Pew says Internet polls often 'more accurate'

[pinebox]

Last edited Mon Nov 16, 2015, 11:26 PM - Edit history (1)

So about those internet polls......

Apparently Pew says online polls can be "more accurate" because people are more candid responding to a computer than a live voice in many cases.
At a minimum, according to a lengthy Pew Research Center study, for most questions there isn't a big difference in the answers people give on the phone or online. That alone is good news for polling.

http://www.pewresearch.org/2015/05/13/from-telephone-to-the-web-the-challenge-of-mode-of-interview-effects-in-public-opinion-polls/

It's one very huge article and has a complete break down of why internet polls are often more reliable than those conducted by telephone.

[Cali_Democrat]

1. There's a difference between a scientific internet poll

and a poll at CNN.com where people can vote 1000 times if they wanted to.

[pinebox]

4. Hey Pew is saying they're more accurate

so I think a polling agency knows wtf they're talking about, not a bunch of armchair pundits but what do I know

[Cali_Democrat]

6. The respondents were selected by the polling firm

The results are based on 3,003 respondents who were randomly assigned to either the phone or Web mode and interviewed July 7-Aug. 4, 2014 for this study.

That is entirely different from a poll at CNN.com where people can vote 1000 times if they wanted to.

This is not difficult to grasp.

[tammywammy]

41. This isn't rocket science, are people really not grasping the difference.

Sometimes the ignorance and stupidity of basic concepts for some DUers amazes me.

[SidDithers]

68. They're being deliberately obtuse...

They must be.

Sid

[Bread and Circus]

18. Curious... how do you vote multiple times in these online polls?

Most online polls only allow me to vote once.

[Cali_Democrat]

19. Clear cache/cookies on your browser...vote again. nt

[DemocratSinceBirth]

20. Now, you added 1,000 more internet votes for Bernie.

[Bread and Circus]

21. I guess... so why don't Clinton supporters do this too?

[DemocratSinceBirth]

23. Not puerile enough?

EOM

[Bread and Circus]

26. Is that the argument, collectively Clinton supporters are better human beings?

[DemocratSinceBirth]

29. You said that.

I believe this tete a tete is going to end as well as our last one for you but I sincerely appreciate your desire to entertain me.

Please continue.

Thank you in advance.

[Scootaloo]

46. Not sure which Clinton followers you've been looking at...

Maybe you could introduce me?

[still_one]

28. go to a different ISP with different IP address.

[Bread and Circus]

36. how do u go to a different isp?

[still_one]

44. Library, McDonalds, Starbucks. Deleting cookies might work, spoofing your IP address. It may not

be worth the trouble though

[Bread and Circus]

54. so people really thing people would do that for any online poll?

[still_one]

61. I was just responding to the issue that if someone could figure out a way to vote multiple times in

an online poll. To your specific question, probably not.

[Bread and Circus]

63. yeah that would be lot of work

[kristopher]

33. You don't. At least, not in the type of poll the OP covers.

I don't understand why some Clinton supporters want to keep people ignorant of how polling is changing.

[restorefreedom]

2. also there is a self selection aspect

some people like myself would never participate in a phone survey but would do so on internet. i just don't like talking to strangers, and i am guessing i am not the only one

[kristopher]

34. That is actually a central part of what the article is about. nt

[restorefreedom]

51. semi reclusive people like me

don't always interact, but most of us have opinions.

will be a very interesting election day(s)

[zappaman]

3. Ok. n/t

[1StrongBlackMan]

13. I guess in this morning's briefing ...

The topic was: find something science-ish to validate the unscientific ... and post it to DU.

[kristopher]

38. Can you read?

If you can, then you should apply that skill to the article in the OP. You are clearly talking up nonsense for some perverse reason and it is getting very, very old.

[1StrongBlackMan]

40. No. I cannot read. Next question. n/t

[kristopher]

42. Then tell your handler to read the article to you.

[1StrongBlackMan]

49. Why are you being such a {term deleted}? n/t

[kristopher]

60. I'm sick of people hijacking threads with jackass behavior.

The OP has a lot of good information we should all know.

[1StrongBlackMan]

66. Sounds like a personal problem ... Get over yourself. n/t

[1StrongBlackMan]

67. Okay ... Now that time has passed ...

what prompted my "jack-ass" post was: 1) Noticing early in the day 2 OPs talking favorably of (un-scientific) online polling; and 2), seeing the first several posters (seemingly) associating this as a validation of clickbait polls, that so frequently get posted on DU.

[MADem]

5. Ah yes, the accuracy of clear cache/vote, clear cache vote...

Or writing a script to give a candidate Lil Kim/Saddam-like approval ratings!

P-ew, indeed!

[Agschmid]

8. Oddly enough on Slate's you don't even have to clear the cache.

I voted for O'Malley like 10 times.

He still didn't win

[MADem]

16. Ha ha ha! You needed to step it up--a hundred votes, from a thousand willing

clickers!

[Bread and Circus]

24. maybe you voted 10 times but it disregarded duplicates from the same ip address?

[Agschmid]

25. Nope.

Vote counts went up and it asked me to pass a Turing test each time...

I passed.

[Bread and Circus]

27. I haven't voted for many of the online polls except I voted once for the cnn poll last time.

[Agschmid]

31. I tried each one this time to see if it would take more than one vote.

I don't normally vote in them.

[Bread and Circus]

35. My question remains why don't Clinton supporters spam these polls with votes?

Seeing as how there are more Clinton supporters and they are super excited about her debate performances why don't they go online and vote?

What is stopping them?

[kristopher]

39. READ THE ARTICLE!!!!!

[Bread and Circus]

52. What will I learn?

[1StrongBlackMan]

48. I can think of a number of reasons ...

not the least of which, is Why take the time to do so? I would think the only reason for doing this is to ... well ... spam a poll.

[Bread and Circus]

50. But the unproven claim is Bernie supporters do, why the difference?

We just care more? We are inherent liars? Clinton supporters are lazy?

I personally argue it is a generational thing not the Sanders as Spammers theory.

Bernie pulls a younger more net savvy crowd. Clintons is more traditional and older.

Bernie wins online polls because his supporters are online.

Clinton wins the landline phone polls because her supporters are more like to have landline phones.

[1StrongBlackMan]

57. Unproven claim? Well ...

I have seen posts on DU soliciting spamming.

[hack89]

58. Because it serves no practical purpose?

there is no point in doing it so why waste my time?

[Bread and Circus]

64. Can't Sanders supporters say the same?

[hack89]

65. If you accept that Bernie is attracting new young voters (the revolution)

It would make sense that Bernie supporters have a bigger presence on social media and the Internet- it is an environment that is second nature to them. Not so older voters.

[Agschmid]

7. *Accurate

Accuracy matters

[kristopher]

43. So does actually reading the material.

[NCTraveler]

9. You should really try to read and comprehend the links you post.

I do know how condescending that sounds but it isn't the first time you have attempted to deceive in this manner. This one is so obvious it just isn't going to work for you. Can you show me the point in your article that backs the accuracy of instant online polls. After all, that is what has been discussed around here.

[1StrongBlackMan]

15. See post #13. n/t

[kristopher]

45. Where was that done by anyone except a select group of HRC 'fans'?

[sufrommich]

11. This has already been explained to them.

Many,many times.

[tritsofme]

12. LOL, this doesn't mean what you wish it does.

This article is not taking about the silly clickbait polls that Sanders is "winning".

[DanTex]

14. They aren't talking about the self-selecting clickbait polls that Bernistas hail as "the people."

They're talking about scientific polls conducted by internet. Big difference.

[DemocratSinceBirth]

17. Pew is referring to polls that attempt to approximate a random sample.

This isn't the stuff of graduate level social science research.

[MaggieD]

22. They aren't talking about click bait polls

LOL!

[randys1]

30. Putting aside the usual Bernie vs Hillary aspect of this, the GOP plans on

stealing elections all across the country using various methods, especially hacking or manipulating electronic vote machines.

They have done this in the past, they did it earlier this month and they are going to continue.

Exit polls will show the discrepancy but we are now told to ignore exit polls, even though before they started to do this exit polls were tremendously accurate.

[kristopher]

32. That's an excellent article on where polling is going.

It's too bad some partisans want to keep people from understanding how things work. Thanks for posting it.

I highly recommend everyone forget for a moment about the primary race and just read the freaking article. It contains information we all should understand.

[Thinkingabout]

37. No substance here, if these same polls said Clinton or O'Malley won the debate, they would be

Declared wrong, we know the game, some may feel better about the candidate of their choice so it can be called a feel good poll.

[kristopher]

47. Another commenter that didn't bother to read the article.

Amazing.

[Thinkingabout]

62. Another commenter who is wrong, amazing how one tries this frequently.

[Orangepeel]

53. "American Idol" lets people vote by phone. Is that a "telephone poll"?

The online poll that Pew is talking about is a scientific poll in which the responses happen to be obtained via the Internet, rather than via a phone or by mail.

There has been uncertainty about how representative data collected with online polls could be, because not everyone had access. (The most commonly cited reason why the polls had Dewey defeating Truman was because the polls were phone based, which at the time, underestimated less affluent voters). That's becoming a smaller and smaller issue with data collected online. Of course there will be a time at which they become more accurate than phone polls, and that time may be here.

Pew is not talking about a website poll that people self select into. Those things would be analogous to taking a poll by telephone and asking anyone who wants to to call in and vote (which is kind of what politicians do when they count pro and con calls to their offices). Those "polls" may indicate enthusiasm by a subset of supporters, but don't necessarily reflect broader public opinion.

[moobu2]

55. Pew isn't talking about those "A or B" polls that turn up on every news site on the internet

after whatever current event has just taken place.

[sabrina 1]

56. You're going to make some people VERY unhappy. But thanks for the article anyhow, we need to

move forward from the regressive polling of the last century and start doing some research on how to use the new media rather than dismissing it as many Hillary supporters try to do.

[DemocratSinceBirth]

59. Res ipsa loquitur

About the Study

Total RespondentsThis study was conducted using Pew Research Center’s nationally representative American Trends Panel (ATP). Panelists who normally take their surveys on the Web were randomly assigned to either the phone mode (N=1,494 completed by phone) or the Web mode (N=1,509 completed on the Web). Each set of respondents was independently weighted to be representative of the U.S. public in an effort to ensure that any differences observed between the groups were a result only of mode of interview effects. Mode differences for each question in the study were measured by comparing answers given by the Web and phone groups using a commonly reported category of each question in the study or the category that shows the largest mode difference — whichever is larger.