18 January 2006

Outlier polls

While polls have their value, there has been entirely too much emphasis on them in media reports during this campaign, particularly at the national level.

That said, it is interesting to see this simple comparison presented in today's nightly tracking report from SES:

Comparison

* SES Research (N=1,012 decided voters, January 14 to 16, 2006)

* Ekos Research (N=2,018 decided voters, January 15 to 17, 2006)

* Decima Research (N=1, 017 decided voters, January 12 to 15, 2006)

* Strategic Counsel (N=1,500 voters, January 14 to 16, 2005)

CPC -- SES 37%, Ekos 37%, Decima 37%, Strategic Counsel 42%

LPC -- SES 30%, Ekos 27%, Decima 27%, Strategic Counsel 24%

NDP -- SES 18%, Ekos 20%, Decima 18%, Strategic Counsel 17%

BQ -- SES 10%, Ekos 11%, Decima 11%, Strategic Counsel 12%

As I have contended all along, the Strategic Counsel research seems to be flattening Liberal numbers by some amount outside the margin of error for the polls and, likely inflates the Conservative numbers by an amount outside the usual margin of error.