Showing posts with label Forum Research. Show all posts
Showing posts with label Forum Research. Show all posts
Thursday, May 22, 2014
New Forum Poll Shows OLP Moving Ahead
Thursday, January 30, 2014
Forum Research Fucks Up, Admits Error
From ThreeHundredEight.com:
UPDATE: As appeared to be the case, Forum Research has confirmed there was an error in its two reports that listed the numbers that should have been Trudeau's as Mulcair's and the numbers that should have been Mulcair's as Trudeau's. So, the error crept in when Forum took the data from their polling and transcribed it into their reports, both of which contained analysis as to why Mulcair was polling ahead of Trudeau on this question.
Worth noting when you consider that their Ontario polling tells a very different story from that told by pollsters who fuck up considerably less often.
UPDATE: As appeared to be the case, Forum Research has confirmed there was an error in its two reports that listed the numbers that should have been Trudeau's as Mulcair's and the numbers that should have been Mulcair's as Trudeau's. So, the error crept in when Forum took the data from their polling and transcribed it into their reports, both of which contained analysis as to why Mulcair was polling ahead of Trudeau on this question.
Worth noting when you consider that their Ontario polling tells a very different story from that told by pollsters who fuck up considerably less often.
Monday, December 02, 2013
Dan On Polling
Probably the best piece I've read on the Forum Polling cock-up in Manitoba:
Instead, what we’re dealing with appears to be flawed methodology. Bozinoff has admitted that some respondents may have been called for 3 consecutive polls, and that likely wouldn’t have happened unless the response rate was in the neighbourhood of 1% (typical for robo-polls, when you don’t do callbacks). Heck, Sunday being the Grey Cup, it may have been even lower. Sampling methodology only works if you assume survey respondents are similar to the public at large – otherwise, these polls are no more accurate than the “self selecting” click polls you see on websites, asking what you think of Miley Cyrus’ antics.
Read the whole thing, and next time you see a poll from Forum Research, just say to your self "Oh, that's just a poll from Forum Research!"
Instead, what we’re dealing with appears to be flawed methodology. Bozinoff has admitted that some respondents may have been called for 3 consecutive polls, and that likely wouldn’t have happened unless the response rate was in the neighbourhood of 1% (typical for robo-polls, when you don’t do callbacks). Heck, Sunday being the Grey Cup, it may have been even lower. Sampling methodology only works if you assume survey respondents are similar to the public at large – otherwise, these polls are no more accurate than the “self selecting” click polls you see on websites, asking what you think of Miley Cyrus’ antics.
Read the whole thing, and next time you see a poll from Forum Research, just say to your self "Oh, that's just a poll from Forum Research!"
Thursday, November 28, 2013
There's Another Forum Poll Out
But I'm not going to link to it or even tell you what its about. Because why bother? I will just say that the one media outlet that's picked it up seems ashamed of the fact. At least, they're not promoting it anywhere.
Tuesday, November 26, 2013
BREAKING NEWS: By-Election Results Offer No Compelling Narrative!!!
Status Quo. Close in Brandon. Whumped 'em in T.O. Center and Bourassa LPC increased its vote share, so that's nice but no big thng. Didn't I tell you so? I did!
PS. Dear Forum Research, your crappy polls were only off by like 30 percent in Brandon. You officially suck.
PS. Dear Forum Research, your crappy polls were only off by like 30 percent in Brandon. You officially suck.
Monday, November 25, 2013
Darrell Bricker, CEO of Ipsos Public Affair, On Recent Forum Polls
The occasion is this story in the Winnipeg Free Press:
Forum released its latest round of polling late Sunday night following calls into the riding during the day. The survey has a relatively small sample size but suggests Liberal Rolf Dinsdale enjoys 59 per cent support, a commanding lead over Conservative Larry Maguire’s 30 per cent.
But, several Brandon-Souris residents report being polled multiple times throughout the campaign as Forum conducted several waves of polling using random, automated robocalls, where voters indicate their candidate preference using their touch-tone phone.
Brandon resident Anne Ross has received six calls from Forum since the campaign began, raising questions about the poll’s reach and its randomness. Ross was called for each of the five polls released so far, including the latest one. She also received a call on Saturday as part of a poll that has not been released.
"I don’t know why I would keep getting called," said Ross, a Liberal.
Ross is the mother of Free Press opinion columnist Deveryn Ross, who said he has been called by Forum Research six times during the campaign, including three times since Friday night.
Two other voters who spoke to the Free Press Sunday, including one Conservative, said they also received multiple calls from Forum Research.
There are 62,000 registered voters in the riding.
Forum president Lorne Bozinoff calls interactive voice-response polls "amazingly accurate." He said while it’s unusual someone might get called six times in a random poll, he doesn’t believe that undermines the poll’s results.
This inspired a series of skeptical tweets from Bricker:
I think what he's getting at is easy enough to follow, but here's a definition of "empaneling". I assume he's wondering if they are not recalling the same list and tracking how voting intentions change during the run-up to the by-election. Not that doing it this way would make the polls wrong, but presumably you would evaluate them differently from a random truly sample.
If that's what Forum is doing. Nobody knows what they're doing. Just that this kind of thing shouldn't happen.
Forum released its latest round of polling late Sunday night following calls into the riding during the day. The survey has a relatively small sample size but suggests Liberal Rolf Dinsdale enjoys 59 per cent support, a commanding lead over Conservative Larry Maguire’s 30 per cent.
But, several Brandon-Souris residents report being polled multiple times throughout the campaign as Forum conducted several waves of polling using random, automated robocalls, where voters indicate their candidate preference using their touch-tone phone.
Brandon resident Anne Ross has received six calls from Forum since the campaign began, raising questions about the poll’s reach and its randomness. Ross was called for each of the five polls released so far, including the latest one. She also received a call on Saturday as part of a poll that has not been released.
"I don’t know why I would keep getting called," said Ross, a Liberal.
Ross is the mother of Free Press opinion columnist Deveryn Ross, who said he has been called by Forum Research six times during the campaign, including three times since Friday night.
Two other voters who spoke to the Free Press Sunday, including one Conservative, said they also received multiple calls from Forum Research.
There are 62,000 registered voters in the riding.
Forum president Lorne Bozinoff calls interactive voice-response polls "amazingly accurate." He said while it’s unusual someone might get called six times in a random poll, he doesn’t believe that undermines the poll’s results.
This inspired a series of skeptical tweets from Bricker:
Anybody else receive multiple calls from these guys? In "random" survey this would be almost impossible. http://t.co/tKNln2G1XF
— Darrell Bricker (@darrellbricker) November 25, 2013
And:
REALLY? Somebody was hit 6 times randomly? Twice, maybe, but 6 times? And multiple people multiple times? http://t.co/tKNln2G1XF
— Darrell Bricker (@darrellbricker) November 25, 2013
And especially:
Journos here's the question to ask - are you empaneling respondents and calling them multiple times? http://t.co/tKNln2G1XF
— Darrell Bricker (@darrellbricker) November 25, 2013
I think what he's getting at is easy enough to follow, but here's a definition of "empaneling". I assume he's wondering if they are not recalling the same list and tracking how voting intentions change during the run-up to the by-election. Not that doing it this way would make the polls wrong, but presumably you would evaluate them differently from a random truly sample.
If that's what Forum is doing. Nobody knows what they're doing. Just that this kind of thing shouldn't happen.
@Bigcitylib2 Saying that responsible pollsters clearly communicate how they do their polls. They wouldn't leave anyone confused like this.
— Darrell Bricker (@darrellbricker) November 25, 2013
Tuesday, March 05, 2013
Kathleen Wynne's Done Good
Full disclosure. It's A Forum Research poll. That said, I can't but like the underlying message:
...the Liberals would again win 53 seats in the 107-member house with the Tories taking 36, down from 37 in the most recent election and the NDP with 18, up from 17 in that vote.
One odd (to me) result:
...the New Democrats have tapped into one possible election-winning issue: reducing auto insurance premiums by 15 per cent.
Last Friday, Forum polled 1,033 people and found 58 per cent supported Horwath’s plan to trigger a vote if Wynne does not move to cut rates. Less than a third — 33 per cent — disapprove of the NDP threat and 11 per cent didn’t know.
...the Liberals would again win 53 seats in the 107-member house with the Tories taking 36, down from 37 in the most recent election and the NDP with 18, up from 17 in that vote.
One odd (to me) result:
...the New Democrats have tapped into one possible election-winning issue: reducing auto insurance premiums by 15 per cent.
Last Friday, Forum polled 1,033 people and found 58 per cent supported Horwath’s plan to trigger a vote if Wynne does not move to cut rates. Less than a third — 33 per cent — disapprove of the NDP threat and 11 per cent didn’t know.
Frankly, the last thing on my mind is auto-insurance rates, and I've always figured I was pretty typical in this regard. Weirdly enough, the OFL's Sid Ryan seems to have got this one wrong too. I guess Ms. Horwarth has better populist instincts than either of us.
Another interesting result from Forum:
Not exactly surprising. I've heard pro-teacher sentiments from people I'd of thought would be unsympathetic But it tells you have far we have come from the 1990s when stirring up anti-teacher and more generally anti-union sentiment was a big part of the PCPO's political strategy. These days Hudak plays the same cards and comes up snake-eyes.
Wednesday, October 31, 2012
They Call Them Political Parties...
...because you always have to look like you're having one. And Justin does. Sure his party sometimes looks like its going to become an all night kegger, but he keeps advisers around to make sure that doesn't happen. Never underestimate how far an outgoing, enthusiastic personality can take you in politics.
As for Forum, does it even matter anymore if their polls are crap, since crappy polls are all that's on offer these days? They're basically treating their surveys as adverts for the company, and its working. Read the papers these days and you'd think they're the only polling firm in Canada.
As for Forum, does it even matter anymore if their polls are crap, since crappy polls are all that's on offer these days? They're basically treating their surveys as adverts for the company, and its working. Read the papers these days and you'd think they're the only polling firm in Canada.
Tuesday, August 28, 2012
A New Poll From Forum
...who poll relentlessly, if not always accurately:
In a random sampling of public opinion taken by the Forum Poll among 1,720 Canadians 18 years of age or older, one third would vote either for the federal NDP or the Conservatives (34% each) if an election were held today. Just fewer than one quarter would vote Liberal (22%). Results of the poll are considered accurate plus or minus 2.36%, 19 times out of 20.
That's it. That's all there is. Can't even find deets at the Forum website.
In a random sampling of public opinion taken by the Forum Poll among 1,720 Canadians 18 years of age or older, one third would vote either for the federal NDP or the Conservatives (34% each) if an election were held today. Just fewer than one quarter would vote Liberal (22%). Results of the poll are considered accurate plus or minus 2.36%, 19 times out of 20.
That's it. That's all there is. Can't even find deets at the Forum website.
Friday, March 04, 2011
On The Credibility Of Polls
This one has the McGuinty Libs and Hudak Tories virtually tied in Toronto.
And if you question the credibility, its by Forum Research who, according to their own website, conduct more than 600,000 surveys each year, or over 1,600 every day. So you know they can be trusted.
And, oh hey! This is the same company that puts Rob Ford's approval rating at 60%. But they do 1,600 of these suckers every 24 hours. They did 300 before you even ate breakfast. So who are you to doubt their authority?
h/t. PS. No, we're not married. Sometimes our interests converge. That's all.
And if you question the credibility, its by Forum Research who, according to their own website, conduct more than 600,000 surveys each year, or over 1,600 every day. So you know they can be trusted.
And, oh hey! This is the same company that puts Rob Ford's approval rating at 60%. But they do 1,600 of these suckers every 24 hours. They did 300 before you even ate breakfast. So who are you to doubt their authority?
h/t. PS. No, we're not married. Sometimes our interests converge. That's all.
Subscribe to:
Posts (Atom)