On the weird Bristol poll

Jacques Poitras has a piece on his blog about the supposedly innovative Bristol poll this election where they will survey the same group of 600+ New Brunswickers to assess any changes in opinion over the course of the election cycle.

He and others are quite critical of this poll approach – as am I (as pointed out in a previous blog) but I don’t know that there has been much backlash in the traditional media. I would have expected a quote from a national polling firm slamming this model. Maybe I just didn’t see it.

The whole purpose of polls is their randomness. The minute you move out of that -all you have is a focus group. The only good that poll will do is it will tell the parties where those specific 600 people will be voting and nothing more.

Except the reality is much more different than that. Polls do influence people and their voting patterns. At least some New Brunswickers want to vote for the ‘winning’ party so as to be in someone’s good graces. If the Bristol poll shows a big swing either way, it could influence voters without having the usual rigor.

And given my mistrust of Al Hogan and Bristol, I almost think this is some kind of political ploy. Al likes to use the witches and card readers to predict the election so I wouldn’t put a little election trickery beyond his reach.

Either way, that’s bad form. But to be expected.

This entry was posted in Uncategorized. Bookmark the permalink.

0 Responses to On the weird Bristol poll

  1. scott says:

    I worked for a polling firm in Ottawa so what Poitras’ claims about Bristol is completely false. The sample which is used over again does not contain the same respondants as previously polled in the databank. What happens is they have a list of eligible voters per riding and their contact info. This is inputed into a what is known as a “demon dialer” which loops the numbers until a connetion is made. Once an individual is called and polled, they are taken off the list and not called back.

    What he is claiming is that the exact same 600+ are being phoned, which is completely inaccurate and not true at all. With the list of thousands of eligible voters that Bristol has at their disposal, calling the same ppl every night absolutely makes no sense, especially given the fact that they need to aim at polling 50% males and 50% females evry evening to have a semi-accurate sample.

    Sounds to me like Poitras is just trying to villify Bristol as he used a blog which is Liberal, not mention by Jacques btw, as an example to back up his claim. (see link on post) If you are going to report about Bristol and CRA in an unbiased fashion guys, first air your political leanings as some may think that you are objective and fair when you are really being considerably biased.

  2. David Campbell says:

    The truth is strange but still true. Here’s the headline from the T&T when the first Bristol poll came out:

    Unique approach to poll; More than 600 have agreed to be surveyed three times over course of N.B. campaign. New Brunswickers will be treated to a unique research undertaking leading up to the Sept. 18 election as the Times & Transcript and Omnifacts Bristol Research track the voting intentions…
    Page: A1
    Times & Transcript (Moncton) – Sat, Aug 26, 2006 – 287 words

    So, Scott, as you say this absolutely makes no sense and I don’t think very many people would argue with you.

    But they are doing it anyway.

    As for my ‘political leanings’ I will air them as I have a dozen times. My family is at least three generations of Tories on each side. My grandfather used to be one of those guys in the 30s and 40s that would be employed by the government when it was Tory and fired when it was Liberal. I have voted for Elsie Wayne and Stephen Harper.

    But in provincial politics, my ‘political leanings’ do lean Liberal these days and I make no bones about it.

    No matter how well intended, Premier Lord is taking us down a dangerous economic path and I am clear about that.

    As for Jacques, he’s right on the Bristol poll but whether he’s ‘right’ or ‘left’ is out of my sphere of knowledge.

  3. scott says:

    Mea Culpa. I stand corrected. My bad. As for political leanings, that was directed at the other site which was referenced by Poitras. He did not make it clear that the site he was referencing was Liberal.

  4. Anonymous says:

    Did we settle what was true about the poll? Is it the same 600 or not? If it is, isn’t Jacques right? So what does it matter who he references? I’ve been at six websites today, should I be saying which ones and their political leanings (or suspected ones) after each point?

    People tend to fixate on whether something is ‘liberal’ or ‘conservative’ as a way to say that its ‘biased’ without even addressing the argument itself (although the above comments did).

  5. Sara G says:

    An acquaintance of mine was a participant in the first poll and at the end of the survey was asked to participate through the length of the campaign. So it appears that they are using the exact same participants unless of course someone said they didn’t want to participate again. I’m not sure of the logic in this process because how much is going to change except for the undecided?

    There’s been a lot of reference to the T&T (and I agree with David’s comments on Al Hogan) but was this not a joint poll (a Canadaeast or Irving paper poll)? Or did the T&T and TJ commission two separate polls fromt he same company that had the same results and were published the same day?