23 Aug

More Confusion

Published by MacDoctor on August 23rd, 2009
13 Comments »

My previous post was designed to add some clarity to the issue of how John Key should respond to the massive “No” vote in the referendum. However, I note that there seems to be much confusion amongst politicians, the media and even in the blogosphere about the result. The Editorial in the Herald on Sunday ”Referenda disservice to voters“ is a fine example of this confusion, rolling a number of fallacies into one column, as it were. The only disservice being done here, of course, is to the science of statistical analysis. In the interests of statistic sanity, MacDoctor feels it is his duty to list the silliness and explain exactly why these people need a course in statistical analysis before looking at a poll ever again.

1. Bradford Mathematics:

“Bradford says even though people care passionately about the issue, just over half of eligible voters cast votes.

““When you add the yes vote and the spoilt vote to the number of voters who didn’t vote at all the figures are about even,” says Bradford.””

I’m sure Ms. Bradford has also worked out that if you add all the people who didn’t vote in the 2008 election to the Green’s tally, then the Greens won the election and National has stolen the government through trickery. Of course, if all the people who didn’t vote in the US presidential election had been counted as McCain votes…

2. The Low Turnout Myth:

“For a start, barely half of eligible voters took part (by contrast, more than 80 per cent of ballot papers were returned in the 1997 referendum on compulsory superannuation)”

DPF has pointed out that the 54% turnout was only 1% off the turnout for the MMP vote. Be that as it may, in statistical analysis a “sample” of 54% is enormous. I’m sure David could calculate the margin of error for this particular “little” sample of 1.5 million people. To imply that this is somehow not a decisive vote because not everyone was counted shows a deep lack of understanding of voting and poll analysis. The only valid assumption you can make about the 46% of the voting population who did not vote is that it is highly likely that 88% of them do not think that smacking as a part of good parental correction should be a criminal offense as well. All else is complete speculation. If you wish to find out why the 46% did not vote, you will have to poll them. It is as simple as that.

3. The spoiled paper myth:

“Second, the number of so-called informal votes is suggestive. In the last general election, the party-vote informal figure was one vote in 200. This included those spoiled in error – when a confused voter gave a tick to two parties, say – and those who had written messages on their ballot paper. In this referendum, the figure was 20 per cent higher than that – nearly 10,000 spoiled papers.”

So the editor’s contention is that this now constitutes a “protest” vote over the “ambiguity” of the question. This is speculation upon speculation as we have no idea that this is a protest about anything, neither do we know what the protest, should there actually be one, is about. The “protestors” could just as equally have been protesting that the vote was postal instead of tied to the last election, or protesting about the “waste” of money or…

None of that matters. The extra spoiled papers numbered less than 2000 out of 1.5 million, making it a laughably weak protest and probably close to the margin of error in even this huge sample.

4. The Ambiguous Question Myth:

The perennial favorite of those who find the results of referenda unpalatable, is the ambiguous, or loaded, question myth. Of course, it usually escapes the objectors notice that the question must first be approved, to ensure that it is not loaded or ambiguous, as part of the process towards a CIR. In the case of the smacking referendum, the process was so long-drawn out, that multiple smaller polls were taken to see where the result may be headed. Many of these did not ask the same question, or asked other questions around it. All found that the displeasure with the repeal of section 59 remained constant at around 80% or so.

As a person who has taken many multiple choice questions in his life, I can tell you that any question seems ambiguous when you stare at it long enough. But if someone bothers to go back to the people who voted “no” and ask them what they thought the question meant, you would undoubtably find thousands of differing answers, but all of them would have a common theme: Smacking is a normal parental behaviour and the government has no right to criminalise it. Sue Bradford, the editor of the HoS and John Key may think otherwise, but the referendum is telling us that 90% of the country think this. There is no other scientific conclusion.

Add a Comment