It appears the federal horse race may have reverted to its three-headedness again, as two new polls suggest a narrowing of the gap separating the three parties.

But one of the surveys provides some insight into what and who is capturing voters' attentions, and what effect it might be having.

The CBC Poll Tracker still has the NDP in the lead with 33.5 per cent support, followed by the Conservatives at 29.1 per cent and the Liberals at 27.3 per cent. The Greens are averaging 5.5 per cent support, while the Bloc Québécois stands at 15.3 per cent support in Quebec.

This is a bit of a reversion to where things stood before the publication of two polls that suggested strong numbers for the New Democrats. These surveys by Forum Research and the Angus Reid Institute, both out of the field a week ago, put the NDP at 40 and 37 per cent support, respectively, among eligible voters. It boosted the party in the average, but the four polls that have been published since have put the NDP between 31 and 34 per cent support. That is where the party was polling prior to the release of these two bullish surveys.

Mulcair not registering as much as Harper, Trudeau

The two newest polls both show a very close race. The survey by the Innovative Research Group put the NDP at 32 per cent, the Conservatives at 30 per cent and the Liberals at 27 per cent. That gap of five points between the three parties was down by three points from Innovative's previous pre-campaign poll, as the NDP dipped and the Liberals gained (though both marginally so).

The poll by Abacus Data published Monday had the gap even closer, at 31 per cent for the NDP, 30 per cent for the Conservatives and 28 per cent for the Liberals. Here again, with the NDP down and the Liberals up, the margin between first and third dropped by six points from Abacus's prior survey.

The Innovative poll looked into a number of different issues beyond the horse race. Of particular interest was the questions concerning whether Canadians had read, heard, or seen anything about the leaders in the few days before the poll was taken. By this measure, Stephen Harper was grabbing the most attention: 60 per cent said they had read, heard, or seen something about him, compared with 52 per cent for Justin Trudeau and 41 per cent for Tom Mulcair. 

But not all publicity is good publicity. The main story that got Harper's name into the news was the Duffy trial, with more than one-third of people who had heard something about the Conservative leader hearing it in that context. And what they did hear was damaging: only four per cent said it gave them a more favourable opinion of Harper and the Conservatives, while 65 per cent said it gave them a less favourable view.

Worse, among unaligned voters (those who say they do not identify with any of the main parties) the days preceding the poll had a negative effect on 54 per cent, with just six per cent saying it had had a positive effect.

Three issues dominated what Canadians had read, seen, or heard about Trudeau: his position on veteran pensions, his plan to increase taxes on the wealthy to give tax cuts to the middle class, and advertisements about him (paid for by both the Conservatives and Liberals). 

Overall, it was a good period for him: 39 per cent of Canadians said what they had read, seen, or heard about Trudeau gave them a more favourable opinion of him, while 23 per cent left with a less favourable opinion.

For Mulcair, his child-care plan captured the most attention. But the effect was a wash: 29 per cent said it left them with a more favourable view of Mulcair and the NDP, and 29 per cent said it left them with a less favourable view. Similarly, Mulcair left 31 per cent of unaligned voters with a more positive view, and 32 per cent with a less favourable one.

These numbers do seem to back up the overall trends that the voting intentions polls are showing: marginally positive for the Liberals, modestly negative for the Conservatives, and stable for the NDP. The campaign has yet to shake things up in any significant way.

CBC's Poll Tracker aggregates all publicly released polls, weighing them by sample size, date and the polling firm's accuracy record. Upper and lower ranges are based on how polls have performed in other recent elections. The seat projection model makes individual projections for all ridings in the country, based on regional shifts in support since the 2011 election and taking into account other factors such as incumbency. The projections are subject to the margins of error of the opinion polls included in the model, as well as the unpredictable nature of politics at the riding level. The polls included in the model vary in size, date and method, and have not been individually verified by the CBC. You can read the full methodology here.

The poll by Innovative Research Group was conducted between August 24 and 26, interviewing 3,274 Canadians via the internet. As the poll was conducted online, a margin of error does not apply. Full question wording and results can be found here.

The questions asked in the polls mentioned in this article were as follows:

Abacus: "If a federal election was held tomorrow, which one of the following parties would you vote for in your constituency?"

Innovative: "If a federal election were held today, which party would you vote for?"