It astonishes me how the media still treat polls seriously, given that they’re utterly fraudulent.

They’re fraudulent because they put a series of selected questions to people which lead to distorted conclusions.

By far the most farcical I’ve seen is one the Dominion Post front-paged a few months back, this concluding half of New Zealanders feel happy, that men were happier than women, that Hawkes Bay was the happiest location and Otago the least, and so on.

This poll was conducted by Stuff, but get this. It surveyed 2000 Kiwis across the nation for God’s sake, this out of 5.2 million.


2,000 may give a reasonable representation of the mood of NZ. But they probably only surveyed 20-30 people in Hawkes Bay from which virtually no conclusions can be drawn on the mood in Hawkes Bay.

..............dave...........smith December 15, 2022 at 12:33 pm

True true SB yes.

Though you may be forgetting how stupid ‘we’ are.

This is really, reaaaaaaaaaaalllllllllllly stretching my memory of Uni stats from many, many moons ago. But, I’ve got a feeling a survey of 2,000 drawn from a pop’n of 5.2 million should have been able to deliver statistically valid results – as long as the bits and bumps associated with conducting the survey passed muster.

Sorry Sir Bob but they’re close to valid. 2400 is a statistically valid sample size with 95% confidence and a 2% error rate.
However, your point is taken as places like Otago Uni often sample 100 or so and then publish the result which I media then usually repeats.

Leave a Reply

%d bloggers like this: