The adjustment is that they ask some of the same questions, phrased somewhat differently. In other words, they seem to try to account for being against A but not for !A.

I noticed that factor towards the end, when I thought, "I answered this question before, but I can't remember what I said!" It was actually a little complicated to suss out the reversals.

That said, I landed in the same place I did before--which I think is just where you landed!

FYI, my score is: 1 left/right -2.9614 (-0.1783)
2 pragmatism +1.7948 (+0.1080)

I'm starting to lose a little faith in these surveys. Somehow, I'm now far more of a lefty than Mike (left/right -5.3424(-0.3216), pragmatism +2.0829(+0.1254)), even though I was to his right on the political compass quiz. Whatever my true political identity is, I find it damn hard to believe they've pinned it to the ten-thousandth decimal place!

I also wonder about the geographic identity precursor question. Does Long Island count as "United States Minor Outlying Islands"?

I love the fact that the survey purports to be "open and unbiased."

Let's see: they show Margaret Thatcher as being approximately similar in left/right politics as Adolph Hitler. That seems like a fair comparison. Off to a good start!

And how about some of the "statements" (and the converses) you have to agree or disagree with:

Art shouldn't receive government funding (vs.) Some art should receive government funding.

Note that there is no "all art should receive goverment funding," which would be a more appropriate converse.

"Children should always obey their parents" vs. "Sometimes children shouldn't obey their parents."

Again, it is not a fair converse to pair an absolute with a relative.

"No woman should be allowed to have an abortion, whatever the circumstances." vs. "Abortion is acceptable in some cases."

A true converse to the first statement would be "A woman should always be allowed to have an abortion, whatever the circumstances."

The list goes on with similar problems.

Basically, this is the same as the earlier political compass - it uses misleading wording to push you towards identifying yourself as "left," and makes statements attributed to the "right" as disagreeable as possible.

left/right, -7.4081 (-0.4459); pragmatism, -0.3359 (-0.0202), somewhere between Tony Benn and Charles Kennedy.  Probably a pretty accurate assessment, although I did find some of the wording a bit questionable.

Wow. I'm not the furthest to the left here. Ashley, I'm impressed!

I'm left/right -7.4081 (-0.4459); pragmatism -0.3359 (-0.0202). It's probably because I was saying "agree" and "disagree" a lot instead of "strongly agree" and "strongly disagree."

And today, I'm a borderline pragmatist. :-)

David

... Or rather, an idealist on the borderline to pragmatism.

Ask me again tomorrow. :-)

Unsurprisingly, I'm the most far-out person here.

http://politics.beasts.org/scripts/results?surveyid=1219809542

-9 on the left (beyond Ken Livingstone)--but---I'm right on the border of pragmatism (.26)

Interesting. And, as Jim suggests, not so useful...

Jimpy, I think you're misunderstanding what the survey creator is trying to do with these converses. He's not trying to cover a broader range of political perspectives; he's trying to eliminate effects where a particular wording makes people more likely to either agree or disagree with the statement, by including two distinct wordings that should produce equivalent (but opposite) responses.

The statements you cite are converses in the sense of propositional logic. The opposite of "For all x, P(x)" is not "For all x, ~P(x)", it's "There exists an x for which ~P(x)". If the first is true, the last must be false, and vice-versa. Similarly, a consistent survey answerer who Strongly Disagrees with "Art shouldn't receive government funding" presumably should Strongly Agree with "Some art should receive government funding". We don't know how that answerer would react to "All art should receive government funding"; that's a logically distinct
statement.

Anyway, here the guy's actually measuring the degree to which these statement-pairs really act as converses in this sense.

Actually, if you scroll to the end of that last page I linked, you get the best bit: the eigenvectors are basically the lists of statements that lead to your placement on the two axes. The first eigenvector, I assume, is the one labeled "left/right" on the rest of the site; that makes the second one the "pragmatism/idealism" axis. Neither name actually seems like that great a characterization of the position-cluster that's been identified, but I can't do better, and the survey would probably be a lot less appealing if it told you "you're .1080 correlative to an eigenvector of responses to the following list of statements..." instead of "you're .1080
liberal".

I confess, I don't have the vocabulary or statistical skills to critique eigenvectors, so I'll offer a simple example of what I see as a damning problem.

Let's say three people take the quiz.

1. "Lefty" believes abortion is Constitutionally protected and should not be legally restricted, under any circumstances whatsoever.

"Righty" believes abortion should basically be illegal, though he would make exceptions in the case of rape, incest or where the life of the woman was in danger (although even then, only in the first two trimesters)..

In our society, we would consider Lefty and Righty to be radically different on this issue.

So they both take the quiz, and the first statement is:

No woman should be allowed to have an abortion, whatever the circumstances.

Lefty will strongly disagree, of course. But what can Righty do? Well, he has to disagree too. The words "whatever the circumstances" make the statement so strong, he can't agree with it even if his only difference is that he would allow first-trimester abortions for rape victims.

What if they got the converse?:

Abortion is acceptable in some cases.

Lefty is going to strongly agree with this. Righty? Well, he has to agree with this too, because he would allow abortions in very limited circumstances.

(more)

So, even though Lefty and Righty are on opposite ends of the spectrum, they give basically the same answers - the pro-choice position.

In fact, I would argue that the only way you won't answer either statement as basically "pro-abortion" is if you literally believe every single abortion must be prohibited - rape, incest, mother dies, abortion pill, etc. Otherwise, the wording of the two statements naturally pushes you to be pro-abortion.

Oh sure, true pro-choice advocates will add "strongly" as a qualifier, but almost everyone, regardless of their true placement on the political spectrum, will answer approximately the same way. So the poll numbers get slanted to make it look like more people are pro-choice (traditionally a left position) than really are.

Interestingly enough, this is shown on that link you sent. Scroll down to the abortion question and you'll see nearly every single person who took the quiz disagreed with the proposition and agreed with the converse. Basically, it is nearly impossible to do otherwise.

So, In my example, pro-life Righty comes through on this quiz as basically pro-choice, which he totally isn't.

Frankly, I think a lot of these questions pull the same trick. I'm the third guy - "Centry." I took the quiz, and halfway through it thought it was such bullshit that I started answering questions as if I was a radical right nutcase - privatize roads! No tax on the wealthy! Immigrants bring no benefits to our country! Minorities aren't needed in government! I still ended up with a negative left/right score.

So then I did the whole quiz, from the start, answering the "righty" answer for every question that pertained to economics or politics (but never strongly), and "no opinion" answer for every question that pertained to religion or society (like sexuality).

That got me a score of +2.9, which according to the "fair and balanced" chart, puts me just to the right of Stalin.

Yeah, I guess if you believe in private trains, you are Stalin. Great survey.

O.K., last thought. Re-reading my comments above, they sound a little harsher than I mean to be (sorry). But, in addition to realizing I sounded like a sourpuss, I was able to put my finger on what was bothering me about the statements:

I don't think it is intellectually honest to oppose an absolute statement with a relative.

For example, what is the "converse" of the statement:

"Some art should receive government funding"

The survey claims the proper converse is:

"Art shouldn't receive government funding."

I would argue that, if that is true, it should be equally true to say that a proper converse is:

"Art should receive government funding."

I would further argue that the best converse would be a simple opposite:

"Some art should not receive government funding"

I guess the problem is that it is really, really hard to agree with absolutes, and really, really easy to agree with relatives. And I was bothered by the fact that the survey seems very heavy in assigning conservative positions the absolute statement.

Imagine if the paired "abortion" statements had been:

"Any woman should be allowed to have an abortion, whatever the circumstances." vs. "It is acceptable to restrict abortion in some cases."

The die-hards wouldn't change their views, but I bet the people in the middle would come out a lot different with that kind of phrasing.

I promise not to talk any more (unless some foolhardy person asks me a direct question).