Huge well-controlled CA exit poll deviates 16% from Dem results, but only .07% for GOP.

[J_J_]

Source.



The GOP exit poll.



EDIT: Forgot to include the Dem exit poll.



EDIT 2: I made a new post about how Bernie will win California, here. This is ABSOLUTELY CRUCIAL INFORMATION that everyone should read!! Please go up-vote it for visibility.


From Reddit
https://www.reddit.com/r/SandersForPresident/comments/4n5o2z/huge_wellcontrolled_ca_exit_poll_deviates_16_from/

[CrowCityDem]

1. 'We only interviewed on campuses, so it was obvious to us he won'. Something like that?

[JoePhilly]

2. We should throw out the votes and use exit ...

... polls to pick nominees!!!!

[CrowCityDem]

4. Yeah, asking a few people and estimating is so much more accurate than asking everyone!!

[LexVegas]

3. ....

[KingFlorez]

5. Stop it

This is beyond a lost cause.

[FSogol]

6. Um hadn't all the GOP candidates dropped out except 1?

Laughably naive.

[J_J_]

9. GOP exit polls were accurate in all the primaries, Dems way off

So GOP tells the truth and Dems lie to exit polls??- THAT is what they expect us to believe?

People were so embarrassed to vote for Hillary they lied about it?

[FSogol]

10. Not true: "GOP exit polls were accurate in all the primaries"

Exit Polls are highly suspect, that's why the media largely ignores them. There are some sociology papers on how the dress, race, or gender of the pollster affects how people respond to them.

[GreatGazoo]

12. the media uses exit polls to call elections before the vote counting

Exit polling is data modeling so it accounts for thos ewho don't answer and the variance between the sample and the larger mix. GOP exit polls were within MOE. On the Dem side exit polls were as much as 16% off and always in favor of Clinton.

[FSogol]

13. Reddit mythology. n/t

[Stevepol]

15. Are you serious?

Statement: "There are some sociology papers on how the dress, race, or gender of the pollster affects how people respond to them."

I'm sure dress, race, etc. affects how people respond to them as individuals (i.e., sociologically speaking), but that has nothing to do w/ their lying about who they voted for.

Exit polls are usually highly accurate. In the US this is not so now because the results are regularly tilted in a Repub direction and the polls have to be adjusted to fit the results spit out by the voting machines. The "Reluctant Republican Responder" thesis has become a standard theory to explain the difference between the exit polls and the machine results. The tilt is now built into the polls done by many pollsters, but the "unadjusted" polls are probably quite accurate and reflect the actual results rather than the machine results.

Of course, it's impossible to "prove" any of this because our election results are almost never VERIFIED and in many cases are actually UNVERIFIABLE. Until the results of EVERY ELECTION are verified by carefully done audits and when the audits reveal highly suspicious numbers the whole election is recounted BY HAND, there really can't be a democracy in the US again. This is what Beth Clarkson, a statistician here at Wichita State University, is running into in trying to get Kobach, the KS Sec of State, to allow her to compare the paper printouts of the elections in Sedgwick County with the machine result totals. The results are highly suspicious all over the country, not just in KS.

[John Poet]

11. I would be, if it were me.

"People were so embarrassed to vote for Hillary they lied about it?"

[NCTraveler]

7. So you are comparing an absentee ballot exit poll....

With a non-absentee ballot exit poll.

Well done.

[Lord Magus]

8. Would you like to grasp at some of these?

[onenote]

14. Am i correct that this is a self-selecting poll?

From the description it looks like the pollsters sent email surveys to voters who had voted absentee (how many? randomly chosen?). And they were dependent on the recipient of the email filling it out and returning it.

Lots of questions there. What was the response rate? What does it mean when it says that the responses were "weighted by age." And so on...

[Hortensis]

16. Any exit poll that had only 0.07% (!!!) deviation

would be purely by accident. And it is not true that exit polls are very reliable.