Liberal support surges

[Cartman]

I thought that Martin was not very bright, but he might just pull off a majority if he keeps this up.

http://tinyurl.com/8ru5g

The federal Liberals had the support of 40 per cent of respondents in a new poll — virtually the same level of backing they received in rolling to their majority government in 2000.

The Leger Marketing survey, conducted Sept. 6-11, pegged Conservative support at 24 per cent, while the NDP stood at 15 per cent and the Bloc Quebecois at 13 per cent.

Ms. Marois said the poll revealed strong growth for the Liberals in Western Canada, including a jump of 16 percentage points in Alberta in two months and an increase of 14 percentage points in British Columbia.

“And as long as he waits at least until the Gomery report is out and as long as the Conservatives still have Stephen Harper leading the party, his chances can only improve.”

[Canuck E Stan]

Interesting,and it could mean a majority for Martin,but what is this?

"The numbers were reached after distribution of the 20 per cent of respondents who were undecided."

Is this manipulation of those undecided into some kind of decided votes? I'm not familiar with this,what this mean?

[Cartman]

"The numbers were reached after distribution of the 20 per cent of respondents who were undecided."

Is this manipulation of those undecided into some kind of decided votes? I'm not familiar with this,what this mean?

It means that they do not believe the undecided will be differently distributed into any particular camp; no reason to believe that the undecided will all necessarily vote NDP, Liberal, Cons or BQ. It is possible, but unlikely that the undecided have any particular preference.

[Riverwind]

Is this manipulation of those undecided into some kind of decided votes? I'm not familiar with this,what this mean?

It means one of two things: you assume the undecides split exactly the way the decideds are going or you use statistics from previous elections which compare the actual vote with the polls immediately before the the election. I cannot remember exactly but I believe the latter approach favours the governing party because 'undecideds' tend to vote for the incumbants.

I do not see either approach as particularily manipulative as long as you know that the undecideds have been factored into the numbers.

[kimmy]

Avast, ye bilge rats. Since today is International Talk Like A Pirate Day, I will run my reply through the English-to-Pirate translator before I post it.

Arrr. A double-digit rise in 2 months in t' west? Based on what? What caused this big surge? That's too much t' be random drift, says I. Either thar be some explanation, or t' pollsters be scurvy dogs.

-k

[shoop]

It means one of two things: you assume the undecides split exactly the way the decideds are going or you use statistics from previous elections which compare the actual vote with the polls immediately before the the election. I cannot remember exactly but I believe the latter approach favours the governing party because 'undecideds' tend to vote for the incumbants.

I do not see either approach as particularily manipulative as long as you know that the undecideds have been factored into the numbers.

<{POST_SNAPBACK}>

While either approach may not be particularly manipulative, they definitely can be manipulated. Presenting the information this way is pretty misleading, as is the title of this thread.

In reality no much has changed on the public opinion front over the summer.

[stubblejumper]

go to http://www.legermarketing.com/eng/

They asked a 2nd question about current leanings.

Those where added in.

[Cartman]

Presenting the information this way is pretty misleading, as is the title of this thread.

In reality no much has changed on the public opinion front over the summer.

The title of this thread is not misleading. This survey documents a major increase in support for the Liberals particularly in BC and Alberta. Double digit increases are most worthy of being characterized as "surges". What would you call them, slight increases? As such, things have not remained the same but have changed considerably. Keep in mind that not too long ago, it appeared that the government might fall. But hey, anytime you want to spend hours generating a mutually agreed upon lexicon, I will contribute.

The information from this survey is also not presented in a misleading fashion. They clearly indicate the percentages for each party and identify those who responded that they were undecided. What are you supposed to do with the undecided when you are making inferences about what they will do when they go to the polls? It would clearly be inappropriate to claim they will all vote for one party or another assuming they vote at all. If they do not vote, then they are not relevant anyways and the results are valid and reliable. Regardless, it appears as though the Liberals are in good shape and I am not sure why Harper is daring Martin to call an election. I am sure the Liberals would really love to go to the polls right now considering the scandals do not seem to be bothering Canadians (particularly Western Canadians) as much these days.

For clarification, I support the NDP and the results do not exactly place them in a great position. Rather than complain about the sound methodology employed or how the results are interpreted, I accept that Layton has a lot of work to do. Don't blame the messenger.

[Montgomery Burns]

I saw this the other day. It's only a poll, but 16 points is a lot and cannot be easily dismissed.

It's distressing.

[newbie]

I'm hoping Martin gets in, just to excite and enrage our very own MB.

[Renegade]

I have mixed feelings about a majority Liberal government. On one hand I really don't agree with many of the liberal spending policies and in many ways they have backed away from the fiscal restraint which has kept them in power.

On the other hand, not supporting the Liberal's risks a minority government, and that minority is subject to blackmail from the NDP, which ended up costing taxpayers $5B this year.

Is it better to support a majorty you dislike or risk a minority which is potentially worse?

[shoop]

The title of this thread is not misleading. This survey documents a major increase in support for the Liberals particularly in BC and Alberta. Double digit increases are most worthy of being characterized as "surges". What would you call them, slight increases?

Problem 1 with methodology. Very small sample sizes for the regional breakdown. 125 respondents probably creates a margin of error greater than +/- 5%. This poll doesn't even acknowledge that, only releases the national margin of error. misleading

What poll are you comparing to? I went to the Leger site and read the results for this poll but didn't see any regional breakdown for previous polls.

But hey, anytime you want to spend hours generating a mutually agreed upon lexicon, I will contribute.

Will do as soon as you explain what the hell you mean by a mutually agreed upon lexicon.

What are you supposed to do with the undecided when you are making inferences about what they will do when they go to the polls?

You shouldn't assume they will break the way the decides broke. Here is a very quick example. In this poll 78% of respondents made a choice, 22% were undecided. You should leave it at that for a number of reasons. With a 22%

Rather than complain about the sound methodology employed or how the results are interpreted, I accept that Layton has a lot of work to do. Don't blame the messenger.

<{POST_SNAPBACK}>

With the limited information provided on methodology, how do you know it is sound?

My questions in my previous post weren't quoted to favour any one political party.

I have two guesses as to why this poll has been given a little more scrutiny than it deserves.

1. The media would love the rush of a Conservative leadership race before the next election. Would definitely sell them a lotta papers.

2. Most journalists, as with some posters, don't know how to read polls or the appropriate questions to ask about their methodology.

[Canuck E Stan]

Now don't get in a huff over this, but how does a polling company make money? Who paid for this poll? The pollsters seem to have some prime office space in this country. So they must be making good bucks doing these polls.

I don't believe Leger does these poll out of interest for the public,so somebody must of requested this poll. Or am I wrong?

Sometimes I think it's like getting an "independent" research report on how great a certain drug is and finding out the research is paid for by the drug company that makes it.I'm not disputing the findings, just looking for some answers about Polling companies.

[Cartman]

Now don't get in a huff over this, but how does a polling company make money? Who paid for this poll? The pollsters seem to have some prime office space in this country. So they must be making good bucks doing these polls.

I don't believe Leger does these poll out of interest for the public,so somebody must of requested this poll. Or am I wrong?

Sometimes I think it's like getting an "independent" research report on how great a certain drug is and finding out the research is paid for by the drug company that makes it.I'm not disputing the findings, just looking for some answers about Polling companies.

Well, that is a good question and I do not have the answer to it. I suspect media outlets probably pay them for such polls when there is little real news to report.

In terms of regional breakdowns, a small sample is not necessarily such a bad thing. It just raises the standard error, but most polls have such small regional samples. As there is greater support in BC and Alberta, I suspect they are measuring something other than error. They could have been deceitful and reduced the level of significance to keep the SE down though. By contrast, extremely large samples are problematic because everything becomes statistically significant.

In any case, this poll does not mean the Liberals will win the next election. It just shows support is high probably because many people have a Gomery hangover and that opposition parties will have their work cut out for them.

[Riverwind]

In any case, this poll does not mean the Liberals will win the next election.  It just shows support is high probably because many people have a Gomery hangover and that opposition parties will have their work cut out for them.

A signficant percentage of Liberal supporters could switch to the NDP if it is clear that the CPC has no chance of winning gov't. So I suspect any poll showing the Liberals far ahead will be followed by a poll showing the CPC in striking distance as strategic voters respond to the polls.

[Canuck E Stan]

Fair enough, but how can the valitity of the poll be truly accepted, when we don't know who asked for it to be done.

E.g. If a national newspaper asks for a poll to be done and the polling company used gives the results to the newspaper, and it is put in print.

Doesn't the paper expect some kind of "newsworthiness" change from the last poll it asked for?

What if the poll said there was no change from the last poll, that would hardly sell newspapers if people read the headlines in the paper.

Would not the pollsters come up with some kind of "change" in their polling method or have some figures to make some kind of change newsworthy? Not that it wasn't true, but somehow more dramatic in it's results.

What about pollsters who have affiliation to political parties?

I guess what I'm asking is how valid is a poll when we know nothing about who runs it or who wanted it, somewhat like the medical research done by drug companies.

[theloniusfleabag]

Dear Cartman and Canuck E Stan,

I'm not disputing the findings, just looking for some answers about Polling companies.

Well, that is a good question and I do not have the answer to it. I suspect media outlets probably pay them for such polls when there is little real news to report.

Check out the "Bid Requests" on the ledgermarketing link. That will tell you 'how' the polls were made. As to 'how they make money', they are a 'target market research group', and they help (larger) companies target potential consumers by collecting information on demographics. What you buy, how much you spend on what, etc is all information collected by many, many firms, whether you are aware of it or not. Hell, that is what 'internet cookies' are for...

So, imagine you are paying to have a poll conducted, and decide for yourself whether or not you want it to serve your interests, or do you want the truth... that is the secret of 'bid requests'.

[Bakunin]

leger marketing is very reliable for the region of quebec, in fact its the most reliable one when it come to know the intention vote in quebec but i guess they didn't do a good job for the entire canada, just like english pollers always do a shitty job for the region of quebec. Its comprehensible.

[Cartman]

The link is broken now but I thought it said that Leger is part of Gallup.

[shoop]

OP,

You referred to the 'sound methodology' of this poll yet have ignored the valid concerns with the methodology.

I suspect this poll may be close to accurate, yet still would like to know where your assertion of sound methodology is coming from.

ps you still haven't explained what you mean by a 'mutually agreed upon' lexicon.

[kimmy]

I say again, I'm skeptical. There's just too big a discrepency with a plethora of earlier polls, and no apparent explanation for what caused this discrepency.

The Liberals aren't 16% more popular in Alberta than they were 2 months ago. I'll believe LiberalMania is sweeping Alberta when I see dealcoholized beer or Roch Voisine records flying off the shelves.

-k

[stubblejumper]

I can't believe that I have to do this, people on this board should be more enlightened than this. Especially, if they are espousing their political ideas.

"Mutually agreed upon lexicon"- negotiated definitions of word or phrases that are vague or in dispute.

[shoop]

I can't believe that I have to do this, people on this board should be more enlightened than this. Especially, if they are espousing their political ideas.

"Mutually agreed upon lexicon"- negotiated definitions of word or phrases that are vague or in dispute.

<{POST_SNAPBACK}>

I can't believe that I have to do this, people on this board should be more enlightened than this. Especially, if they are going to react so condescendingly to questions.

Read post number 8 in this thread and you'll see how your definition doesn't clarify what the f**k Cartman meant with "mutually agreed upon lexicon."