Polls-- Neutral? Or another form of endorsement?

[Boru]

Strategic

17/01/2006

Liberal 24

CPC 42

NDP 17

Bloc 12

Green 5

SES CPAC

17/01/2006

Liberal 32

CPC 37

NDP 18

Bloc 10

Green 4

I find it hard ot believe that two polls could yield two results so differently. Do you think it's possible that even polls are being run in ways to favour one aprty over another? I'm not saying any one is more valid than the other at this point, but as you can see, one poll displays an 18 point lead, whiel the other a 5 point lead. I don't think any margin of error can quite cover that spread.

I have no idea, but just thought it interesting. What do you think? Is this one conspiracy theory too many?

[shoop]

Yes, it is one conspiracy theory too many.

The only real *different* results are the Liberals and the Conservatives.

There are real questions with the SES methodology.

That being said I feel the true number lies somewhere between the Strategic Counsel and SES numbers at this point...

I find it hard ot believe that two polls could yield two results so differently. Do you think it's possible that even polls are being run in ways to favour one aprty over another? I'm not saying any one is more valid than the other at this point, but as you can see, one poll displays an 18 point lead, whiel the other a 5 point lead. I don't think any margin of error can quite cover that spread.

I have no idea, but just thought it interesting. What do you think? Is this one conspiracy theory too many?

[geoffrey]

What are the margins and poll numbers for these?

I wouldn't say people are outright lying about outcomes. And its hard to influence a poll question that dramatically, at least in regard to "If the election was held today, which party would you support?" (The question I was asked by Strategic Counsel about a week ago).

Though taking the polls at different times of days (days will include more stay at home moms/dads and unemployed, ect..) may have an effect? Timezones would play into this.

No poll has numbers like that 32 to the Liberals though (other than SES).

[seabee]

Cell phones are becoming a problem for pollsters; they are not available in directories. So people who are now commnicating only with cellphones are not contacted by pollsters; and these people represent one particular group. That alone changes the reliability of polls. Even with people who have both landlines and cellphones, this affects reliability of polls; if the pollsters phone daytime, there may be no-one to answer to the phone, as these people, during working hours, rely on cellphones. This may have more of an effect in urban areas than in rural ones, and in the cities, more on high-income people than low-income people.

Polls may be interesting entertainment, with some value, but are taken with more than a bit of salt.

[scribblet]

Strategic

I find it hard ot believe that two polls could yield two results so differently. Do you think it's possible that even polls are being run in ways to favour one aprty over another? I'm not saying any one is more valid than the other at this point, but as you can see, one poll displays an 18 point lead, whiel the other a 5 point lead. I don't think any margin of error can quite cover that spread.

I have no idea, but just thought it interesting. What do you think? Is this one conspiracy theory too many?

It depends on the questions asked, I do believe that push polls can sway voters. Isn't that the reason for polls not being allowed 24 hours prior to election day? I'd like to see it more than that, maybe 3 days before.