Vote Compass: Canadians predict who they think will win their riding

With the federal election mere days away,parties are campaigning furiously to galvanize their bases and try to sway undecided voters.

Candidates have their own idea of how well their respective campaigns aregoing and who is going to win these local contests.But what about the people living in those ridings?

CBC/Radio-Canada has producedan interactive feature that tries togauge local sentiment about which party voters think will win their respective ridings on Oct. 19.

The interactive belowis not a poll, and it does not display voter intentions. Rather, it uses data gathered through CBC's Vote Compass tostate which party people across the country believe is most likely to win the riding they live in.

The results are based on 68,481 responses gathered through Vote Compass between Oct. 9 and Oct. 12.

• Find your riding
• Poll Tracker: Where do the parties stand?
• Pledge to Vote: Take the challenge

The map below contains all of the ridings in Canada. The colours correspond to the party that most respondents believe willreceivethe most votes.

In addition to making a prediction, Vote Compass respondents were asked to express how confident they were in that prediction. That dataisreflectedby clicking on a ridingon the map.

Click on the map or use the zoomtool to find your riding.

Developed by a team of social and statistical scientists from Vox Pop Labs, Vote Compass is a civic engagement application offered in Canada exclusively by CBC News. The findings above are based on 68,481 responses gathered through Vote Compass between Oct. 9 and Oct. 12.

The results are derived from the responses to the question: In your opinion, how likely is it that the people in your riding will elect (party)? Users are asked this question for each party (with the caveat that only users in Quebec receive the Bloc Qubcois) and provided with a response scale from 0 to 10, where 0 means not at all likely and 10 means very likely.

The results are not an indication of who the constituents of a riding actually intend to actually vote for, but who they think will win the election in their riding. They are not a forecast of the election results.

Unlike opinion polls,respondents to Vote Compass are not randomly selected. Similar to opinion polls,however, the data are a non-random sample from the population and havebeen weighted in order to approximate a representativesample. Vote Compassdata have been weighted by geography, gender, age, educational attainment,occupation, religion, religiosity and civic engagement to ensure the sample'scomposition reflects that of the actual population of Canada according to censusdata and other population estimates.