Ipsos/Reuters Poll w/ Filters Cleared: Bernie 34%, Hillary 32%, O'M 6%

[HerbChestnut]

http://polling.reuters.com/#poll/TR131/dates/20150808-20160125/type/smallest

Has basically the same number of responses as the other poll recently posted. Let's not overreact to polls...

[Renew Deal]

1. I clicked your link and it says 44-26 Hillary.

A lot of people aren't voting which drives her numbers even higher. And the numbers are jumping all over the place. What is that?

[HerbChestnut]

2. I just clicked my link to be sure, and it says 34-32 Bernie. Weird.

Jumping numbers are probably a result of low response rates and typical polling variation (exacerbated by low response rates)

[Renew Deal]

4. It seems to be rotating.

Maybe it's a partial days results?

[Attorney in Texas]

11. If you accidentally drag your mouse over the graph, it gives you the data from the date where your

mouse is touching.

[HerbChestnut]

9. Also make sure the results are being displayed for December 1st.

If you move your mouse around the graph the results will show from previous dates.

[Bjorn Against]

10. That is what I saw when I first went to the same site less than an hour ago...

Now when I went there again I see the same results the OP reported so it looks like the numbers changed drastically for no clear reason. Also if you put on the "Tea Party affiliated" filter it shows Hillary getting 43% among Tea Party affiliated voters, I know she is conservative but even so I have a really hard time believing she has 43% support from Teabaggers. The numbers in this poll do not make sense.

[NurseJackie]

3. "Likely Democratic Primary Voters" seems like a meaningful one.

[Attorney in Texas]

12. Reuters is reporting its poll as 57% Clinton to 29% Sanders (i.e., no real change). The reason why

Reuters is reporting its own poll contrary to the way that you would like to read it is because if you put the "likely voter" filter on (which is a good filter if the polling is broad enough to successfully apply the filter), it reduces and skews the sample size so that the data is no longer reliable.

I suspect that DemocratSinceBirth is well aware of this.

[Agnosticsherbet]

5. well, it changes- Clinton 42.4 Sanders min 25.7

Overall

RESPONDENTS

Former Sec. of State Hillary Clinton42.4%
Wouldn’t vote26.9%
Ver. Sen. Bernie Sanders25.7%
Mar. Gov. Martin O’Malley2.5%
NY Gov. Andrew Cuomo1.2%

[Renew Deal]

7. Wouldn't vote should be ignored completely

Meaning that all candidates numbers go up.

The 26.9% that won't vote don't count. So Hillary is 42.4 of 73.1 (100-26.9). 58%

[Agnosticsherbet]

8. That puts up wh r e she has been running.

The polls seem to be very stable.

[Renew Deal]

6. Just figured it out. Unfiltered means unweighted. And that's for part of one day.

"Wouldn't vote" is in there which doesn't normally count. This means that actual numbers for all candidates are higher.

5 day average of just Registered Democrats: 62-29
5 day average of all Democrats: 57-29
5 day average of likely Dem voters: 77-17

This is a pretty handy application.

[Attorney in Texas]

13. The problem si that the sample size of registered and likely Democrats is too small (and too

divergent from the population) and so Reuters is publishing the all Democrats number.

You can look at the cross-tabs but the more you drill down, the smaller the sample size.

[HerbChestnut]

15. LOL no. I posted it in response to the other poll.

The reason being is that other OP is basing results off of 111 respondents to the poll, which is an absurdly low number. My OP has 113 respondents, which is why I mentioned them. So if you look at it that way they two are comparable.