[N_Fiddledog]

Critiques, you mean, Doc.

It's a bogus, push-poll from a non-profit with an agenda, and CBC is offering up some more Liberal party propaganda.

[DrCaleb]

N_Fiddledog N_Fiddledog:

Critiques, you mean, Doc.

No, those are direct attacks on both the credibility and integrity of those involved. Ad Hominem. 'Critiques' involve picking apart the results or methods, not the organizations.

N_Fiddledog N_Fiddledog:

It's a bogus, push-poll from a non-profit with an agenda, and CBC is offering up some more Liberal party propaganda.

It's a paid poll from a group (Environics) whose job it is to do such things, from a news service whose job it is to ask such questions. Just because the answers don't fit your agenda doesn't mean it can be dismissed as 'party propaganda'.

[DrCaleb]

Survey of Muslim Canadians rebuts lazy generalizations with hard data

[andyt]

martin14 martin14:

1 million muslims in Canada, and they post results from 600.


And we are supposed to take these results seriously ?

Every time you righties see a poll you don't like the results of, you bring out this trope. Have you ever taken any statistics courses?

A better critique would be to ask how honestly the radical Muslims answered the question.

But pretty interesting result that more Muslims are proud to be Canadians than non-Muslims are.

[Lemmy]

That's the "Righty Rulebook" phenomenon again. Any information/research that doesn't jive with the rulebook is immediately dismissed. When they can't discredit the message, they resort to discrediting the messenger. But really they just discredit themselves when they do so.

[andyt]

So you should know. The 1200 sample size for all of Canada (30,000,000) seems pretty skimpy. What is an accepted confidence level and margin of error for these things usually?

Goes to show how much statistics I've forgotten. I found a sample size calculator. For any population of 500,000 and above, a sample of 600 will yield the same margin of error and confidence level.

But again, that leaves the design of the survey. Did they even manage to get radical Muslims to respond? Did they use interpreters or only survey those who spoke good English? How did they frame the questions?

[2Cdo]

Lemmy Lemmy:

That's the "Righty Rulebook" phenomenon again. Any information/research that doesn't jive with the rulebook is immediately dismissed. When they can't discredit the message, they resort to discrediting the messenger. But really they just discredit themselves when they do so.

Do you honesty think this is just a "righty" phenomenon?

[andyt]

Never seen a lefty on this forum use that bullshit excuse. It's a FOG special, and Martin seems to fall for it too. Guess lefties are just smarter than that. I've pointed out lots of questions you could ask about the survey, but just yammering about sample size, when it's more than adequate, won't cut it.

[DrCaleb]

andyt andyt:

So you should know. The 1200 sample size for all of Canada (30,000,000) seems pretty skimpy. What is an accepted confidence level and margin of error for these things usually?

I'd think it's more about cost. For 1 Million Muslims, a sample population of 600 means 4% margin of error. To get 1% margin of error, you'd have to poll 9,000 people. That costs money.

The same would be true of the Canadian Population. 1200 from 34 million would be about the same error rates, so you'd have to poll 12,000 to get a 1% margin. If CBC did that, everyone would bitch how much money they spent on a poll.

So 4% error rate is plenty good enough.

[martin14]

andyt andyt:

Never seen a lefty on this forum use that bullshit excuse.

Bullshit.
It gets constantly pulled out in every global warming thread.

[andyt]

That the sample size of a survey was too small? Really? Got a link for that?

[Lemmy]

2Cdo 2Cdo:

Do you honesty think this is just a "righty" phenomenon?

Absolutely. There's been lots of research and writing done on this phenomenon. It's one of the defining characteristics that makes someone a conservative.

"Why Liberals and Conservatives Clash" - Bruce Fleming, ISBN: 13-978-0-415-95353-5

"Moral Politics: How Liberals and Conservatives Think" - George Lakoff, ISBN: 0-226-46771-6

"Predisposed: Liberals, Conservatives and the Biology of Political Differences" - John Hibbing & Kevin Smith, ISBN: 1136281215

"The Great Divide: Why Liberals and Conservatives Will Never, Ever Agree" - William Gairdner, ISBN: 1594037655

"The Righteous Mind: Why Good People are Divided by Politics and Religion" - Jonathan Haidt, ISBN: 978-0-307-4557-2

The overwhelming conclusion is that the fundamental difference between conservatives and liberals is the conservatives' need for a rulebook. And the authors cited above come from across the political spectrum. Start with Gairdner. He's as far right as they come.

[andyt]

Well actually the conservatives in this case are going against the rule book. Statistics is a well established branch of Mathematics. (what you do with them is another matter). But sample size calculation is a basic part of Statistics, no controversy about it, nobody demanding that creationism (as in just create your own sample size "facts") be taught beside statistics.

Seems the righties only like the rules when they favor their biases. In that of course they are no different than the lefties. But not in this case.

[Lemmy]

No, the "rule" in this case is "The CBC is biased and untrustworthy". That rule has them immediately dismiss anything the CBC reports. It's not a dismissal of statistics, it's a knee-jerk reaction to attempt to discredit what their rulebook tells them to discredit.

[andyt]

FD went that route, but Martin did a straight out FOG. Only 600 out of 1 million surveyed, can't possibly be valid.

the problem, as we've seen with political polls, isn't the sample size, but how they go about collecting the sample. Is it truly random. Is it a push pull poll as FD says, and so on.

I don't think the rules the authors you mention are talking about are just biases. They're the official rules, ie conservatives are more authoritarian, and the well established common rules a society has.