Federal Election Polls

[WWWTT]

Also very hard to believe that all of a sudden, overnight the conservative support can jump near 10%!

Do 10% of voters have kids and are suddenly persuaded the exact same day that they are to allegedly receive the cheques and then receive a telephone poll?

WWWTT

[WWWTT]

Haven't seen another repeat poll putting the conservatives at 38% yet.

In fact, the poll I linked below shows the conservatives and liberals tied for second at 27%!

http://poll.forumresearch.com/post/324/ndp-minority-seen/

WWWTT

[waldo]

apparently, notwithstanding the criticism of the methodology used by the relative unknown Mainstreet polling company (particularly with regard to national polling), that Mainstreet poll only presented results from decided voters, not including the key segment of so-called leaning voters. Imagine that!

still haven't seen Postmedia outlets present anything on the Forum poll (that had the Conservatives a percentage point less than the Liberals), released just one day prior to that Mainstreet poll they hyped to no-end and still present prominently. Imagine that!

Forum poll: NDP (34%), Liberals (29%), Conservatives (28%)

[WWWTT]

This picture says a lot!

https://www.facebook.com/theglobeandmail/photos/a.157369913903.114299.140961138903/10153413973073904/?type=1

WWWTT

[Bryan]

Also very hard to believe that all of a sudden, overnight the conservative support can jump near 10%!

Do 10% of voters have kids and are suddenly persuaded the exact same day that they are to allegedly receive the cheques and then receive a telephone poll?

WWWTT

It's probably a combination of several factors. The UCCB surely helped to boost CPC numbers, but so did the NDP bringing up coalition talk. A weakening global economy also historically favours the incumbent (even those who favour change start feeling like this might be the wrong time to do it.)

But keep in mind that most of the polls we have been seeing are by pollsters like EKOS who have a history of exaggerating opposition support. As Derek pointed out here: #2994 , you generally have to add between 5 and 8 points to their CPC numbers to get what actually happens on voting day. That alone makes the Mainstreet numbers (if accurate) actually not much of a change at all.

[WWWTT]

But keep in mind that most of the polls we have been seeing are by pollsters like EKOS who have a history of exaggerating opposition support. As Derek pointed out here: #2994 , you generally have to add between 5 and 8 points to their CPC numbers to get what actually happens on voting day. That alone makes the Mainstreet numbers (if accurate) actually not much of a change at all.

LOL!

You and Derek2.0 can make up whatever you want, believe it,and then spew it out as many times as you want!

http://rabble.ca/blogs/bloggers/djclimenhaga/2015/07/two-polls-widely-different-results-and-postmedia-only-reports-on

Mainstreet is obviously a conservative partisan hack poll that other pollsters shutter at.

Most didn't even receive their cheques yet when the poll was conducted?!?!?!?! Oh and those allegedly receiving a cheque account for an 10% surge ahead?!?!?!

Lets see how well Harper does at the televised debates.

WWWTT

Edited July 25, 2015 by WWWTT

[Bryan]

You and Derek2.0 can make up whatever you want, believe it,and then spew it out as many times as you want!

Derek didn't make anything up. You can go and look at the polling leading up to any election.

Here are the numbers for 2011: https://en.wikipedia.org/wiki/Opinion_polling_in_the_Canadian_federal_election,_2011

The votes that CPC actually gets are consistently higher than most of the pollsters report. When you see EKOS or Forum report CPC support, you have to add 4-6 points to get the real number.

Look, I'm not defending Mainstreet, I know nothing about them. I'm just saying that their numbers don't show as big of a jump as it might appear.

[waldo]

Look, I'm not defending Mainstreet, I know nothing about them. I'm just saying that their numbers don't show as big of a jump as it might appear.

huh! A 10 point jump from other pollsters isn't huuuuuge... including Forum that released results one day prior to the Mainstreet poll; results that had the Harper Conservatives in 3rd place. A poll that has yet to see the light of day, as I'm aware, in any Postmedia outlet! Again, the announced Mainstreet poll that only includes results from decided voters, not 'leaning' voters... the Mainstreet polling process that is being criticized for it's approach... the Mainstreet polling company that has no real proven track record, particularly in a national context.

.

[Bryan]

huh! A 10 point jump from other pollsters isn't huuuuuge...

It's not, and I fully explained why. If EKOS says CPC support is 31%, that consistently has been shown to mean that it's really 35 or 36. Mainstreet is only a showing a couple of points above that.

Again, the announced Mainstreet poll that only includes results from decided voters, not 'leaning' voters... the Mainstreet polling process that is being criticized for it's approach..

That's demonstrably false. The poll shows both decided and undecided (as well as several other factors), and gives detailed breakdowns for each. Polling process? EKOS and Forum both use IVR too.

[WWWTT]

Derek didn't make anything up. You can go and look at the polling leading up to any election.

Here are the numbers for 2011: https://en.wikipedia.org/wiki/Opinion_polling_in_the_Canadian_federal_election,_2011

Yes he did make stuff up. Not the first time either.

Lots of polls showed the conservatives last election well within their margin of error.

Also, Derek 2.0 implies that this low conservative polling numbers of 4-5% is consistent with every federal election since the birth of conservatives. Another false.

Polls are not an exact science, just an indicator.

But according to the both of you, polls are now an exact science where the conservatives can never lose! Even when the polls show the conservatives are losing!

And this is probably the most comedic things going on in MLW right now.

WWWTT

[waldo]

It's not, and I fully explained why. If EKOS says CPC support is 31%, that consistently has been shown to mean that it's really 35 or 36. Mainstreet is only a showing a couple of points above that.

That's demonstrably false. The poll shows both decided and undecided (as well as several other factors), and gives detailed breakdowns for each. Polling process? EKOS and Forum both use IVR too.

ya ya, you claimed EKOS bias... and completely blew off, and continue to ignore the reference I made to the Forum poll that placed Harper Conservatives in 3rd place. If you want to speak of presumed bias, you could also comment on Postmedia's actions.

if, as you say, it's "demonstrably false", you should... demonstrate it. That Mainstreet poll may show many things... but what was reported and hyped by Postmedia is the isolated decided vote ONLY. There is no "one of" IVR process... identical process... that all polling companies use. Each companies IVR methodology is unique to their process, particularly in terms of randomizing, identifying cell customers, call-back, including cell phones, etc..

[Bryan]

ya ya, you claimed EKOS bias... and completely blew off, and continue to ignore the reference I made to the Forum poll that placed Harper Conservatives in 3rd place. If you want to speak of presumed bias, you could also comment on Postmedia's actions.

Forum is not much different, they also consistently under report CPC support, right up until election night.

Post is displaying a break from the bias it used to have, by not using a company that has proven itself conclusively to be unreliable.

if, as you say, it's "demonstrably false", you should... demonstrate it.

The links have been provided several times, it's not my fault you won't read them, nor my job to make you. If you deliberately choose to live in ignorance, good for you.

[Bryan]

Lots of polls showed the conservatives last election well within their margin of error.

None of the pollsters said that their MOEs were anywhere near as big as how far off they were.

Also, Derek 2.0 implies that this low conservative polling numbers of 4-5% is consistent with every federal election since the birth of conservatives. Another false.

That is exactly what has happened. Derek and I both posted the links for you. Look at the pre-election numbers, then look at the election results: CPC real support is ALWAYS higher than the most of the polls say they will be (especially EKOS and Forum).

Polls are not an exact science, just an indicator.

An indicator off by 4-6% to the left almost every time.

But according to the both of you, polls are now an exact science where the conservatives can never lose!

They are not exact at all, they are consistently wrong, well outside of their stated MOE.

[Evening Star]

You know, even if 40% of the population voted CPC on May 2, 2011, that doesn't necessarily mean that Forum was wrong to report 35% support on April 30 or even 36% on May 1 (with a 1.6% margin of error). People could have changed their minds in between those times for a number of reasons, even on Election Day itself. CPC voters could have been more motivated to actually go out and vote, etc. If your argument was "the CPC tends to do a little better on Election Day than the polls show leading up to it so we can adjust poll numbers a little to account for this", I could understand that. However, these factors would apply to all polls, including Mainstreet, unless you have reason to believe that something about the polling methodology actually underrepresents CPC support and that Mainstreet is doing something different in their methodology that fixes this. (Since you have admitted that you don't know much about Mainstreet, I don't think you can say this.) I don't think you can really say "the Mainstreet poll shows the CPC 10% higher than the other polls BUT we can automatically add 5% to CPC numbers in the other polls anyway so Mainstreet is not that far off".

[waldo]

Forum is not much different, they also consistently under report CPC support, right up until election night.

Post is displaying a break from the bias it used to have, by not using a company that has proven itself conclusively to be unreliable.

all other polling companies are biased against the Harper Conservatives? You self-acknowledged you knew nothing about Mainstreet... but by-golly, by-goshum... you sure likee that company now, hey!

.

The links have been provided several times, it's not my fault you won't read them, nor my job to make you. If you deliberately choose to live in ignorance, good for you.

don't eat that Elmer, that there's bull-shyte! I'll ask you more directly/pointedly: is it your claim that the much hyped Postmedia articles are not specifically targeting only decided voters from that Mainstreet poll... to the exclusion of undecided/leaning voters? Is that what you're claiming?

.

[Evening Star]

None of the pollsters said that their MOEs were anywhere near as big as how far off they were.

All of the Nanos, Harris-Decima, and Abacus polls from the few days before (Apr 28-May 1) accurately predicted CPC support, within the margin of error, using your own Wiki link.

An indicator off by 4-6% to the left almost every time.

Most of the polls you linked from the same time period predicted the NDP's support accurately, within the MOE, so I disagree with this.

Edited July 26, 2015 by Evening Star

[Bryan]

If your argument was "the CPC tends to do a little better on Election Day than the polls show leading up to it so we can adjust poll numbers a little to account for this", I could understand that.

Sure, I can go with that.

However, these factors would apply to all polls, including Mainstreet, unless you have reason to believe that something about the polling methodology actually underrepresents CPC support and that Mainstreet is doing something different in their methodology that fixes this. (Since you have admitted that you don't know much about Mainstreet, I don't think you can say this.)

You can't say it either way, because unlike the other pollsters, we haven't seen the difference between their polls and election results. I was simply pointing out that their numbers aren't as far divergent from the other recent polls as it appears at first glance, not that I have any insight as to their accuracy.

[Bryan]

I'll ask you more directly/pointedly: is it your claim that the much hyped Postmedia articles are not specifically targeting only decided voters from that Mainstreet poll... to the exclusion of undecided/leaning voters? Is that what you're claiming?

Absolutely, yes. Both decided and undecided voters are factored in, and both sets of data are presented.

http://news.nationalpost.com/news/canada/canadian-politics/conservative-support-among-voters-jumps-as-childcare-cheques-doled-out-poll

[waldo]

Absolutely, yes. Both decided and undecided voters are factored in, and both sets of data are presented.

absolutely? What screamed out in the Postmedia headlines... that's right, the 38% figure... and those graphics appear to only exist in the NP - example of headine/no graphics - here: What got carried forward in other non-Postmedia headlines... that's right, the 38% figure. What's freakin' incredible is the "journalist" can't even get the poll specifics right - here, let the waldo help you out, directly as linked from the Mainstreet website: 34% compared to 30% for both the NDP & Liberals. Talk about a sweet Postmedia ride, hey!

[Bryan]

absolutely?

Yes. ABSOLUTELY. The undecided numbers are right there in your own Edmonton Journal link:

Factoring in undecided voters, 30 per cent said they would vote for the Conservatives if a federal election were held today, with the NDP at 22 per cent and Liberals at 20 per cent. Five per cent of respondents said they would support the Green party and three per cent the Bloc.

Same as the graphic I posted. Same as the actual poll data.

[waldo]

Yes. ABSOLUTELY. The undecided numbers are right there in your own Edmonton Journal link:

Same as the graphic I posted. Same as the actual poll data.

you are being anal and disingenuous! Again, what figure was blasted across media, whether Postmedia newspapers or other? Again, the 38% figure. Again, as I just linked to you directly... Mainstreet's own article speaks to the 34% number when undecided are factored... as compared to 30% for both the NDP and Liberal.

[Bryan]

you are being anal and disingenuous! Again, what figure was blasted across media, whether Postmedia newspapers or other? Again, the 38% figure.

Both decided and undecided numbers are in the article. It's nobody's fault if you don't read them.

Mainstreet's own article speaks to the 34% number when undecided are factored... as compared to 30% for both the NDP and Liberal.

That is not the undecided data, those are the combined first and second choice numbers. Last time you voted federally, were you asked to combine or rank your choices, or to pick just one?

Edited July 26, 2015 by Bryan

[waldo]

Both decided and undecided numbers are in the article. It's nobody's fault if you don't read them.

again, you're being (purposely) disingenuous! You choose to ignore Postmedia headlines were misleading; that the results were hyped to singularly focus on decided voters only. That other media outlets (newspaper and TV) only presented or emphasized the same misleading reference.

That is not the undecided data, those are the combined first and second choice numbers. Last time you voted federally, were you asked to combine or rank your choices, or to pick just one?

and it represents the linked verbiage per the interview with the head of Mainstreet... where the guy only speaks to the same decided only vote. Interestingly, per my reference highlight, the guy also speaks to their "Momentum Tracker"... something that gets absolutely no play by Postmedia. Again, the 34/30/30 percentages - go figure, hey!

Edited July 26, 2015 by waldo

[Bryan]

again, you're being (purposely) disingenuous!

I'm being 100% upfront and honest, which is 180 degrees from where you're coming from. Nobody claimed anything about a momentum tracker, or combining first and second choices. YOU claimed the undecided data was not in the Post stories, and that was categorically false-- It's right there, in the first few paragraphs. Moving the goal posts by pretending that you meant something else doesn't change that.

[WWWTT]

None of the pollsters said that their MOEs were anywhere near as big as how far off they were.

LOL!

You never even looked at the link you provided!

Nor have you provided any links for any other previous elections!

Don't let the facts get in the way of your partisan projection of reality.

WWWTT