New Hill-Harris X poll, which I hope is wrong, shows 5 top candidates all losing support

[highplainsdem]

https://thehill.com/hilltv/what-americas-thinking/457430-biden-sanders-warren-support-dips-in-new-poll


This poll has shown support for all 5 wobbling back and forth, as many polls do.

But this is the first one I've seen that shows Biden, Sanders, Warren, Harris and Buttisgieg all losing a bit of support in the same poll.

It does show an uptick in support for some candidates who hadn't been doing very well in earlier polls.

The survey showed that 31 percent of likely Democratic primary voters back Biden, marking a 3 point drop from an identical poll conducted two weeks ago.

-snip-

Sanders’s support dipped 4 points to 16 percent, while Warren’s support dropped 2 points to 10 percent.

Sen. Kamala Harris (D-Calif.) retained her spot in fourth place but her 7 percent support is a decrease of 2 points since the July 29 poll. South Bend, Ind., Mayor Pete Buttigieg rounded out the top five with 4 percent, a 1 point dip over the past two weeks.

Former Texas congressman Beto O’Rourke, meanwhile, held onto to sixth place with 4 percent support, which is unchanged from two weeks ago.

The number of undecided voters ticked back up to 10 percent, after briefly dropping to 8 percent.

-snip-

Gov. Steve Bullock (D-Mont.), former San Antonio Mayor Julián Castro (D), and Sen. Amy Klobuchar (D-Minn.) all ticked up one point to two percent. Best-selling author Marianne Williamson, who had a breakout moment during the second round of Democratic debates, saw a two-point jump to 2 percent.

Some of the other candidates got votes but are still polling at 1% (Yang, Ryan, Sestak, Gillibrand, Steyer, Booker, Delaney, Gabbard) or at 0% though still getting 1 vote each (Inslee, Bennet) or 0% and getting 0 votes (Moulton, Hickenlooper, De Blasio, Messam).

Of the top candidates, Biden lost the smallest percentage of support, dropping from 34% to 31%, a drop of about 9%. The other four lost 20%, 17%, 22%, and 20%, in the order named.

It does seem odd, after debates that focused so much attention on our candidates, to have the number of undecided voters increasing, even a little.

[qazplm135]

1. As I have been saying all along

Microanalyzing polls this far out is pointless. A couple point rise or fall is statistical noise. The meaning of polls this far out is extremely limited.

[JoeOtterbein]

3. +1

[TheCowsCameHome]

2. You'll drive yourself crazy

following all these useless polls.

Come back in six months when they might mean something.

[JoeOtterbein]

4. Yep.

Too little, too soon.

[kurtcagle]

5. Not really

The debates served to highlight some of the other candidates, and a lot of people who likely have not been paying that much attention yet are beginning to determine which candidates they like.

[delisen]

6. This is the Mark Penn poll, is it not?

Is it the one with a C+ rating from 538?

Mark Penn has become such an avid defender of Trump I tend to discount the polling done by his outfit.

[pnwmom]

7. Why does anyone pay attention to a poll 538 rates as a C+ poll?

[highplainsdem]

9. C+ rating, true, but an 86% accuracy rate according to 538, not 56%

[pnwmom]

11. You're right. They're doing better than they used to. But it's still a C+ online poll. n't

[thesquanderer]

13. C+ matters, online doesn't.

The grade reflects the quality of the poll, and takes numerous factors into account. A poll can be online and A rated, or not online and more lowly rated.

[pnwmom]

18. An online poll can never state a statistical margin of error. If they were as good

no one would ever do a real probability poll, because they're much more expensive.

[thesquanderer]

19. An online poll can have a margin of error...

...here's a recent yougov online poll with a 2.6 MoE.

B-rated.

https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/c1mjtqtl2f/econToplines.pdf

[pnwmom]

20. Where does that poll say how it was conducted? n/t

[thesquanderer]

21. On the last page. (n/t)

[pnwmom]

22. I still don't understand how this online panel poll can claim a margin of sampling error.

From the American Association of Public Opinion Research

https://www.aapor.org/AAPOR_Main/media/MainSiteFiles/Margin-of-Sampling-Error.pdf

Note that margin of sampling error is always expressed in percentage points, not as a percentage- -for example, three percentage points and not 3%. And the margin of sampling error only applies to probability-based surveys where participants have a known and non-zero chance of being included in the sample. It does not apply to opt-in online surveys and other non-probability based polls. More about this in a moment.

SNIP

The Credibility Interval

As online surveys and other types of nonprobability-based polls play a larger role in survey research, another statistic has emerged that is often confused with MOSE. It is called the “credibility interval” and it is used to measure the theoretical accuracy of nonprobability surveys. While both MOSE and a credibility interval are expressed in the familiar “plus-or-minus” language, they are very different.

The credibility interval relies on assumptions that may be difficult to validate, and the results may be sensitive to these assumptions. So while the adoption of the credibility interval may be appropriate for non-probability samples such as opt-in online polls, the underlying error associated with such polls remains a concern. Consequently, AAPOR urges caution when using credibility intervals or otherwise interpreting results from electoral polls using non-probability online panels.

[thesquanderer]

23. I wonder if it depends on the definition of an "opt-in online poll"

For yougov, you can opt in to the pool they draw from, but you cannot opt into a specific poll.

OTOH, any poll you can specifically opt into (even assuming controls to assure no one votes more than once) I think would be so suspect from the outset that MoE or credibility interval would be moot.

Another possibility is that the info is outdated (the doc is at least 3 years old) and online polling's selection/representational capabilities have improved since then.

Or maybe yougov is calling it MoE when they really mean credibility interval.

I don't know. It's a good question, though.

[Laura PourMeADrink]

8. I find this very interesting! Thanks for posting.

Perhaps no one has captured the essense of what people out there are looking for - not for the rest of their lives..but for this moment in time. This day, when they just want Trump in the rearview mirror. Some wise people on KLBJ in Austin said it best. Don't be a policy wonk...it's simple ..all you have to do is not be trump.

[highplainsdem]

10. Thanks, Laura! My best guess about this poll is that it reflects the downside of a primary

race that runs too long, has too many candidates, and has seen multiple candidates hurt by a circular firing squad.

[Laura PourMeADrink]

12. Totally agree. Yet we, for many years, are always huge proponents of giving everyone a chance

to speak, everyone a chance to debate. It’s so ironic that for this moment in time, it actually hurts us. In a way, it seems rather self-serving to me too. Given the absolute imperative to remove trump, why pick this critical time in history to cobble together the minimum money and support to muddy things up and diminish the front runners?

Remember being shocked that our debates started so early but even more shocked that there was little mention of trump. Like he doesn’t even exist.

[qazplm135]

17. or it's simply statistical noise

in a poll that is still half a year away from the actual vote.

[thesquanderer]

14. re: "which I hope is wrong" -- Why?

This early in a campaign, with so many candidates still barely known, I would not be surprised to see some lesser known candidate start increasing their support, and some of that support naturally would come from people who had previously favored other, more widely known candidates. If some lesser candidates start to increase their numbers, you'd expect that would be accompanied by some leading candidates' numbers dropping off.

I don't see anything troubling about our top candidates losing support among each other. I would only be troubled if they were losing support relative to Trump (i.e. if they started doing more poorly in head-to-head poll matchups against Trump).

[highplainsdem]

15. See my reply 10 above.

I agree that it's a more serious worry if our candidates start doing worse in matchups against Trump.

OTOH, there was a NC poll yesterday or the day before which did just that -- showed all our candidates losing to Trump in NC, after earlier polls showed them beating him.

https://www.democraticunderground.com/1287239147

Circular firing squads never help.

[thesquanderer]

16. That post about North Carolina was meaningless.

there was a NC poll yesterday or the day before which did just that -- showed all our candidates losing to Trump in NC, after earlier polls showed them beating him.

Not a valid comparison, explained at post #11 in that thread.