In their 2011 chapter for the Cambridge Handbook of Intelligence, Stanovich et al. review the evidence suggesting that intelligence and rationality are not the same thing and that rationality is often more important than intelligence. They then lament the fact that there are no standard tests for measuring one's "Rationality Quotient." Then they take a few steps toward such a thing by suggesting some important rationality skills (actively open-minded thinking, fine-grained emotional regulation, tendency to seek information and fully process it, etc.) and rationality 'mindware' (probability theory, scientific process, economic thinking, etc.).

Here are those pages in particular: first a graphic of some important rationality skills and mindware, and then a table of the components of rational thought: rationality components, relevant literature citations, and example word problems that would test for each rationality component.

New Comment
22 comments, sorted by Click to highlight new comments since: Today at 2:49 PM

Ok, I'm getting this book.

Yeah, I bought the Kindle version a few days ago. Very handy for getting up to speed very quickly on our understanding of intelligence. I wish there was an Oxford Handbook of Rationality that wasn't from 2005 and didn't suck.

Do you mean handbook published in 2004? And could you please describe in a few words why does it suck?

It's too focused on philosophical issues, not sufficiently aware of contemporary rationality science, and its chapters aren't handy broad overviews but instead relatively narrow in scope. Luckily, we now have the Oxford Handbook of Thinking and Reasoning, which is quite good.

  • Intelligence quotient
  • Rationality quotient

I wonder what IQ and RQ you need to make up for a low social/emotional/ethical intelligence quotient. The latter 3 are in theory comprised in the former 2 but probably demand massive resources if they are not hard-coded.

ETA:

Autistic children, for example, are sometimes extremely clever. They're very good at making observations and remembering it all. However, it is argued they have low social intelligence. Chimpanzees are very clever at the level of being able to make observations and remember things. They can remember better than humans can, but they, again, are inept at handling interpersonal relationships. So something else is needed. What is needed is a theory of mind, a theory of how other people work from the inside.

Wikipedia

Chimpanzees are very clever at the level of being able to make observations and remember things. They can remember better than humans can, but they, again, are inept at handling interpersonal relationships.

Chimpanzees aren't that bad at handling interpersonal relationships. Especially compared to their cleverness in other areas. They have approximately the opposite problem as the autistics.

I am genuinely surprised. The parent is the last of my recent comments that I would expect to be disapproved of but it managed to get to -2. And not out of any evident general assassination either from what I can tell. That usually means that either people disagree strongly or, more commonly, that it goes against people's politics. Since I am not aware of any group that would find that particularly objectionable the obvious hypotheses are either two people somehow disagree about the relative strengths of chimpanzees or the meaning of my words are less transparent than I thought.

Given their level of cleverness chimpanzees are more clever at dealing with other chimpanzees than they are generically clever at tasks other than interpersonal relations. Is that not both true and trivial? The corollary seems equally obvious. Given their level of cleverness and relative to other individuals of their species autistic individuals can be expected to be better at non interpersonal tasks than interpersonal tasks.

(I didn't downvote the grandparent.)

I disagreed with that because it appears the difference between chimps and humans is that humans are better at social things, including modeling other humans. Chimps can learn, but they don't teach- probably because they don't understand that their children can learn.

The claim you're making seems subtler- that if you divide a chimp's ability to model reality by a human's, you'll get a lower fraction than if you divide a chimp's ability to model chimps by a human's ability to model humans. I don't know enough about their ability to model reality, but the impression I get is that chimps are pretty clever.

Chimps can learn, but they don't teach- probably because they don't understand that their children can learn.

Citation needed - this seems to disagree, and, more generally, I would expect the ability to learn to go hand in hand with the ability to teach. Anybody whose cat had kittens knows that teaching exists in animals.

From what I've read in Chimpanzee Politics, I got the impression that chimps are pretty good at modeling other chips (better than many other primates), so I wouldn't be ''that'' sure that the relative increase in humans was greater in the ability to model peers than in the ability to model reality (though most likely it is).

Citation needed - this seems to disagree, and, more generally, I would expect the ability to learn to go hand in hand with the ability to teach. Anybody whose cat had kittens knows that teaching exists in animals.

My cat used to teach us how to catch mice. It was adorable!

[-][anonymous]11y 3

Upvoted for adorable.

Citation needed - this seems to disagree, and, more generally, I would expect the ability to learn to go hand in hand with the ability to teach.

That seems to suggest that teaching is rare among chimpanzees near the end. Regardless, I'm not a zoologist, and my model of chimps is fuzzy. It may be that someone was commenting just about male chimps and I extended that to all chimps, or that most chimps don't teach except for this variety, or a number of other options.

Most of these were familiar to me, but I'm not sure of the basis or assessment criteria for this one:

Select the correct answer: Absinthe is (a) a precious stone or (b) a liqueur. What is the probability that the alternative you selected is correct?

My smartarse answer would be "1 minus epsilon", but as far as I can see, the only informative classes of answer in fitting with the cited measurement paradigms are either "1" or "< 1". Is that it?

What is the probability that the alternative you selected is correct?

I hate "gotcha" questions. How can we know that we have equivalent priors unless they are explicitly stated? And how can you establish probabilities without known priors?

[-][anonymous]11y 0

what?

Excellent, this would make it possible to evaluate the connection between IQ and "RQ" further and to determine the payoff to rationality, especially to determine how rational rationality is.

Indeed. As autistics and social intelligence was brought up earlier, maybe one also could find correlation between RQ and emotional quotient and RQ and social intelligence. It would definitely to be interesting to see if highly rational people do better in negotiation and understanding others.

Isn't rationality just a subset skill of the general intelligence encapsulated by the brain, intelligence being the set of skills your brain can handle or the processing power of the brain? Just curious as I would see some who is relatively more rational as just more generally intelligent. Essentially is intelligence processing power? And are labels like RQ, IQ, EQ just different ways of demarking[sic] the tasks that a particular brain is good at handling?

Intelligence is the raw processing power of brain. It is biological; it remains more or less constant during adult life. It is like a brain speed and short-term memory capacity. Computer analogy: processor frequency and RAM memory. IQ tests are trying to measure this brain power, not a specific skill. Different IQ tests may use different types of simple tasks, and yet their results highly correlate.

Rationality is how you use your brain. It can be learned (and LW is trying to teach it). Computer analogy: software with its features and bugs. Rationality allows you to reach nontrivial goals, where the raw brain power is not enough, it must be also used properly. Without rationality, you can expend a lot of mental energy and yet never reach your goals, as if you are running quickly but in a wrong direction.

EQ measures a specific area of skills.

In a very simplified metaphor, I would say that in a landscape of thoughts intelligent people "explore quickly" and rational people "navigate carefully". Low intelligence + low rationality = stays at the start or slowly follows the mob. High intelligence + low rationality = runs wildly everywhere, sometimes falls in a pit and cannot move from there (see Mensa). High intelligence + high rationality = avoids dangerous places and walks towards a distant goal.

And what about low intelligence + high rationality?? What does that equal to?

Not sure how much that combination is possible; whether some degree of intelligence is not necessary to be able to follow the rules of rationality.

But if it is possible, then "low intelligence + high rationality" would be going in the right direction, just very slowly. Then it would depend on the specific situation, on the distance to the specific goal, whether going slowly one has enough time to reach the goal and benefit from reaching it.

The Millionaire Next Door might be an example of average intelligence, high rationality. These are people who are unusually good at remembering their goal and doing reasonable things to achieve it.