2016 Postmortem
Related: About this forumEmerson College poll: Clinton 65 - Sanders 26 - O'Malley 2
One candidate seems to be going up, and one seems to be going down.
http://media.wix.com/ugd/3bebb2_b85d13974aeb4901bd68916b963cea3c.pdf
JaneyVee
(19,877 posts)brooklynite
(94,719 posts)JaneyVee
(19,877 posts)Jarqui
(10,130 posts)the demographics are even worse on this poll
42.6% are 65 or over on this poll (compared with 15-16% or so in the general election (in 2008 and 2012)
http://media.wix.com/ugd/3bebb2_b85d13974aeb4901bd68916b963cea3c.pdf
That's the age group where Clinton is strongest and Sanders weakest.
That's a dramatic skewing with demographics - to the extent, it's just plain silly. No wonder they're an outlier.
mattvermont
(646 posts)I was about to point that out. What is the rational for such a spread?
Jarqui
(10,130 posts)laziness? that's just how things turned out?
I have no idea. It's pretty much a useless poll it's so slanted (except for the majority of folks who don't look under the hood ...).
JaneyVee
(19,877 posts)Jarqui
(10,130 posts)42+% of those who voted in 2012 were not 65 plus - it was closer to 16%
http://media.wix.com/ugd/3bebb2_b85d13974aeb4901bd68916b963cea3c.pdf
(see question #18 asking age and it's responses)
JaneyVee
(19,877 posts)30% of eligible voters voted in 2014, what is 16% of 30%? More than half.
NuclearDem
(16,184 posts)The 16% figure represents the contribution of that age group to the total votes cast of all age groups in the election.
The other figure concerns a percent within that specific age group. 51% and 49% respective to gender of eligible voters within that age group voted.
That 51% and 49% of voters in that age group who actually voted made up 16% of the total vote.
onenote
(42,752 posts)Imagine if the total universe of potential voters was 1000 and only 100 (10 percent) of them were over 65 and 900 (90 percent) were under 65. Then imagine that past turnout was 30 percent. If turnout was evenly distributed over all ages (i.e., 30 percent of those under 65 voted and 30 percent of those over 65 voted), a poll of 300 voters that had 270 participants under 65 (30 percent of 900) and 30 participants over 65 (30 percent of 100) would be an appropriate division, even though it would mean that, as reflected in the total population, 90 percent of those polled were under 65 and only 10 percent were over.
If turnout varied by age group, so that only 30 percent of the 900 under 65 voters in the example turned out (270) and 100 percent of the over 65's turned out (100), a poll of 300 people would be appropriately weighted if 81 of those participating (27 percent) were over 65 and 73 percent (219) were under 65.
While turnout is higher among the available population of over 65's, that group still makes up a significant minority of the voting age population and turn out among the over 65s is not 100 percent. There is no way that it makes sense for more than 40 percent of those polled to be over that age threshold.
NorthCarolina
(11,197 posts)Just kidding....thanks for the synopsis.
NurseJackie
(42,862 posts)... depending on who's giving them. They're smart enough to know if they've "overpolled" a certain demographic, and "underpolled" another demographic, then they'll give the overpolled group less weight (depending on other historical data and actual distribution of the different groups within the population.)
Thus, the "margin of error" that you've heard so much about.
Jarqui
(10,130 posts)But I haven't seen it
When you're crossing a couple of tables or using one table to break out another, you would think something wouldn't add up because of that smoothing. But it always seems to - working with the lopsided demographics they provided.
Sancho
(9,070 posts)Hillary continues to gain endorsements at an all time record pace from members of Congress and Governors.
Hillary is wrapping up union support every which way.
Hillary is winning the debates by all neutral assessments.
Hillary is killing the polls, particularly in the Sunbelt leading up to Super Tuesday.
Bernie is struggling to keep control of his staff and campaign. Investigation ensuing.
O'Malley is in the dump.
The GOP has found someone crazier and dumber than Palin to nominate.
Maybe I should buy some bubbly?
virtualobserver
(8,760 posts)JaneyVee
(19,877 posts)Sancho
(9,070 posts)have you looked at the LV (likely voter) polls - those who are registered and vote in every elections and the ones that the Sanders campaign wants to steal? Have you looked at the committed super delegate count? Oh well, probably not.
Seriously in favor of Hillary. FL, NC, OH..that's the electoral votes.
Frankly, no one cares about HN since there are less voters in NH than in my county in FL.
Geraldine Ferraro had really big rallies!!! What happened to her? Who?
There are 40 ball games with 10 times more people than any candidate next weekend. Have fun at the rallies. Get excited at the rallies. Be a Hillary Hater if you want.
Just for fun:
Nevada:
CNN/ORC 10/3 - 10/10 253 LV 50 34 12 0 0 0 Clinton +16
Gravis Marketing 7/12 - 7/13 416 RV 55 18 5 1 1 0 Clinton +37
Massachusetts:
Boston Globe/Suffolk 11/19 - 11/22 241 LV 6.3 54 29 3 Clinton +25
Emerson College 10/16 - 10/18 265 LV 6.0 59 25 3 Clinton +34
Minnesota:
PPP (D) 7/30 - 8/2 426 RV 4.9 50 32 4 Clinton +18
Suffolk 4/24 - 4/28 100 LV 9.8 63 -- -- Clinton +48
PPP (D) 1/18 - 1/20 373 RV 5.1 59 -- 1 Clinton +45
North Carolina:
PPP (D) 12/5 - 12/7 555 RV 4.2 60 21 10 Clinton +39
Elon University 10/29 - 11/2 514 RV 4.3 57 24 3 Clinton +33
Oklahoma:
Sooner Poll/News 9/News on 6 11/12 - 11/15 389 LV 5.1 47 12 2 Clinton +35
The Oklahoman 10/19 - 10/22 350 RV 5.4 30 21 1 Clinton +9
Louisiana:
WWL-TV/Clarus 9/20 - 9/23 57 22 7 2 1 0 Clinton +35
Michigan:
FOX 2 Detroit/Mitchell 9/18 - 9/18 538 LV 35 28 22 -- -- -- Clinton +7
MRG 9/9 - 9/14 600 LV 41 22 22 0 1 0 Clinton +19
Florida:
Florida Atlantic University 11/15 - 11/16 297 LV 5.6 66 22 4 Clinton +44
Bay News 9/News 13 10/28 - 11/1 922 LV -- 66 24 3 Clinton +42
UNF 10/14 - 10/19 632 RV 3.9 55 16 0 Clinton +39
Missouri:
PPP (D) 8/7 - 8/9 352 LV 5.2 53 25 5 5 1 Clinton +28
Ohio:
Quinnipiac 9/25 - 10/5 40 21 19 1 0 0 0 Clinton +19
Quinnipiac 8/7 - 8/18 47 14 17 -- 0 1 0 Clinton +30
Quinnipiac 6/4 - 6/15 60 13 10 -- 1 0 0 Clinton +47
PPP (D) 6/4 - 6/7 61 -- 13 -- 2 1 2 Clinton +48
Quinnipiac 3/17 - 3/28 54 9 3 -- 3 2 -- Clinton +40
Quinnipiac 1/22 - 2/1 51 7 5 -- 1 0 -- Clinton +37
virtualobserver
(8,760 posts)can they vote from that chair?
Sancho
(9,070 posts)and we all have a land line (for hurricanes and alarm systems, etc.). We all have caller ID too, so we know it's a poll calling.
virtualobserver
(8,760 posts)walking in lockstep
Sancho
(9,070 posts)Lots of immigrants, lots of women, lots of union members, and lots of minority.
virtualobserver
(8,760 posts)We have a primary and a caucus on the same day......so the mail won't work for me.
brooklynite
(94,719 posts)Answer -- none.
I'm sure it's nice to rev up the fans you already have, but some people feel it's important to go out and convince people to vote for you.
virtualobserver
(8,760 posts)We know where Bernie has stood on issues for decades......Trying to figure out where Hillary stands is the big challenge for voters,
and trying to generate excitement within the voting population is Hillary's challenge.
brooklynite
(94,719 posts)Something other rally candidates didn't seem to achieve...
virtualobserver
(8,760 posts)we have the excitement of 2 million plus donors, huge crowds and the much maligned online polls.
you have polls indicating a lack of trustworthiness.
Thank god that you have Debbie Wasserman Schultz working for you, limiting the debates and picking
the worst possible dates for the meager number that she has scheduled.
brooklynite
(94,719 posts)I'll start with that.
virtualobserver
(8,760 posts)brooklynite
(94,719 posts)Pretty exciting.
virtualobserver
(8,760 posts)I am happy for them.
We shall see how the polls hold up.
Iliyah
(25,111 posts)Matter of fact the whole lot of them (GOP) are nuttier than hell.
workinclasszero
(28,270 posts)Go Hillary!
Alfresco
(1,698 posts)Happy Holidays!
coyote
(1,561 posts)67.4% of the poll participants were over 50 years old. Again the millennial are totally underweighted. The demographics simply do not match past presidential elections. The poll is trash.
NurseJackie
(42,862 posts)... calculate the numbers appropriately. Over sampled demographics count less, under sampled demographics count more based on actual population demographics and voter turnout likelihood.
ismnotwasm
(41,999 posts)Impressive indeed!
HoneychildMooseMoss
(251 posts)I don't even know where she actually stands, because she keeps changing her position.
Fuck that.
I want someone I can count on.
That's Bernie.
seabeyond
(110,159 posts)Alfresco
(1,698 posts)Happy Holidays!