UMR = Underwhelming Marketing Rubbish

You’ve seen or heard the UMR poll into the Auckland Mayoral race, right? 800kb PDF is available from Scoop here:

Len Brown – 1 point ahead of John Banks!

Sounds nice and newsworthy, doesn’t it?

But take a closer look at the poll of 482 Aucklanders.

It was actually two separate polls of 241, taken over June and July (two weeks between the polls).

Now, I’m not a polling whiz. But David Farrar is. He might shed some analysis on whether two polls of 241 is statistically reliable as an indicator on Auckland voting intentions. But we can note the following:

From what I can see, a poll of 241 has a margin of error of around 6.5%. And what’s more a poll of 241 means that each area of the city would have had around 50-60 people polled in each. That’s right, 60 people to represent Manukau’s opinions, 50 people for Waitakere, you get the drift. There’s hardly any useful information to be got from such a small sample.

This poll was not done to provide useful information. How could it provide anything of use with such low samples, high margins of error and spread out over three weeks? (first polling started 18 June, last person sampled 18 days later)

The poll was done to see if Labour could talk up Len Brown. (probably the Labour Party people who would have paid for the political questions in the omnibus polling about toothpaste, the tunnel at Victoria Park and attitudes on bank fees)

As a side note, check out the PDF – nice big photo of Len Brown with wife on the front page, smaller photo of Banks by himself. Mmm, subtle betrayal of the heart there by UMR.

The media got once again got scammed by Labour and UMR and have again put their credibility at risk as well. It’s just underwhelming marketing rubbish.

What a shame the mainstream media didn’t take closer looks at the samples and methodology, instead of just running the headlines.