Skip to main content
public editor

Getty Images/iStockphoto

Writing about opinion polls is kind of like sports coverage. You need to get the score right and you need to understand if a team is on the rise or stuck in a rut, but, whether that day or the next, a writer needs to explain what it really means.

This year, with a federal election looming, the coverage of opinion polls will be important.

One reader complained this week about confusing coverage. "So, according to Lysiane Gagnon the latest national polls show the Conservatives surging ahead of the Liberals (Even Harper wants Quebec, January 21, 2015)," he wrote. "However, Adam Radwanski informs us that the Liberals are running neck and neck with the Conservatives (Centre-left battle rages on unfamiliar turf, January 21, 2015). What is a body to believe?"

Adam Radwanski says that "matters have been complicated by opinion polls consistently showing the Liberals neck-and-neck with the Conservatives and the NDP back in third."

What he is discussing is a range of polls and the overall view.

Meanwhile, Lysiane Gagnon is pointing to "the latest national polls showing the Conservatives surging ahead of the Liberals. But then, the election is eight months away, and there is ample time for the Conservatives and the Liberals – by far the two leading parties nationally – to trade numbers."

The two are not mutually exclusive.

The latest poll Ms. Gagnon mentions is by Ipsos, which says the Conservatives had a strong poll showing. "Rallying Harper Tories (35%) break stalemate and shoot to early 2015 lead over Liberals (31%) and NDP (24%)."

However, if you scroll down on the above Ipsos link, you will see that the Liberals and Conservatives have switched places in the lead over the past year.

In this case, both Mr. Radwanski and Ms. Gagnon are right, and using terms such as surging or dropping are fair descriptors, in my view, as long as you are seeing a significant change of more than a point or two. Pollsters will talk about a virtual tie, which means the numbers fall within the margin of error of the poll.

In any original reporting about a poll, I believe the margin of error needs to be included. The reporting also needs to include the date on which the poll was taken, whom the survey was done for and what question was asked.

This will become more important during this year's federal election campaign when we start to see polls that may seem contradictory. It's worth remembering today, well before any election campaign officially starts, that for many reasons polls aren't the gospel and caution should be used in reporting them, especially on the eve of a vote.