Found a link to The Political Compass Quiz over at Ilyka's place. The idea behind the compass is to show another dimension of political ideography beyond Left/Right. Specifically it separates the traditional social metric from the traditional economical metric. As the site says:

My initial impression of the quiz that plots you on the compass was favorable. Don't get me wrong, there are some obvious and very large problems with it. It is, however, an excellent step forward towards a very good idea. The most obvious problems are insufficient number of questions and several very convoluted questions. The first is a problem because a small sample field cannot be used reliably in statistical measurement. The larger the sample, the more accurate the result. The smaller the sample, the greater the chance of statistical anomaly. Confusing, poorly worded or misleading questions throw monkey wrenches into the works. This is exacerbated by the small size of the original sample.

A larger but less obvious problem with the quiz is that the questions lack weight. You may agree or disagree with a quiz statement or you may strongly agree or disagree. The quiz does not ask you how much that statement matters to you. Let me show you why. The four items in the following table are scored in the same manner as the ones on this political compass test.

I take this test and my answers are A,D,A,D. The results of my test indicate that I am dead center on both the social and economic scales. My agreements cancel out my disagreements leaving me at 0,0. But wait. I don't give a fig about the issue on question #1, I just strongly disagree with the statement. Now question #2 I am both strongly in agreement as well as passionate about the topic. If we can record this "weight" for each question we can reduce the effect of topics that don't matter (question #1) and increase the effect for topics that do matter (question #2). With the points weighted I'm suddenly a good distance to the right on the social metric.

Weighted polls like I describe have been going on for a long, long time. They have never really caught on because they are generally not needed - most polls are simply constructed to assign beans. "Will you vote for Bush? Yes/No". It doesn't matter how much you care about voting for Bush because you have one vote and it will either be cast for him or it won't. You can't half vote or 2/3 vote. They care only about the absolute values.

The poll/quiz they are using for this compass cannot work without weighting the questions as it is trying to determine relative values. As I illustrated in the example above it is far too possible for the results to skew without weights. So overall I don't put much stock in the results I got for my Political Compass but I like what they are trying to do here; the inclusion of a second major metric in political placement is absolutely necessary in my book. I'll wait impatiently for a proper quiz. In the meantime, go to the site, take the one that's there to see what it's all about and then place yourself on the graph where you feel you should be. It'll probably be more accurate than where the quiz places you.

By the way (in case you haven't guessed already from the post title) my results were Economic (Left/Right): 0.12 and Social (Libertarian/Authoritarian): -0.10. Those are on a scale of -10 to 10 so I'm about as dead stick stuck at the center as it's likely anybody could get.