Latest polls on Sask. Party leadership race marred by small sample, narrow definition

Two new third-party polls on the Saskatchewan Party leadership race feature questionable polling methods and therefore offer no clear picture of a front-runner less than a week before Saturday's leadership contest, says CBC poll analyst Éric Grenier.  

The methodology issues, including small samplings of party members and a potentially narrow definition of what constitutes a Saskatchewan Party member, raise questions about the reliability of the polls released Monday by Insightrix and Mainstreet, he said. 

"I think people can look at them to give them a little bit of what the political landscape looks like," said Grenier. "I wouldn't suggest that members of the party should make their voting decision based on these results."

Insightrix poll

The online Insightrix poll, done in partnership with CJME and CKOM, surveyed 1,004 randomly selected Saskatchewan residents between Jan. 9 and 17.

Only 104 of those polled were Saskatchewan Party members. As a result, "You're looking at a very, very large margin of error," said Grenier — +/- 9.6 percentage points, 19 times out of 20.

Among all 1,004 voters polled — including 900 non-party-members — Ken Cheveldayoff was seen as the frontrunner by the largest contingent of respondents, 16 per cent.

Among only the 104 party members quizzed, Alanna Koch led the race.

Grenier said it's easier to do general population polls because statistics are kept on demographics. It's harder to poll members of political parties because that information is something a party keeps to itself.

Another problem with the Insightrix poll, he said, is the ambiguity about what makes a person a member of a political party. 

"Let's say they've donated to the party or they've voted for them all their lives — they might consider themselves to be a member of the party even if they don't have a membership card and won't be able to vote," Grenier said.

Mainstreet poll

The Mainstreet poll, which combined three surveys, polled 764 Saskatchewan residents between Jan. 4 and 6 using "interactive voice recording" and had a margin of error of +/- 3.54 per cent, 19 times out of 20.

Among those polled were 452 former donors to the federal Conservative Party thought to represent Saskatchewan Party members — another red flag, said Grenier. 

"It might be people who are more conservative than the average member of the Saskatchewan Party," Grenier said. At the same time, people who vote for the Saskatchewan Party at the provincial level but for the Liberals at the federal level would not be represented either.

"The fact that it is only looking at people who have donated to the federal Conservatives means that it might not be looking at a completely representative sample of the Saskatchewan Party."