by David J. Climenhaga
Wow! This just in! Alison Redford is so popular….
How popular is she?
Alberta’s new premier is so popular she’s more popular that her popular party. And they’re so popular they’re called the Popular Conservatives, right? So that means, surely, that “the Conservatives have basically recovered from the Stelmach era.”
That quote belongs to a political scientist, so I guess you can take it to the bank. It’s one of several similar comments used by the Calgary Herald to bolster its story about its counter-intuitive new poll, which says that 44 per cent of decided voters and those leaning a particular way would vote for Ms. Redford’s Conservatives if an election were held right now
What’s more, according to the exclusive poll, done for the Herald by Angus Reid Public Opinion, Ms. Redford herself has the approval of 55 per cent of Alberta’s voters.
So it’s all over but the weeping and the gnashing of teeth, right? Woe unto the Wildrose Party, saith the Herald, since it’s now sitting at a non-heady 22 per cent, the Alberta Liberals at 16 per cent, the NDP at 13 per cent and the Alberta Party at 2 per cent.
Indeed, things look so bad for the sad-sack Wildrosers, said the Herald, that the right-wing party has come a long way down from “the zeitgeist of 2009 and 2010, when the party seemed to be surging ahead of the long-governing Tories.” The zeitgeist of 2009? I guess it’s a good thing that the days are gone when they horsewhip editors for letting stuff like this into their newspapers. But, still…
Well, as they say, just a you-know-what minute.
First of all, of course, these results are based on a poll conducted from Oct. 17 to 19, before Premier Redford’s run of flip-flops and flip-flop-flips began to make the news.
Be that as it may, back in September, in the midst of the Ontario election campaign, Darrell Bricker and John Wright of the Ipsos Reid polling company had a little sage advice for journalists during election campaigns. To wit: “All polls are not created equally.”
There is, the pair of pollsters said, “a disturbing trend of late in which questionable polls find their way into an outlets coverage because they appear to match an editorial line, or present a counter-intuitive perspective.”
Messrs. Bricker and Wright called for journalists to be more grown-up and responsible about the way they report polls, and they provided readers with a handy list of six easy rules to help us, the “consumers” of the product produced by newspapers, judge their likely accuracy.
Among those rules, they suggested pollsters who use on-line methodologies to predict votes should be asked to provide “unweighted” data, results prior to adjustment for demographics and political support. “You will find some heavy thumbs are being applied to adjust for under-represented voting groups. While the weighting can produce very good results, it really amounts to no more than an educated guess. And if that’s the case, the results should be reported as such.” By the journalists reporting on them, that is.
Just for the record, the Herald story noted, “the Angus Reid poll was conducted as an online survey among 802 randomly selected Alberta adults who are Angus Reid forum panelists.” There’s no information in the Herald story about how the poll was weighted.
Another suggested rule the pair provided was for readers to take the traditional margin-of-error disclosure with the proverbial grain of salt. Merely disclosing a margin or error or listing the questions asked “doesn’t represent meaningful disclosure,” they stated.
They advised journalists: “Be honest when something looks dodgy – either don’t publish it, or publish it with an editorial disclaimer.”
Not that I’m saying anything looks dodgy about this poll, which the Herald reported as if it came down from Mount Olympus with a chorus of political scientists in white robes to reinforce its conclusions. Its margin of error, by the way, “is plus or minus 3.5 percentage points, 19 times out of 20,” noted the Herald.
Just remember, as the Herald doesn’t seem to have done, that Angus Reid doesn’t exactly have a sterling record when it comes to reporting Alberta election results. On Feb. 29, 2008, for example, they released a poll showing the following results: 43 per cent for the PCs, 28 per cent for the Liberals, 13 per cent for the NDP, 10 per cent for the Wildrose, 7 per cent for the Greens.
Three days later, on March 3, in the “only poll that really matters,” as politicians are forever saying of general elections, here’s what the voters actually did: 53 per cent for the PCs, 26 per cent for the Liberals, 9 per cent for the NDP, 7 per cent for the Wildrose, 5 per cent for the Greens.
Oh well, what’s 10 per cent, give or take?
It’s worth noting that most other pollsters got the 2008 election right within a couple of percentage points. It’s also worth remembering that several other Angus Reid polls have been dramatically different from the surveys done by other pollsters with a record of success in this province.
Does this prove Angus Reid got it wrong? Of course not. For all we know, they could have nailed it this time.
But it is a good reason for a word of caution, or a qualifier or two from the credulous Calgary Herald. Well, if so, don’t waste your time looking for it in the pages of that paper, because you won’t find it.