Compared to Forum's previous poll of September 5, the Liberals dropped two points, while the NDP was up one and the Conservatives were unchanged. None of these shifts were outside the margin of error.
It is worth noting, however, that the Liberals have dropped in three consecutive polls from Forum, down from a high of 44% in July. That number was unusually high, however, and other polls have not hinted at a similar Liberal slide, or relative Conservative strength for that matter. But this is the first poll done in the context of the Iraq mission and the Liberal and NDP decision not to support it, so we should wait and see what other polls will say in the coming days.
With the aggregate levels of support, the Liberals would likely win 145 seats, with 121 seats going to the Conservatives, 68 seats to the New Democrats, 2 seats to the Greens, and 2 seats to the Bloc Québécois.
The aggregate had previously been updated on October 2 with two new polls by EKOS Research for iPolitics and Abacus Data.
The EKOS poll, the fresher of the two, showed very little change since the last EKOS poll of July 16-23. The Liberals led with 38.3%, down 0.4 points, with the Conservatives down 0.7 points to 24.9% and the NDP up one point to 24.4%. None of these shifts were outside the margin of error.
As far as I can tell, that 24.9% was the lowest support the Conservatives had registered in any poll in over nine years. That might set alarm bells ringing - but recall that EKOS had the Tories at 25.6% in July. So, from that perspective, this was just an insignificant dip.
The other poll, by Abacus Data, put the Liberals in front with 38%, up three points from Abacus's last poll of August 15-18. The Conservatives were up one point to 30%, while the NDP was also up one point to 23%. None of these shifts were outside the margin of error of a similarly sized probabilistic sample, but the trends had been in the Liberals' favour over the last few Abacus surveys.
The methodology used to weigh polls can be found here. By including polls in the average, no representation as to the accuracy or equivalency of the methods used is implied, nor should inclusion be seen as an acceptance, endorsement, or legitimization of their results. The weighting scheme takes reliability partly into account. See here for a complete rundown of the latest polls in Canada (external link).