A news.com.au story asserts
“Australians would re-elect Prime Minister John Howard in a landslide if votes were cast purely on policy – not personality or party loyalty – according to the first results from a “blind voting” tool developed by NEWS.com.au.”
Almost 35,200 of the 72,300 participants who have completed the Vote-a-matic were matched with the Coalition, compared with 28,700 participants who were identified as a fit with Kevin Rudd’s Labor Party.
A further 6700 participants have been matched with the Democrats, while around 1700 went to the Greens and 175 to Family First.
The trends in results are a reversal of findings in the authoritative political polls which indicate the Coalition could face catastrophic defeat in the election, due to be held before the end of the year.
This is not even close to true. Firstly, the results come from an on-line survey and are not representative of the electorate. Secondly, because of the way the tool works, the results don’t even represent the policy preferences of those that take it.
To see why, look at the questions in the “Leaders” section:
(1) Who do you spend the most time talking to at parties?
A. Someone you’re familiar with (Coalition 1 point)
B. Someone new (Labor 1 point)
C. Someone with the same interests
D. That friend you’ve lost touch with (Democrats 1 point)
It’s not a good question in any case, since it’s about parties rather than what you want in a Prime Minister, but the scoring is biased against the Greens and Family First — there is no way they can get any points, no matter what you answer.
(2) The most important thing in a prime minister right now is…
A. Experience and a strong track record (Coalition 3 points)
B. Important friends (Coalition 1 point)
C. A fresh approach (Labor 2 points)
D. The Dalai Lama on speed dial (Greens 1 point, Democrats 1 point)
This one is even worse — the Greens and Democrats only get points from the joke answer, two of the answers give points to the Coalition, and only the Coalition and Labor can get more than one point from this question. And even if Family First’s policies perfectly match your preferences, this question gives points to some other party.
The scoring on the other questions is similarly strongly biased against the minor parties and somewhat biased in favour of the Coalition. It may be that voters prefer the Colition policies, but this survey is useless as a guide to voter’s preferences.
Also unimpressed with the survey are Phil Gomes, Jeremy Sear, and Tim Hollo, while Bryan Palmer has a much better test that matches you with a political party.