This is our monthly place to discuss topics that haven’t appeared in recent posts.
What are the books recommended by Overcoming Bias to get a better background in the area of recognizing and overcoming bias?
Yesterday I saw something that reminded me of Robin’s many posts on signalling, so I figured I’d share it with his readers and thought this Open thread would be the proper place. It was a study on civic and citizenship educatedness, which received quite a bit of media attention here in Finland – naturally, since we did quite well in it. But when I looked into it, I was surprised about the way the survey was conducted. Here’s an example item (from here, page 63).
A local school has a volunteer day. On this day parents volunteer to come to the school and paint the
classrooms. The parents are not paid for their work.
is a parent who does not like painting, but he volunteered anyway.
Write the best reason to explain why volunteered to help paint the classrooms.
Code 2: Refers to either or both of the two categories of reason listed below.
RC1. a concern for the common good
RC2. a desire to participate in the local community.
RC1. Suggests a concern for the common good as a motivation for volunteering. [Note: A desire to help others or to help the
school is seen as sufficient for this category.]
• He wants to contribute to the school.
• It’s the right thing to do.
• Because he’s helping other people, even if he doesn’t want to he would feel better
knowing he helped them.
RC2. Suggests a desire to participate in the school/local community as a motivation for volunteering.
• He wants to get involved in his child’s school.
• He wants to feel part of the local community and the school.
• So he can socialize and have fun.
Code 1: Suggests only immediate self-interest or benefit as a motivation for volunteering.
• He wants to look important/wants to look good.
• He wants to learn new skills.
• To make sure he knows the environment his child is working in.
“He wants to look good” gets only partial credit? So seeing people’s choices as signalling is deemed to be a sign of poor civic education. But I guess civic education isn’t about learning to see things as they are.
This comment isn’t really intended as a reply to George and doesn’t really advance by the discussion much, but Albert Camus’ short novel is a brilliant exposition of much of what Robin writes about.
I was wondering if Robin or any other commenters have read it and if so what they thought. I see Robin as being very similar to Jean-Baptiste Clamence, though I assume he still has many possessions!
>Albert Camus’ short novel
The name of the novel is The Fall.
Thanks Tobiasz. I’m an idiot…
Time to compare the amateurs on ‘Less Wrong’ with the professionals. Readers may recall how some ‘Less Wrong’ amateurs (you know, those people with no qualifications or real world achievements) rave on about how ‘Bayesian Induction’ is the foundation of rationality itself, the ultimate model for normative rationality.
Take a gander at this paper from two professionals. Notice how this paper is filled with detail: exhaustive arguments, numerous examples, impressive math, and crystal clear careful statements. I want everyone familiar with ‘Less Wrong’ to read this very carefully, (and remember, be sure while you’re reading to compare it to the standard of the typical ‘Less Article’ article). Here’s the link:
‘Philosophy and the practice of Bayesian statistics’
Let’s hear an abstract of what they think about ‘Bayesian Induction’ shall we? Quote:
“A substantial school in the philosophy of science identifies Bayesian inference with inductive inference and even rationality as such, and seems to be strengthened by the rise and practical success of Bayesian statistics. We argue that the most successful forms of Bayesian statistics do not actually support that particular philosophy”
“We examine the actual role played by prior distributions in Bayesian models, and the crucial aspects of model checking and model revision, which fall outside the scope of Bayesian confirmation theory“
It gots mafs!
Time to compare the amateurs on ‘Less Wrong’ with the professionals. Readers may recall how some ‘Less Wrong’ amateurs (you know, those people with no qualifications or real world achievements)
Ouch, mjgeddes. I’ve written a soon-to-be-published scifi novel, beaten up a gay basher, and talked my way out of numerous traffic fines and worse – doesn’t that count for something? Just because I elected not pay $15 000 for one more year of University and a piece of paper shouldn’t *completely* discount me.
Reading the article now.
mjgeddes, you have definitely a valid point in that “bayesianism” seems, sometimes, to be an ideologi down at Less Wrong rather than as would be more appropriate, a research program for formalisation of epistomology. (Non-degenerated research program in my view — but I am just an amateur.) Your criticism regarding amateurism at Less Wrong is, however, of the point. Less Wrong is a _blog_ — not a scientific journal, and has (to my knowledge) never pretended to be. Critique such as that of yours here is of course very valueble but could have been even more so if it had been more carefully worded and less rethorical. It would of course also have been more natural to post a crtique as this one directly on Less Wrong rather that at Overcoming Bias? Fortunately nhamann at Less Wrong noticed your post here, so there does now exist a discussion thread on your (and, particularly, Gelman´s and Shalizi´s) critique:
It should be clear that the whole ‘Less Wrong’/SIAI ideology is misguided.
Look, the paper clearly shows that Bayesian Induction can’t create new models for you, it can only assess the probability that a given model is true. This fits with my earlier intuitions that decision theory/ Bayes can be redefined in purely passive terms, as merely the means for selecting actions/assessing models from a predefined set.
To repeat: Rationality can’t create new models of the world or consider novel goals. It can only assess or choose pre-defined models/goals. Only Creativity can create new models and novel goals. Creativity requires categorization and categorization requires consciousness. Ergo, general intelligence without consciousness is impossible.
Any good studies that talk about bias that results from being brought up in supernaturalistic house holds?
Just a quick thought I had yesterday: if cigarettes continue to become more expensive, do you think at some point the price will be extreme enough to make smoking a strong signal of wealth? If so, will that cause smoking to swing back into fashion?
Robin, do you play video and computer games at all? If so, what are some you especially like?
I still enjoy your posts now and again, but I must say that I’m less interested than I was a year or two ago.
I like your stuff on big questions (the future, the origin of life, physics). I like stuff on predictions markets and alternative institutions. I like your published stuff on agreement and Bayesian rationality.
I’m generally tired of your signaling analyses and discussion of evolutionary psychology. And discussion of meta-studies showing that research in some area is biased. (I think this stuff is worthwhile, but it just doesn’t hold my interest as it did when I first read your stuff on medicine).
There are some things you don’t write much about but that I think I might find very interesting. Here are some ideas:
1. Description of how you came up with the ideas you’ve had. What kinds of reading or social experiences led you to important new ideas or changes of view. Some narrative detail might be good here. Did individidual teachers or colleagues convince you of things, or did you mainly come to your own conclusions?
2. Ideas on how economics and social sciences relate to natural sciences. You don’t talk much about actually coming up with some model for an economic situation and then testing it across a wide range of situations. What do you consider the empirical successes of economics? What empirical case-studies would you provide to the curious but skeptical non-economist?
3. Great ideas that might not be widely known or well explained. You’ve studied a range of fields. Which ideas have been most productive? Obviously signaling theory, Bayesian inference, evolutionary psychology, basic micro. But I’m sure there are other ideas less relevant for your day to day blogposts that you consider powerful and important. This might be somewhat tedious for you, but I feel like short posts summarizing the ideas and why they are important would be very interesting for your readers.
Aaron M —
What you actually find, all over the place, is people coming up to a guy with a cigarette in his hand and asking to buy a cigarette for 25 cents or so. More or less, this corresponds to “chipping” — a practice from maybe 50 years in which occasional users of narcotics would contribute money to a regular user to facilitate purchasing narcotics for a group.
Nicotine is the new heroin.
Cute bit of business speak irony (National Interest corp. to be a wholly owned subsidiary of Global Business Services).
Interesting from organizational studies perspective. National Interest Corp. is interesting, part of their mission claims to be to serve the American national interest, they hire from National Clandenstine Services (part of the CIA where classic cia agents work, including former directors of that agency). I’m interested in intelligence agencies as part of different social epistemological efforts going on in the world.
“ARMONK, N.Y., March 2 /PRNewswire-FirstCall/ — IBM ( IBM) today announced that it has completed the acquisition of National Interest Security Company, LLC (NISC), a privately held company headquartered in Fairfax, VA. Financial terms were not disclosed. The companies had announced a definitive agreement in January to pursue the acquisition.
The business will be integrated into IBM Global Business Services, operating initially as a wholly-owned subsidiary.”
Robin, who is your favorite evolutionary biologist? Will you please recommend some books on the subject?
Where’s Counselor Kagan’s demonstrated quantitative competency?
Who in the class of quantitatively competent public affairs experts are beating the drum to introduce this criteria for evaluating public administrative candidates?
‘Why Nobody Likes Nerds and Why you’re Justified In Hating Them’
(‘Less Wrong’ links in the above article)
… be a charity angel.