adma
Superstar
Well, unlike Farrow, he didn't use white type for his surname, so it's readable.
|
|
|
2014 Municipal Election: Toronto Council Races
- Thread starter ShonTron
- Start date
Well, unlike Farrow, he didn't use white type for his surname, so it's readable.
nfitz
Superstar
I'm surprised no one mentioned the Toronto Star/Forum ward by ward poll results (for selected wards) that came out last night:
http://www.thestar.com/news/city_ha...s_for_toronto_city_council_election_2014.html and
http://poll.forumresearch.com/post/178/in-toronto-ward-races--most-incumbents-rule
The samples sizes are small, but it's hard to see how Rob Ford won't get elected in Ward 2 with over 40% of the vote. Domise and Abukar tied near 15% each. Digging deeper, the strong Doug Ford support in both Ward 1 and 2 suggests Mike Ford might easily win as well. Sigh ...
No other huge shocks - though Ward 12 is a complete toss-up. Yikes, Nunziata might get in.
http://www.thestar.com/news/city_ha...s_for_toronto_city_council_election_2014.html and
http://poll.forumresearch.com/post/178/in-toronto-ward-races--most-incumbents-rule
The samples sizes are small, but it's hard to see how Rob Ford won't get elected in Ward 2 with over 40% of the vote. Domise and Abukar tied near 15% each. Digging deeper, the strong Doug Ford support in both Ward 1 and 2 suggests Mike Ford might easily win as well. Sigh ...
No other huge shocks - though Ward 12 is a complete toss-up. Yikes, Nunziata might get in.
toraerach
Active Member
I'm not sure how much faith I'd put in this. That they have Bravo coming in third in Ward 17 sounds suspicious.
WislaHD
Superstar
Yah the forum research polling has been getting a looooot of hate online.
JGHali
Active Member
Well, these are all very small samples. For a sample of 216 as in Ward 2, the sampling error is *at least* about +/-7% and thanks to non-response bias, incomplete sampling frames, and an inability to stratify the sample, the actual error/bias is certainly a lot higher.
salsa
Senior Member
What is it with suburbanites and their shitty councillors that they vote for?
ShonTron
Moderator
Member Bio
- Joined
- Apr 24, 2007
- Messages
- 12,487
- Reaction score
- 9,328
- Location
- Ward 13 - Toronto Centre
The sample sizes are tiny. Some results are really suspect. How come the Tory/Stintz backed candidate in Ward 16 not even register in the poll? (I'm rooting for JP Boutros personally.)
nfitz
Superstar
Sure - but Rob Ford has a huge lead in Ward 2. It's hard to believe as bad as the poll could be - that it's that bad.
There can be no doubt then, that Rob Ford will be re-elected to council.
Greg Banks
Active Member
Really? With a an electors list of 33,073 in Ward 2, a survey of 216 would seem to offer a more accurate result than say a survey of 3569 of 1,560,527 (Mainstreet's poll of Toronto with an accuracy of 1.64% 19 of 20 times)
christiesplits
Senior Member
A few of the polls seem questionable. Bravo in third place in Ward 17? That is highly unlikely. I also expect the race between Augimeri and Cusimano is closer.
Ward 12 looks insane. Even before the poll I suspected it was a four way race between DiGeorgio, Nunziata, Dominelli and Olawoye. I have my fingers crossed that Lekan Olawoye can end up on top.
Ward 12 looks insane. Even before the poll I suspected it was a four way race between DiGeorgio, Nunziata, Dominelli and Olawoye. I have my fingers crossed that Lekan Olawoye can end up on top.
JGHali
Active Member
What? The sampling error has little to do with the size of the population. As a general rule of thumb, sampling error is proportional to 1/sqrt(n), with n denoting sample size. This error doesn't, however, account from the bias introduced by non-random non-response.
One problem with these polls is that I'm not sure how they determine where someone with a mobile number lives. Names attached to cell phones don't appear in the phone book, and the exchange of each number doesn't reflect the neighbourhood where they live.
adma
Superstar
Or, re *mayoral* choices, that it'd be 3-way in Ward 14 (Perks' ward)...
Greg Banks
Active Member
Then, aren't you in fact blowing smoke with + or - 7%, if you don't have a clue on the sampling?
Long Island Mike
Senior Member
Everyone: How is Andray Domise (See: www.andraydomise.ca) doing in his effort to win the Ward 2 City Council
seat against RF? This will probably be the most-watched Council race in Toronto tomorrow and I hope that AD wins
to send RF a message - that his health and well-being is more important then politics...LI MIKE
seat against RF? This will probably be the most-watched Council race in Toronto tomorrow and I hope that AD wins
to send RF a message - that his health and well-being is more important then politics...LI MIKE
JGHali
Active Member
Well, that margin of error only truly applies if it's a true, unbiased random sample, which is why +/- 7% is an underestimate of the actual error. Assuming a random sample of size 216, though, we get an approximate standard error of 0.068, which I rounded to 7%. Similarly, for that Mainstreet poll, a random sample of 3569 gives an approximate standard error of 0.0167, close to the error margin they calculated more precisely.