Page: 82
[226]
Dr. Corbin would eliminate more than half of the survey respondents as lying or
mistaken based on a comparison of tables of survey respondents who reported receiving calls
voted. Excluding those she concluded had lied from the data and adjusting the data to
compensate for under-age voters, cell-phone owners, ineligible voters and other concerns such as
faulty memory, the remaining alleged voter suppression effect is reduced to zero. In the result,
Dr. Corbin found no statistical difference between the subject ridings and the comparison group
ridings. In her analysis, the EKOS Survey is, overall, methodologically unsound and could have
a margin of error of between 8% and 20%.
[227] A sur-reply affidavit by Dr. Corbin was tendered on October 29, 2012 in response to the
revised report by Graves and EKOS calculated without the data from the seventh riding. In the
revised report, some of the numbers had changed but not to the extent that the results were
dramatically altered. Changes from the original report were indicated with strike-throughs and
side-bars describing the adjustments made.
[228] Dr. Corbin added the following arguments, among others:
There is now stronger evidence of no material voter suppression in the collection
of six subject ridings.
There is new unequivocal evidence that if non-voters had been targeted by
illegitimate phone calls, those most likely to be targeted were people with
Conservative Party leanings, rather than non-Conservative Party leanings.
That
2013 FC 525 (CanLII)
which caused them not to vote, and tables of survey respondents who said they had actually
Select target paragraph3
Connect to a paragraph
Connect to an entity
Disable highlights
Add to table of contents