Monday, August 29, 2011

Bias in the Political Compass Test

I was taking the Political Compass Test, and I noticed something interesting… there’s no “neutral” option for answers, which is bad because some of these questions seem to be biased, particularly toward libertarianism.  What I mean is, I seem to be forced into appearing to disagree with a statement that is actually a far more complicated issue than is presented to me.  (Skip down to the bottom of this post if you want the smoking gun.)

Here are some examples of question bias:

  • “Military action that defies international law is sometimes justified.”  Is the international law corrupt, or is it justified?  And what is the goal of the military action?  ”Sometimes” is a very vague concept.
  • “Controlling inflation is more important than controlling unemployment.”  When?  How bad is unemployment?  How bad is inflation?  This isn’t always an either/or situation.
  • “The freer the market, the freer the people.”  Does this even mean anything?  It’s like saying “fancy is as fancy does” or some other such BS.
  • “All authority should be questioned.”  Asking questions doesn’t mean being anti-authority or even necessarily disagreeing with said authority.  This is far too ambiguous a statement to base a political ideology upon.
  • “All people have their rights, but it is better for all of us that different sorts of people should keep to their own kind.”  Which people?  There’s a difference between expecting blacks to keep to their own kind and serial rapists to keep to their own kind, for example.
  • “The most important thing for children to learn is to accept discipline.”  I don’t know of anyone who would say this is literally the most important thing a child could learn.  This question is obviously biased so as to be anti-authority (pro-libertarian) and also has nothing to do with political ideology anyway.
  • “When you are troubled, it’s better not to think about it, but to keep busy with more cheerful things.”  Not even relevant!  Come on!
  • “A significant advantage of a one-party state is that it avoids all the arguments that delay progress in a democratic political system.”  That may actually be a significant advantage… doesn’t mean I’d actually prefer a one-party state.  Nuance, people, nuance!
  • “Abstract art that doesn’t represent anything shouldn’t be considered art at all.”  Lol wut.
  • “Astrology accurately explains many things.”  I quit.

So now, the ultimate challenge: I “strongly disagreed” with literally everything on the list and ran the score.  The results were way into libertarianism, though on the line between left and right.  They did manage to balance the left/right scale, but they were not at all balanced on the libertarian/authoritarian scale (or as my friend Jay put it: “This compass definitely needs some declination adjustment”).  A well-written test will score perfectly center if the questions are all answered the same.  This test was clearly written in such a way as to make people squeemish about agreeing with the statements, thus allowing them to nudge their own score into libertarian territory.

In short, the Political Compass Test is a recruiting tool for an underdog political agenda, and is not to be taken seriously.