Page: 81 [223] I note that call-backs to verify the information received would run counter to the concern expressed by the Supreme Court that efforts to establish causal effects must not breach ballot anonymity. While there were multiple call backs to obtain a response, EKOS did not thereafter attempt to verify the responses received from those reporting non-voting behaviour. Dr. Corbin views this as a weakness but it is a point that would be problematic in any live survey that could not assure anonymity. [224] Dr. Corbin challenges Mr. Graves’ assertions that the survey provides a reasonable basis for estimating the impact of the calls. She states that there is no scientifically defensible basis for estimating or inferring causal impacts from the report. Following a detailed analysis of the discussion of the results in the EKOS Survey, she found the explanations "subjective, predisposed to a hypothesis of voter suppression, and inconsistent with rigorous statistical reasoning for determining cause-and-effect." [225] Dr. Corbin’s analysis of the data led her to conclude that at least two ridings, SaskatoonRosetown-Biggar and Nipissing- Timiskaming, could immediately be eliminated from consideration as targets of suspicious phone calls, because there were fewer complaints about such calls than in the comparison ridings. Her analysis of the remaining ridings indicates that overall, there was not enough difference between voters in the comparison group and voters in the subject ridings to demonstrate that suspicious calls had made any difference. She points out that in Winnipeg South Centre and Vancouver Island North, the magic number is not reached even if Mr. Graves’ findings are accepted. 2013 FC 525 (CanLII) secrecy. One of the features of the IVR approach was that the respondents were assured

Select target paragraph3

Connect to a paragraph
Connect to an entity
Disable highlights
Add to table of contents