A marketing research expert retained by the party’s law firm has given a detailed analysis questioning the methodology and conclusions of a poll by Ekos Research that found “robocalls” and live calls suppressed the non-Conservative vote by about 1.5 per cent in these ridings.
The Ekos survey, conducted by pollster Frank Graves, was prepared at the Council’s request to back up a series of legal challenges in Federal Court that seek to have the election results set aside in select ridings won by Conservatives.
Most of the ridings were won by small margins. The applications are brought by people in each of the ridings who all said they received calls.
The Ekos poll is a key piece of evidence in the cases. To trigger byelections, the applicants must convince a judge that not only was there interference in the vote, but also that it affected the outcome.
Graves said his poll of 3,297 people showed a “highly statistically significant” trend of non-Conservatives receiving the calls.
But Graves’ work suffers from “profound problems,” according to an affidavit filed in each of the cases by Ruth Corbin, the CEO of market research firm CorbinPartners.
“Nothing can be concluded from the survey with respect to the incidence of voter-suppression phone calls or with respect to any cause-effect relationship between such phone calls and outcomes of the 2011 Federal election.”
Corbin cites numerous sampling problems with the survey, including the exclusion of people using cellphones, and says the work contains “serious violations of published industry standard.”
She questions the automated interactive-voice response (IVR) system Ekos used to conduct the poll, calling it “a kind of ‘robo-call’ technique of its own.”
She says IVR has very low response rates and “permits no verifiable control over who answers the survey — it could be a 12 year old child without permission to vote, for example.”
Graves conducted the poll in April, after Chief Electoral Officer Marc Mayand told a parliamentary committee that Elections Canada had fielded more than 700 complaints about election calls from about 200 ridings across the country.
The poll asked respondents if they had received misleading calls during last May’s election.
In her 35-page affidavit, Corbin says Graves’ results cannot be relied on because the survey was conducted so long after the fact.
“It is implausible that consumers would remember not only getting a call a year after it occurred, but specific details of the call.”
Graves had said in his affidavit that he dealt with the possibility of “over-remembering” by comparing his results in the seven ridings with a control group of 1,500 people in other ridings.
Corbin suggests Graves didn’t make his raw data available for others to review. However, the poll results have been posted on ekospolitics.com since shortly after his affidavit was filed.
The lawyer for the applicants in the cases says he will be presenting evidence next week to rebut Corbin’s affidavit.
“Mr. Graves takes the allegations very seriously and will be replying to them in detail next week,’ said Steven Shrybman.
Original Article
Source: ottawa citizen
Author: GLEN MCGREGOR AND STEPHEN MAHER
No comments:
Post a Comment