Rationality versus Intelligence | Hacker News Reader

Rationality versus Intelligence

project-syndicate.org

36 pointstokenadult25 comments

For those truly interested in improving their rationality, http://www.lesswrong.com is in a similar format to Hacker News and is dedicated to the subject.

More directly addressing this article, K.E. Stanovich has conducted some excellent research showing relationships between cognitive abilities and thinking styles. It is fully available online. General reasoning/rationality works are available here: http://web.mac.com/kstanovich/iWeb/Site/Research%20on%20Reas...

Here's a particularly relevant piece: http://web.mac.com/kstanovich/iWeb/Site/Research%20on%20Reas... (Various cognitive bias tests are conducted to explore relations between cognitive functions and challenge assumptions on normative models of rationality. Really interesting exploration of competence-performance dichotomy with focus on effects of thinking dispositions vs. algorithmic limitations. The effect of thinking disposition in achieving normative rationality is an important consideration that applies to considerations for improving your own thought/action models)

To the results oriented person, a physics class is not an intellectual exploration of the properties of physics, it is a series of lectures, concepts, exams, and a grade.

Someone who focuses too much on the learning/conceptual aspect might be tremendously insightful about some of the concepts, but may omit others because they seem less interesting, etc.

If tested on peak reasoning ability, the "learner" might do far better, yet might do significantly worse academically.

Many people have a mix of both intellect and results orientation, but they are independent characteristics. I've known some relatively non-bright people who are very successful -- they find what works and run with it.

IQ alone is only good for doing well on IQ tests. That doesn't mean it's a useless measure, but it's far more signicant as a measurement of potential than of anything else.

This is why valedictorians get into lots of colleges -- they are proven results getters and have proven that they are able to identify the prize and win it. Someone may be brighter but not care about the prize or be too lazy to do what it takes to win.

You nailed it. This idea or distinction is certainly not new, anyone who's attended university or even high-school knows this somewhere inside, but this is the best verbalization I've yet seen. Thanks!

Now the question is, when is what strategy better?

Better for what? Valedictorians will suffer and sweat under an overseer of their own making -- goal from the future. Failed intellectuals will enjoy the immediate fun of solving a complex problem or satisfying their innate curiosity yet rapidly changing priorities as soon they get bored. At what price?

I personally enjoy the carpe-diemishly failed intellectual's path; I don't deem myself smart but I do enjoy hacking C++, Ruby and LISP out of pure fun and my college GPA is around 2.4 with several failed classes. I never finished high school due to bureaucratic crap I haven't liked to deal with. I've done drugs too, just out of curiosity. I probably look like a typical failed teenager but I can't look at myself with a pair of eyes other than mine.

My friend is a totally different story. Before his freshman year of high school he set a goal -- get a 4.0 GPA and get into a totally awesome university.

I watched him -- it was easy since we shared schools -- months and months, digging into assignments, spending nights at the library, working a part-time job and paying an SAT tutor. He was like a bulldog; he grasped the system by the very throat (yet he seemed like a prisoner of his own mind). He got 3.99 GPA and 2200 SAT (up from 1680). He kept focus for four years. I consider myself lucky if I keep focused on something for more than two months. He went to Berkeley.

Me? I consistently skipped school, learned stuff on my own, been the typical slacker, failed classes, and somehow picked up programming contracts while in high school.

Now he's finishing Berkeley and I'm taking classes in community college, just for intellectual stimulation, working part-time as a C++ programmer and towards my goal: frugal passive income from web-startup that would give me freedom to explore and learn, not being stuck in office (like he might).

My friend has prestige, respect and he is very succesful but I wouldn't live his life even for a second. I value the freedom from all that crap too highly :)

Different people want different things, I guess. Hope that answered your question.

You seem to have this implicit conclusion that being results-oriented is the better way to go. You highlight the non-bright people you know who are successful, and declare that they are results-driven. You highlight bright people who aren't successful, at least with the examples of prizes you pick out, and declare that they are not results-driven

What about the other permutations?

I can see some of the valedictorians doing something that you don't account for in your description, nor really in your description of the dumb but successful: after identifying the prize, they determined if it was a prize worth having at the cost it demanded. They became valedictorians in part because there were many other prizes that they identified, but decided not to pursue.

Not every prize is worth winning, and this is where a dumb, but results-driven perspective can get people into trouble. It doesn't matter what they are winning, so long as they have won. This strikes me as just as wrong headed as not caring about results at all.

I totally agree with you, actually. I was not meaning to make a value judgment.

However I think that it would be nice if everyone thought about what they really wanted,... such that they became more self-actualized, not necessarily better able to pursue the various gold stars that society offers.

The lead in is about Daniel Kahneman who won the 2002 Nobel Prize for his study of non-rational thinking in economic decisions. It's amazing that the noncognative aspects of economic decision making have been neglected for nearly 250 years. Those of you who have taken an introductory economics course have noticed the emphasis on "utils" and perfect markets. I guess this is because economics always wanted to be a science, even if dismal.

The article goes on to discuss the non-correlation between IQ and non-rational decision making and the lack of measures for rational/non-rational decisions. Non-rational doesn't mean irrational, rather it encompasses emotions, gut feelings, biases, and such in decision making.

Web developers and marketeers have some feel for this stuff. That's why we use web analytics and tinkering with design to see what gets the best click throughs.

Actually, in the terminology of the article, and in the terminology of Kahneman, the issue investigated IS cognitive. For the authors who have done this body of research, "cognition" is a broader term than "intelligence," and what goes awry in the wrong decision-making that Kahneman observed is cognition that is of poor quality, that is irrational.

The article claims that there ARE measures (tests) to distinguish rational decisions from irrational decisions. Many forms of irrational cognition, as the author of the submitted article acknowledges, are biased to go wrong in one way rather than another because of emotional factors and the like, but the error is still fundamentally a cognitive error.

It is a very good idea for web developers to test their assumptions with carefully designed observations. Sometimes those observations can lead to surprising, counterintuitive results that produce a significant business advantage. The author of the submitted article says (in other writings) that individuals developing strong skills of rationality is important in a world in which businesses try to take advantage of irrational thinking to sell products to consumers.

This comment is not precisely "on topic," but when I took an AI class at university, we used Artificial Intelligence: A Modern Approach (AIMA), by Stuart Russell and Peter Norvig.

I thought the book (website: http://aima.cs.berkeley.edu/) contained a good discussion of Rationality vs. Intelligence, which ended up coloring my views on AI in general.

Worth the read if you're into that. (And for those who may not know: it's a textbook, and very code intensive. Not for general consumption.)

So how is rationality defined and how is it measured? Does it correlate with any important life outcomes such as income, academic performance or law-abidingness? He says that irrationality can be "fixed." Does this fix result in measurable changes in peoples' lives?

He says that IQ and rationality are "very imperfectly" correlated. What is the number, exactly?

You seem pretty determined to persuade us that IQ is unimportant. I don't claim to be an expert on the subject, but HN readers should know there's more to the debate:

http://www.sullivan-county.com/id5/murrey.htm

Yes, I would have liked to see and example of what a question from a "rationality test" would look like. IQ tests are already accused of cultural bias - I can't imagine an objective rationality test.

Kahneman's test of the conjunction fallacy

http://books.google.com/books?id=FfTVDY-zrCoC&pg=PA66...

is culturally fair, in that people from all cultures fall for the fallacy, and there is no question about what the right answer is to a test that tests on that fallacy.

Did you read the article?

According to the article, you are acting rationally if you are behaving in ways that maximise achievment of your goal(s). So rationality does (tend to) correlate with doing well in the fields you mentioned (excluding law-abidingness, perhaps).

The article seems to be using people's susceptibility to the cognitive biases identified by Daniel Kahneman and Amos Tversky as a measure of rationality.

The author states that there are methods of "fixing" these errors, and that they should be a priority in education, instead of focusing on "intelligence".

As mentioned elsewhere, the community blog http://www.lesswrong.com is dedicated to the pursuit of instrumental rationality.

behaving in ways that maximise achievment of your goal(s)

But that's exactly what IQ tests measure, at least in the context of academic and job performance. IQ tests are used because they are the most economical predictor of achievement in certain areas.

Stanovich seems to be saying that he can predict some kind of achievement (or "well-being," whatever that is) by examining people's susceptibility to certain mental quirks like the sunk cost fallacy. Ok, it's a fine hypothesis, but where's the data?

behaving in ways that maximise achievment of your goal(s)

But that's exactly what IQ tests measure, at least in the context of academic and job performance.

The longer book, published only at the beginning of this year, by the same author includes numerous examples of high-IQ individuals NOT achieving their goals because of nimble but irrational thinking. The kind of tests developed by Kahneman (and replicated in practice by many other investigators, repeatedly) show that IQ scores are essentially devoid of predictive value in showing who will make rational decisions most consistently. (Most human beings don't make rational decisions particularly often, and having a higher IQ doesn't lower the rate of irrational decisions on many kinds of tests of rationality.)

The data can be found in the abundant citations to the primary research literature in the book by the same author. I'm all too well aware that you know how to find the book

http://news.ycombinator.com/item?id=622924

so rather than repeat myself here, I'll simply mention that the book is well evidenced, well written, interesting, and a good contribution to the popular literature by scholars on cognitive psychology.

After taking the Myers-Briggs test again several years ago as part of a team bonding exercise, I remember reading the description of my subtype (ISTJ) and it saying 'in times of stress, you may tend to act rashly and irrationally'.

I believe that even the most logical people will act emotionally under stress.

Isn't the meaning of the cliche that 'someone is human' that they make mistakes/mess up/show weakness?

Logical people aren't just irrational emotional people underneath who use logic when it's easy. I think really logical people fall back on that as their core behavior, even under stress. That's their nature.

Logic isn't something we abandon under stress, it's something we cling too even tighter because that's when you need it the most. I've had jobs where lives literally hung in the balance; extreme stress. It only made me more rational and less emotional, that's how you cope.

The author's response to both comments above would be that all human beings reliably make many irrational ("illogical") decisions every day, and don't even notice themselves doing so.

I've always disliked the idea that the defining aspect of being humans is that we suck.

I've also observed this; I summarized it as: It's possible to be analytical without being scientific.

Good to hear that I'm not alone.

"Ironically, the Nobel Prize was awarded for studies of cognitive characteristics that are entirely missing from the most well-known mental assessment device in the behavioral sciences: intelligence tests."

. . . .

"Nevertheless, recent progress in the cognitive science of rational thought suggests that nothing--save for money--would stop us from constructing an 'RQ' test."

"nothing--save for money--would stop us from constructing an 'RQ' test."

And politics / political correctness. Anyone who actually made and promoted such a test would probably get shot.

The individual items for such tests have already been made, without any shots having been fired. The money barrier is that a test publisher would have to conduct a validation study of a battery of such items on a norming population, and a publisher will only do that if it expects a market to develop for such tests. If people don't want to know who is more rational and who is less rational--perhaps because of what they would find out about themselves--it may be difficult initially to establish a market for rationality tests.

Interesting. And now that I've read Measurement in Psychology I understand what that means. (Thanks for the tip, btw.) My thinking though was more along the lines that such a test would prove very controversial if it effectively allowed employers to screen out, say, all religious people or something like that. (Without actually asking any questions about whether they believed in god or what religion they were.)

Incidentally, do you know if anyone has ever made a test of intellectual curiosity?

Incidentally, do you know if anyone has ever made a test of intellectual curiosity?

Google tells me that people have worked on the issue.

http://www.eric.ed.gov/ERICWebPortal/custom/portlets/recordD...

http://epublications.bond.edu.au/greg_boyle/9/

http://www.informaworld.com/smpp/content~content=a785038693~...

I'm not aware of any test battery related to curiosity in wide clinical use.

Reply on news.ycombinator.com