Well done Herald

The Herald reports:

Most parents of secondary students believe the Government should agree to at least some of their teachers’ claims for a 4 per cent pay rise and smaller class sizes, according to a Weekend Herald readers’ poll.

But a majority still disagree with the teachers’ strike action.

A Nielsen Consumer Research non-scientific survey, conducted online for the Weekend Herald, asked nearly 400 parents of high school-aged children whether they agreed with the decision to strike.

More than half of the parents from the APN reader panel (comprising readers of the Weekend Herald, NZ Herald, Herald on Sunday, nzherald.co.nz, Listener, NZ Woman’s Weekly and APN regional papers) said they “disagreed” or “strongly disagreed” with the rolling strikes, which will continue for the rest of the term.

I hope the Herald keep the story as a template for how they report online panel surveys. The online panels are a valid way to get some idea of opinion on an issue, but they should not be reported as “poll findings”. All too often I see results from online panels reported as if they were scientific random selection polls.

What I like in the story is that the Herald has called it a readers poll in the lead paragraph, giving early indication it is not a random poll.

Then they specifically call it a non-scientific survey, tell us it was conducted online and how many parents they asked.

Later on they also tell us who makes up the panel.

Overall a really good example of how to correctly report on the results of an online panel.