?

Log in

No account? Create an account

Previous | Next

More leftwingery

Another political quiz, via nwhyte:



Crime and punishment, internationalism

Your position on this axis is -5.8
You are likely to be very internationalist and rehabilitationist.

Economics, etc

Your position on this axis is -1.2
You are likely to be slightly socialist and anti-war.



More interesting because it gives you a lot of data and comparisons - my results are here. The results for the vertical axis are much more clumped together than the horizontal axis, and the combination of economic and military answers often balances out, I think.

I'm Very Left Wing, apparently. More shocking news. 95.3% of the population are significantly to my right on the horizontal axis, which is a little bizarre - I don't think of my opinions as particularly extreme. I was going to say web survey = self-selecting sample, source bias, etc etc, but it turns out the comparisons are based on a properly representative 2,000 person poll of the population - there's a very interesting PDF with graphs and contour lines of those results. I particularly like the three at the bottom that are presented without comment. (And if you're avoiding it because it's a PDF, get Foxit Reader immediately and you will never have to worry about PDF problems again.)

Clicky here to take the test.

Tags:

Comments

matgb
Jan. 22nd, 2009 11:46 am (UTC)
No, no fallacy at all. I had a discussion about my qualms over it with the original author (RIP).

It's based on strong survey data with a lot of backup, asking people to prioritise their issues and take their opinions into account.

A lot of people are both eurosceptic and in favour of tough punishments.

I came out in the middle of the economic graph as well.

My main dispute is the socialist vs markets thing—socialism works best within a competetive market.

Read the (excellent) pdf linked—I'm going to contact Tom later asking if I can help do an updated version, I suspect the axes will have shifted again.
zagreb2
Jan. 22nd, 2009 05:50 pm (UTC)
>A lot of people are both eurosceptic and in favour of tough punishments.

What if someone isn't though (and such people certainly exist). Just because many people hold these views doesn't mean they should be positioned as inclusive on a graph because they can be mutually exclusive. Your point about markets and socialism is similar.

This is the whole problem with this sort of "on a graph" positioning of beliefs.
zagreb2
Jan. 22nd, 2009 05:57 pm (UTC)
The best way to work out this sort of thing would be to have a range of sliding scales like so:

Economy: Free ---- Regulated
Redistribution: Redistributive ----- Non-redistributive
Foreign Policy: Isolationist ---- Interventionist
Law and Order: rehabilitaion ---- punishment
Social issues: liberal ---- authoritarian
matgb
Jan. 22nd, 2009 06:01 pm (UTC)
This is the whole problem with this sort of "on a graph" positioning of beliefs.

Well, obviously. It's an aggregate abstraction that gives a rough idea of where you are overall, if you take it as a definite "I am this" then it's not at all helpful, but as an overall positioning tool it's not that bad.

I score roughly in the middle on the up/down axis because I have strong views one way and strong views the other way, which put me roughly in the middle. Chris wrote about how he originally sampled the data a lot more here:
http://ex-parrot.com/~chris/wwwitter/20050415-my_country_right_or_left.html
And Joe has come up with a better explanation of how the stats themselves work plus a critique of the question set here:
http://joeotten.blogspot.com/2009/01/principal-components-of-principles.html

It's an abstraction, and thus is flawed, but it's a much better abstraction than any other I've seen, and is at least very open about how it works and why you're positioned the way you are.
zagreb2
Jan. 22nd, 2009 06:06 pm (UTC)
But the notion that people split easily into political camps on "two axes" is completely flawed for the reasons I said. What about people who agree strongly with one side of "their half" but not the other? They are thrown in with it whether they like it or not. Thus my hardline socialist became "anti-war". What if he was strongly pro-war but capitalist? He might have ended up in the same place and that makes no sense.

It's one thing to say "this side is capitalist, this is socialist" (and even that's not that simple, as you've correctly stated) quite another to say "this side is socialist and pacifist, this side is capitalist and pro-war. For that reason I don't think this graph does what it's supposed to, except for those with very orthodox political views in which case they might as well be on a simple left-right spectrum.
matgb
Jan. 22nd, 2009 06:15 pm (UTC)
It's an abstraction and simplification. Of course it can't be taken as read.

But the statistics analysis shows that, as an abstraction, it is fairly accurate as a predicitve tool. If you have one view associated with a side of an axis, you are likely to have other views on the axis.

You could split it up into a massive multidimensional array and cover many many issues, it still wouldn't be 100% accurate. But it's much more accurate than the old "high taxes=left wing" approach, which is partially what the author set out to do.

The methodology and predictive ability is sound and testable (they did actually test it against other survey data), the margin of error isn't tiny, but it's not massive either.

Yes, there will always be outliers—several people are "close to the BNP" including the guy who married an immigrant, but that will always be the case.

The notion isn't "completely flawed"—it's generally fairly true. I'm in no way claiming it's flawless, but if you want to dismiss it out of hand that's your choice; I find it a useful analytical trick to get people to think, which is always useful when you seek to persuade people to vote for a different party to the one they've always voted for.
zagreb2
Jan. 22nd, 2009 06:23 pm (UTC)
But the statistics analysis shows that, as an abstraction, it is fairly accurate as a predicitve tool. If you have one view associated with a side of an axis, you are likely to have other views on the axis.

But then, as I said, you might as well divide the thing into "left and right" like we've been doing for centuries now. That's a simplification but those on the "left" and those on the "right" tend to be fairly predictable in the main.

Thing is, it's not that simple. The people who put this together recognise this, try and make it all a bit more complex and then create the problem of not having a space for my Yorkshire socialist chap and end up completely misrepresenting him. Hence why I don't think it's any good.

As I said, if we really want to break down political allegiances we have to do it as a series of sliding scales. A graph like that can't cover it without being innaccurate.

Interestingly, in the PDF the authors say "it used to be very simple, there was left and right" which is simply wrong. Even when "left" and "right" were created back in the days of the French revolution you can bet that those who supported the ideals of the revolution were no more homogenous than those who opposed it. And it simply got more and more complex. The "old left" of the radical liberals was challenged by the (rather different) socialists who have largely supplanted them. Similarly the "old right" of the pro-aristocratic, pro-church conservatives were challenged by the business class and then later new and more radical movements like fascists and religious fundamentalists. Left and right have never been "simple" terms in the past any more than today, they just make handy generalisations.

Profile

bad wolf
pickwick
Notes from extinction

Latest Month

November 2010
S M T W T F S
 123456
78910111213
14151617181920
21222324252627
282930    

Page Summary

Powered by LiveJournal.com