An Ontario election polling firm responded to a statement from an industry counterpart on Thursday that some in the business were "hucksters selling methodological snake oil," after releasing vastly contradictory results.
In a letter posted online Wednesday, Ipsos CEO Darrell Bricker and senior vice-president John Wright called into question the untried methodology used by some marginal pollsters and their failing to police their own results.
The statement comes against the backdrop of the most recent Ontario provincial election polling results, which saw most results suggest the Liberals were gaining momentum and in a statistical tie with the once-leading Progressive Conservatives.
Another poll suggested the Liberals had jumped ahead, while Abacus's own results found the Conservatives still maintained a strong lead.
The online Abacus poll was the only one that did not suggest the Liberals were gaining momentum in the campaign. Instead of asking respondents who they were planning on voting for, Abacus asked how likely each person was to vote for each party.
In a statement on Thursday, Abacus said the election results would determine which company was more accurate, defending its less common line of questioning as innovative this early in a campaign.
"Pollsters have been using a sliding scale for years to gauge purchasing intent and brand loyalty," the post read. "Is it flawed? No. Just because a few competitors don't like it or haven't seen it before, doesn't make it ‘snake oil.'"
Nik Nanos, president and CEO of Nanos Research, said on Thursday that he shared some of the concern voiced by Wright and Bricker. He questioned Abacus' decision to introduce its new methodology during the Ontario election.
"You don't go into the Super Bowl and start trying things you have never done during the season. You go into the Super Bowl with the tried-and-true methodologies that got you there," Nanos said. "Although there is nothing wrong with trying out new ideas."
Nanos said when a news organization commissions a trustworthy company, it is able to ensure accuracy and understand the context and values of the results.
"When a media organization sponsors its own polls, there is always a process to discuss the content, the questions, the question or and the methodology," Nanos said. "They know exactly what they are getting; they know the quality of the work."
Nanos Research is commissioned by CTV News and The Globe and Mail to do their exclusive polling during the Ontario provincial election.
Ontario election polling results painted a confusing picture for observers earlier this month when both Ipsos and Nanos Research polling suggested Liberal support was on the incline and were statistically tied with the Conservatives.
A Harris/Decima poll meantime suggested the Liberals had suddenly secured an 11-point lead over the Conservatives, while the Abacus poll suggested the Conservatives held a nine-point lead.
"Polls are kind of complex organisms, many times they cannot be compared," Nanos said. "The reality is you should be looking at the trends that a pollster had in terms of the direction of the campaign."
Abacus also defended its decision to conduct an online poll, as opposed to a random telephone survey, saying the method has provided accurate polling in the past.
