Let's make a politics/ideology test

Seeing as other political/ideological tests have been found wanting, we are going to try creating our own test that would evaluate your general political beliefs based on your positions on different issues.

The goals:

• Based on your answers, the test will rank you on three or more scales, to be determined.
• Based on your positions on these scales, the test will suggest political parties/groups that fit most closely with your views.
• The test will be not be nation-specific, insofar as possible (although it should suggest political parties from the nations in question, if possible).

The format:

• There will be, I think, more than 10 and less than 100 questions. However many we need, basically. They will probably be broken up by topic (economics, defense, public works, etc).
• Question format may vary as needed (checkboxes, radio buttons, etc). Yes/no is discouraged. Each question will have a don't know/don't care option.
• There will be the ability to rank the importance of each question (or topic, depending on the number and specificity of questions).

Other guidelines:

• People of all political beliefs are encouraged to participate. I promise to turn over a new leaf and not be sarcastic or obnoxious.
• This is not a forum for discussing what political beliefs are right or wrong. Take it elsewhere.
• The exception to the above: discussing the importance of different beliefs/issues is allowed, in order to decide which questions, topics, and scales should be present on the test.

Are we able to do statistical analysis or do we have to rely on intuitive scales? I.E. is it possible/practical to make a lot of questions and see which ones correlate and create axes on that or do we think up the axes first and ask questions later?

Scales proposed.

-Measures for

• Civil Liberty (Personal conduct)
• Economic Mobility (Monetary transaction)
• Government Authority (Power Structure).

Civil Liberty would measure the freedom of individuals on a personal basis, what they can or cannot do.

Economic Mobility would measure the restrains on individuals or companies when making business (If at all).

Government Authority would measure the power and scope of the national government. Leaning from Authoritarian in nature to Corporatist. Depending upon how much society relies/is subject to the latter or former power hierarchy.

edited 8th Feb '11 10:59:42 PM by CommandoDude

My gut feeling would be that there would be no meaningful way to do statistical analysis to create the axes because you would be thinking up questions in the same way you'd be thinking up axes (ie, you put in personal freedom questions because you're thinking about personal freedoms, so you may as well just make a personal freedoms axis).

But if you think generating axes statistically based on our questions would be doable/better, we can try it, by all means.

I think axes for religion and foreign policy might be good to include.

The simplest way seems to be to create categories for each question relating to the corresponding axis. Then have a 5 point response (Strongly/Agree-No Feeling-Disagree/Strongly) for each question.

I would refrain from multiple choice questions considering that some people may feel that they would select multiple answers if allowed.

edited 8th Feb '11 11:23:45 PM by CommandoDude

I'm not a big fan of numeric scales, honestly. Can be real unclear how "agree" and "strongly agree" differ.

I like "check all that apply" questions better.

Maybe numeric scales for things that can be meaningfully numeric, though. Like ideal tax rates.

One thing these tests always seem to miss: an "agree" answer tells you what a person thinks. A "disagree" answer tells you what that person doesn't think, but it doesn't tell you what they do think.

edited 8th Feb '11 11:34:19 PM by BobbyG

I remember a test that was done by essentially taking the important questions, seeing what people answer and getting a left-right axis and a pragmatism-idealism axis (British-based) on exactly the methods I'm talking about so it certainly can be done. Getting enough of a sample might prove problematic though.

On the questions, I think that taking an issue, listing the most common types of solutions and letting people either choose which ones they'd support, be ambivalent about or oppose, or alternatively rank the desirability of each solution (incl. "none of the above) might be a good way to avoid the problem of agree-disagree questions.

On scales, economy is the obvious one. Then there's civil liberties versus security; liberal values versus traditional values; interventionism versus isolationism (maybe "imperialism" versus "humanitarian interventionism" versus isolationism triangle?); centralization versus decentralization (as in "urban or rural" and another as in "anarchy or tyranny"); environment versus economy; paternalism versus right to be stupid; etc. We obviously need to consolidate some and ignore some. It might also be good to include some borderline strawmannish questions to avoid extremes, if someone scores at an absolute edge of the scale it should be a case of actual extreme views instead of failure to extend far enough.

I would like this, so that we in the world might have an even better one than the political compass test.

People should have an option for an ultra simplified test, to as many questions as possible. Maybe even 500 or more. I would certainly answer that many for accuracy. So it would be nice if the goal wasn't less than 100. I'd love to see a test that has as little as 20 and as many as 500. Really, it takes a lot of questions to understand one politically. And we should strive, if we were to make something anything remotely better and more precise than the political compass, to ask as many questions as possible to people whom want to spend the time answering as many questions as possible.

And to make it global, there are all sorts of Nationalism. All sorts of Leftism. All sorts of Rightism. All sorts of Conservatism. All sorts of Libertarianism. A lot of questions to these weren't very well asked. If you ask something like "do you think that certain races are superior to others", most people are going to say no. Nationalists included.

To be accurate for the whole globe, many sorts of cultural questions should be asked. Whether they support the Chinese government. Whether they believe most gender roles are socially constructed. Divisive issues within the Libertarian Left such as Consumerism should be addressed. Of Localism and Isolationism versus Internationalism and Globalism. About integration, borders, separatism, multiculturalism, and so forth.

There needs to be questions which distinguish between Conservatism and Libertarian Rightism with Conservative leanings. Whether it is more important to preserve culture and family values, or allow the Market to be free. Whether one supports the European Union, Israel, and so forth, are all also important political positions.

Agreed.

I would also make a point of putting in items about moral relativism, ideals of governmental structure, fatalism and its importance in one's political perspectives, one's position on the importance of hierarchy in faith or politics, respect for versus deference to authority figures, and other subconscious determining factors, most of which are far more important to political position than superficial opinions on topics that one may not be well-rehearsed in.

An axis that denotes attitudes on freedom (that is to say, negative vs. postiive liberty) might be interesting.

@Gloomer Agreed. I actually think negative vs. positive liberty/freedoms is actually one of the biggest gaps that exists in politics (and I'd argue that it's actually the source of a lot of the conflict..the two simply don't play nicely together at all)

Should there be something on whether or not you find something important or not?

To make it politically unbiased and international, I'd say, I'm not sure if we should include questions of the "do you support Israel" kind. I'd rather write it in terms of Hypothetical Country X than Israel or China.

And then, a question about modifying the Constitution is a very different thing in, say, the US and France. It'd be good to catch things like these, and try to think up a neutral wording.

That's just my rambling, don't know if you find any worth in it.

Also include something about conceptions of justice; that might be important. Is justice about what is "right" or is it about what is "good"? Is justice essentially fairness, based on fundamental principles that all rational human beings would agree to from an position of complete ignorance (ie. the Rawlsian conception) or is it a system of values informed by the context of society (ie. the communitarian conception of justice).

How about questions relating to the sources of power in society? Is the liberal social contract justified by the threat of Hobbes' "state of nature"? Are the feminists correct to suggest that sex and gender are the most important factors in the allocation of political power? The possibilities for such a test are endless, really.

OK, let's first of all make a crowner to vote on what axes we should use. I like the idealism/pragmatism one someone mentioned: There can be a big difference between what you think would be ideal and what you think should actually be done.

Riffing on Uko's suggestion, perhaps we can have sections of the quiz be expandable. So it might start with a single question on a topic (like "Should the government be responsible for education?") and then you could expand it to get questions about vouchers, charter schools, home schooling, universities, etc. That way you can give more detailed answers for things you care about, but won't have to answer long lists of questions for topics you don't care about.

(I have no idea how to make crowners, btw.)

Are we going for a two-three axes forming a graph, like most political compasses do, or a large number of linear, I'd say, sliding scales, like a certain test I once saw?

I'd think axes encourage labeling extremes, rather than identifying nuanced philosophies.

It'd be more interesting to see if correlations can be found through analysis, rather than trying to determine axes from the start and try to assign weights to fit them.

Individual axis lead to extreme labels, but multiple axis create a position which could, arguably, be very difficult to set a label to, and the least of which is hardly ever 'extreme'

It seems as though we'll have, at a minimum, 5 axis already, as oppose to most quiz's two. (some three)

edited 9th Feb '11 1:33:04 PM by CommandoDude

Tangent: Be aware that we have a small sample size. We would need hundreds of people to take the quiz to create axes by statistical analysis, wouldn't we?

(It looks like we will have about 5 axes to me as well. A large part of the purpose is to address the deficiencies of 2-axis systems.)

edited 9th Feb '11 1:45:44 PM by jewelleddragon

Three axes can be safely put on a 2-D graph, but I'm not sure if more.

I've been giving this some thought recently. The axes for the scale/graph could be arranged as if they were the spokes of a wheel, and the position of the respondent could be denoted by how far the spaces in between each line are shaded in.

It would be akin to the kind of "character stats" graphs you sometimes encounter in supplementary material.

I'm not sure how to illustrate it.

It's been used in a test for the latest European elections and the site is still up if someone wants to check it out. Not a bad idea.

edited 10th Feb '11 2:00:52 PM by ChurchillSalmon

The "Analyse Your Nearest Party" option at the end is closest to what I had in mind.