According to Mainstreet, Ford won the debate (35.1%, vs Horwath's 24.3% and Wynne's 19.3%).
https://www.mainstreetresearch.ca/ford-wins-the-debate-according-to-torontonians/
But what's *really* interesting, though given no play at all, is that they threw a voting-intention question into it, and the PCs are at 36.6% to the Libs' 31.1% and the NDP's 23.1%. Remember: the last Mainstreet poll went 44.9-28.2-21.3; and before that 50-24-18. Maybe a different *kind* of sample, but...
Mainstreet has been discussed before.
For the hell of it, I went and looked at their stated methodology and details.
Some notes.
Poll was IVR, it was of City of Toronto residents only, with a solid sample size of 2,000
There were 4 geographical subsets, roughly, the old City of Toronto (downtown in Mainstreet parlance) along with Scarborough, Etobicoke and North York.
The subset data indicates the NDP in a very strong position in the old City, while the Conservatives have the inner burbs in a run-away.
Overall margin of error 2.9%, but varied in the 4-5% range for the geo sub-sets.
IF one took that data at face value, it wouldn't skew things much as they show the Cons up by 12 or so in Scarb/NY and even more in Etobicoke.
They show undecideds at 8%, the top line numbers are leaning and decided and so exclude that group.
I have real issues w/Mainstreet as a pollster, that said, if they asked the questions the way they said, its not a bad sample size.
But, I have some real questions of 'weighting'. The published results are weighted for gender, age and region.
For instance, it appears as though they used the census numbers for weighting.
Though Elections Canada suggests that both age and gender show variable turnout rates.
Methodology language from Mainstreet:
The analysis in this report is based on results of a survey conducted on May 7th at the
conclusion of the rst provincial leaders’ debate among a sample of 2000 adults, 18 years
of age or older, living in Toronto. The survey was conducted using automated telephone interviews (Smart IVR). Respondents were interviews on landlines and cellular phones.
The survey is intended to represent the voting population of Toronto.The survey was conducted by Mainstreet Research and was not sponsored by a third party.The sampling frame was derived from both a national telephone directory compiled by
Mainstreet Research from various sources. The survey was conducted as a stratied dial of
the following regions: Downtown Toronto (collloquially known as the old City of Toronto), Etobicoke, North York, and Scarborough. Respondents were dialed at random. The questionnaire used in this survey is available in this report and online at www.mainstreetresearch.ca. Questions are asked as they appear in the release document. If a question is asked of a subset of the sample a descriptive note is added in parenthesis preceding the question.The sample was weighted by population parameters from the Canada 2016 Census for adults 18 years of age or older in Canada. The population parameters used for weighting are age, gender, and region.
The margin of error for this poll is +/- 2.19% at the 95% condence level. Margins of error are
higher in each subsample. The margins of error for each subsample is as following: Males: +/- 3.37%, Females: +/- 2.88%, 18-34 age group: +/- 7.49%, 35-49 age group: +/- 5.92%, 50-64 age group: +/- 3.94%, 65+ age group: +/- 3.2%, Downtown Toronto: +/- 4.21%, Etobicoke: +/- 5.32%, North York: +/- 4.31%, Scarborough: +/- 3.99%.In addition to sampling error, one should bear in mind that the wording of questions and
practical difficulties in conducting surveys can introduce error or bias into the ndings of
opinion polls
Elections Canada Turnout rates:
http://www.elections.ca/content.aspx?section=res&dir=rec/eval/pes2015/vtsa&document=table2&lang=e