Page: 85
[234] He described her general approach, however, as “cross[ing] the boundary that separates
The Corbin report lists a litany of claims concerning the reliability,
validity, and scientific reporting which render the data as invalid,
unreliable and so not a foundation from which useful inferences
can be drawn. My view is that this is too harsh a judgment. Many
of the claims upon which those sources of unreliability and
invalidity are based are factually incorrect. Others, as it turns out,
have been addressed.
[235] Dr. Nevitte concluded that the EKOS Survey “for the most part, makes relatively modest
claims that are advanced, usually, with related caveats and proper caution”, and criticized the
initial Corbin report on the ground that it had not squarely addressed the central question of
whether voters for different parties had different probabilities of receiving deceptive calls.
[236] Dr. Nevitte acknowledged that some of Dr. Corbin’s criticisms raised valid issues,
notably: (a) memory lapse over time; (b) the suitability of IVR technology; (c) the randomness of
the survey sample; (d) the nature and suitability of the comparison group; (e) the question of
causation; and (f) the issue of data disclosure.
[237] Memory unreliability could be a problem but a call related to a general election would be
more memorable, in Dr. Nevitte’s view, than most daily events. He notes that it would not be
logical that supporters of one political party would have consistently less accurate memories than
others. IVR interviewing is automated and impersonal but he did not see that as a problem and
considered that the methodology used was standard for the industry. The fact that the sample was
random was not an issue so long as the normal precautions were taken. While the comparison
group consisted of 106 ridings in which few complaints were filed, rather than ridings in which
2013 FC 525 (CanLII)
honest disagreement from strenuous, and not so credible, disagreement.” In his view:
Select target paragraph3
Connect to a paragraph
Connect to an entity
Disable highlights
Add to table of contents