.

Most cell phone users have caller id. I personally do not answer cell phone calls from numbers I don't know. That's why pollster predict many number outcomes. Bloomberg/Selzer could not mustard up enough LVs under 50 so therefore, they predicted that since 50 and over overwhelmingly would vote for t-rump and using a model of over 80% of white voters. Ohio is much closer than ya think.

[MyNameIsKhan]

8. NC with be finally HRC +4, Early votes will tell us by mid October.

[Loki Liesmith]

12. They try to make up for it with their internet survey.

They know it's a problem. Let's see if they have a solution.

[molova]

13. Still, if we equate internet with cell phones, are there only 20% cell phone users?

In North Carolina, versus 80% landline?