01.19.2012 08:58 AM

Bucket post

I’m back – and, it looks like online protest does work, sometimes. Nice.

Anyway, I’m not saying that Forum Research is a bucket shop. I’m not saying that. I will, however, draw your attention to this and this and, of course, this. Never mind the fact that the pollsters, who use a discredited methodology (IVR, or robo-polling), actually show the federal Libs behind the leaderless NDP.

As my friends John Wright and Darrell Bricker have written, media organizations are getting sloppier in their reporting of these public opinion surveys.  And political organizations should therefore remember that many of these polls are worth what you pay for them.

Nothing.

16 Comments

  1. Lance says:

    WK, brace yourself for the pile on in 3…………2…………..1…………

  2. Marc says:

    Gord, everyone I’ve asked this morning things your post is stupid and you should lose internet privileges for one week.

    Aren’t you interested in my polling methods? Would knowing that I was the only respondent affect how you view the poll? Wouldn’t you demand to have more people who represent a wider array of viewpoints be asked? Wouldn’t you like to know the context of the poll question?

    • George says:

      I’m a conservative Marc…and even I agree with your post. Gord needs his own blog and he sure doesn’t speak for me!

      • Jason King says:

        Im pretty sure that there would be enough of us around even to help make Tulktalks.com a reality. Anyone want to host a wordpress blog for Mr Tulk?

        Put your money where your mouth is Gord. Get a blog and lets actually see how popular you are.

        Any other way is just being a chicken

        • Ted B says:

          That’s for sure.

        • Jason King says:

          Interesting Gord, so I take it you realize that your commentary, put on its own blog, wouldnt be read by many users.

          Im not surprised. Conjecture and generalization wrapped in a pompous self righteousness is pretty tiring online.

          At least your persistent, and consistent.

  3. Marc says:

    (PS – I don’t really think your post is stupid or that you should lose internet privileges… just demonstrating a point 🙂 )

  4. Ted says:

    We should take the same approach in all fields.

    Why require footnotes and research for a PhD dissertation in the humanities. Just go with your gut! And since Harper has gotten rid of the long form census, and it will cost us millions due to inaccurate information anyway, why not go further and just use guess work to determine the population and demographics.

    And why not apply the same approach to pure science? Why should a bunch of Nobel prize winners and PhD brainiacs relying on hundreds of years of refining methodology be dictating to the rest of us how “science” must be practiced. It smacks of elitism. What is wrong with having that guy out in Alberta declare he’s proven that evolution is a fraud because he was able to fossilize a teddy bear. Why should we impose peer review and regulation on such genius.

    To protect us from rogues, we just need to write a couple of Latin phrases. That’s all.

    Established tandards and accepted norms and rigorous methodology are elitist and undemocratic.

    And 99 out of 100 Canadians agree with me.

  5. billg says:

    Ted: When did Harper get rid of the Long Form Census? And your worried about polls?

    • CanadaDave says:

      Sure does! It’s just now voluntary, self selecting, and statistically worthless.

      • Ted says:

        I wouldn’t go so far as to say the NHS is worthless. All data has some value when you are making decisions.

        But it is certainly far less valuable than a census, far less accurate, far less reliable.

        And as a result it will increase waste and inefficiencies in government.

        My brother is a water engineer. When they go to design their projects, they rely on historical data on soil, weather, precipitation, water levels, etc. Most often this is provided by the provincial government. In his world, to this day, they still talk about the “dark years” of Mike Harris: not as any commentary on his government, but on the lack of reliable data from those years because of Harris’s cut-backs. They have to build in assumptions and buffers into their designs that add to costs. They also had to seek out less perfect data from multiple different sources which also adds to costs. They won’t build something they can’t stand behind so there is no danger, but it is definitely less efficient and more costly.

        THAT is what Harper has replicated, but on a much much larger scale by getting rid of the Long Form Census.

        THAT is what happens when ideology trumps good policy.

      • Ted B says:

        Actually, it did, Gord.

        You had extremely high response rates as a result. Negligible margins of error. High reliability.

        A survey? Not so much.

    • Ted says:

      The Long Form Census was replaced by the National Household Survey by Harper.

      Surveys and censuses are not the same thing, though close.

      A census attempts to produce, as much as possible, an actual representation of the demographics, not merely a sampling. Some extrapolation does end up being required, but the results are very solid and reliable.

      The NHS is a survey with survey methodologies built in. The results are far less reliable as much more extrapolation based on random sampling is required. Harper’s decision, like so many of his decisions, has made government decision making less precise and more wasteful and expensive.

      Ask any statistician and they will tell you the two are very different.

    • Ted says:

      Nope. Do please keep up with current events, Gord.

      The Long Form Census was replaced by the National Household Survey.

      A census and a survey are not the same thing.

      • Ted B says:

        That’s like saying it’s semantics to say the Toronto Maple Leafs and the Toronto Marlboroughs are different since they are from Toronto, play hockey, and wear blue and white.

        They aren’t the same thing because they aren’t the same thing.

  6. Tiger says:

    I was wondering what the slams on Forum came from — now I see, a poll putting the PCs up at 41% again with the Liberals down at 33%.

    Well, whatever. NDP’s still in third in that one, and so Horwath has no reason to force an election. And McGuinty has (alas) turned it around during campaigns before.

    Mind you, it’ll be tougher to portray Hudak as not ready for the job in the next campaign — smart parties stick with their leaders after campaigns like the last one, and work on improving them.

    Next election, whenever it comes, should be fun.

Leave a Reply to Ted Cancel reply

Your email address will not be published.