Forum is most accurate with
an average deviation of 0.7%
TORONTO, OCTOBER 20th, 2015 – Out of
all the polling firms tracking the 42nd Canadian general election,
Forum has proven to be the most accurate. Forum successfully predicted the
election outcome within 0.7% average deviation. Furthermore, Forum is the only
firm to predict a Liberal majority government giving the Liberal party 40% of
the vote, and 171 seats in the October 18th final report. When it
comes to sample size and margin of error, Forum predicted the outcome within a
margin of error of +/- 3%.
“Once again, Forum’s IVR polling
methodology has proven itself to be the most accurate means of polling public
opinion, and it allowed us to accurately predict the Liberal majority outcome,” said Forum Research president Dr. Lorne
Bozinoff.
Lorne Bozinoff, Ph.D. is the president and founder of Forum Research. He can be reached at lbozinoff@forumresearch.com or at (416)
960-9603.
|
%
|
Lib
|
Cons
|
NDP
|
Bloc
|
Green
|
Other
|
Average
Deviation
|
|
Results
(rounded)
|
40
|
32
|
20
|
5
|
3
|
0
|
|
|
Results
(first
decimal)
|
39.5
|
31.9
|
19.7
|
4.6
|
3.4
|
0.9
|
|
|
Forum Research
|
40
|
30
|
20
|
6
|
3
|
1
|
0.7
|
|
Nanos
|
38.2
|
30.1
|
21.2
|
4.9
|
4.7
|
0.8
|
1.1
|
|
CBC Poll Tracker
|
37.2
|
30.9
|
21.7
|
4.9
|
4.4
|
0.9
|
1.1
|
|
EKOS
|
35.8
|
31.9
|
20.4
|
4.9
|
5.6
|
1
|
1.2
|
|
IPSOS-Reid
|
38
|
31
|
22
|
4
|
4
|
1
|
1.3
|
|
Leger
|
38
|
30
|
22
|
6
|
4
|
0
|
1.3
|
|
Mainstreet Research
|
38
|
33
|
21
|
4
|
5
|
1
|
1.3
|
|
Angus Reid
|
35
|
31
|
22
|
5
|
5
|
2
|
2.0
|
|
Innovative Research
|
38
|
30
|
22
|
3
|
6
|
1
|
2.0
|
|
Abacus Data
|
32
|
33
|
24
|
4
|
5
|
2
|
3.0
|
|
Environics
|
29
|
26
|
34
|
4
|
8
|
0
|
6.2
|
Methodology
Forum only
used the final election prediction numbers published by each organization prior
to Election Day.
To calculate
the average deviation Forum Research compared survey results per party to
actual election results. The deviation per party was then averaged to determine
overall accuracy in predicting the election.
Most firms provided their predictions using
rounded numbers however a few predicted to the first decimal. In order to
fairly assess each organization’s election prediction we use comparable
election numbers. We have provided the election results in both rounded and
first decimal values.
|
%
|
Avg. Deviation
|
Sample Size
|
Date Published
|
Methodology
|
Source
|
|
Forum
|
0.7
|
1,373
|
18-Oct-15
|
IVR
|
Source
|
|
Nanos - 3 day
|
1.1
|
2,176
|
18-Oct-15
|
Telephone
|
Source
|
|
CBC Poll Tracker
|
1.1
|
N/A
|
18-Oct-15
|
Data
Aggregator
|
Source
|
|
EKOS
|
1.2
|
2,122
|
18-Oct-15
|
IVR
/ Telephone
|
Source
|
|
IPSOS-Reid
|
1.3
|
2,503
|
17-Oct-15
|
Online
/ Telephone
|
Source
|
|
Leger
|
1.3
|
2,086
|
16-Oct-15
|
Online
|
Source
|
|
Mainstreet Research
|
1.3
|
5,546
|
15-Oct-15
|
IVR
|
Source
|
|
Angus Reid
|
2.0
|
2,022
|
16-Oct-15
|
Online
|
Source
|
|
Innovative Research
|
2.0
|
1,143
|
11-Oct-15
|
Telephone
|
Source
|
|
Abacus Data
|
3.0
|
2,000
|
6-Oct-15
|
Online
|
Source
|
|
Environics
|
6.2
|
1,001
|
15-Sep-15
|
Telephone
|
Source
|
