Overcoming Bias - Latest Comments

Re: French Fertility Fall

Bock — Thu, 23 May 2013 21:56:54 -0000

Surely Balzac had something to say about this somewhere.

Re: French Fertility Fall

IMASBA — Thu, 23 May 2013 19:12:45 -0000

"But the problem with your hypothesis about women is that men don't care about stuff like a woman's education, or job*. We care about "hotness."

Have you ever had a multi-year relationship with a woman? There are a lot of things besides "hotness" that start to count after a while, even as far as back as the Roman Empire the women of the elite were educated so as not to bore their future husbands.

Re: French Fertility Fall

IMASBA — Thu, 23 May 2013 18:42:28 -0000

"High status folks focused their resources on fewer kids, and your kids had a big chance to grow up high status too if only you would also focus your energies on a few of them."

Yes, but this is no secret: people from third world countries will sometimes flat out say they have a lot of children to make sure there'd be someone to look after them in old age. With inceased income equality and social mobility fewer children are needed to look after two parents. I've also heard third world women say they'd rather have gotten less children but their husbands forced them to have lots for religious reasons.

So the reasons for lower fertility are this:

1) less children needed for old-age security (this factor is increased by the existence of pension systems)

2) loosening up/disappearing of religious pressures to "go forth and multiply" and religious rules against the use of contraception (Iran is a good example)

3) the availibity and low price of contraception

4) the emancipation of women: if you put off having your first child you reduce the maximum number of children you can have in your lifetime, the same happens when people are no longer forced to marry when they're young (they spend years looking for a suitable partner)

5) the increased technological level of society necessitates more time to be spent receiving education, during which it is not practical to have children (for social and financial reasons)

6) changing legislation regarding slave women, concubines, inheritance, divorce, alimony, DNA testing, etc... that make it much harder to just deny the existence of unwanted children

7) probably minor factors, but still, pollution reducing sperm count and gay people being not being forced into heterosexual marriage anymore (though some gay couples use medical science to have a child and adoption by gay couples may decrease infant mortality in society, perhaps even negating their lower fertility), finally I know some people explicitly have less children because they worry about overpopulation and environmental damage, a worry that didn't use to exist when science was in its infancy and most people couldn't read

Re: French Fertility Fall

Noumenon72 — Thu, 23 May 2013 17:35:40 -0000

I simply don't understand what this paper is trying to say, despite Robin's summary. The last line here says the landholding pattern is crucial, but the excerpt doesn't say which kind of landholding goes with low fertility or why. The same with correlation of wealth across generations, which way does it send fertility and why?

Re: French Fertility Fall

theakinet — Thu, 23 May 2013 17:15:51 -0000

"It seems to me this roughly fits with the fertility hypothesis I put forward."

But the problem with your hypothesis about women is that men don't care about stuff like a woman's education, or job*. We care about "hotness." And that trait peaks in the High School/College years. So that's when women should try to lock down a mate. The Mormons are right.

*Unless a woman makes more money, then we care about her job and don't give the relationship our all, because it's embarrassing, as a man, to be out-earned by your mate.

Re: French Fertility Fall

Doug — Thu, 23 May 2013 16:35:22 -0000

Two factors strike me as highly important that don't seem covered in the paper:

1) Primogeniture vs. equal inheritance. If only the oldest son inherits the bulk of the property than the legacy of the house can be preserved even with a lot of children. In contrast if legal or social rules promote fairly dividing the inheritance than high fertility will quickly ruin a house.

2) Rights of illegitimate children: Fathering bastards seems like a common way that rich and high status men in medieval Europe achieved high fertility rates. As women gain social status the cost of bastards is higher.

Re: Best To Mix Odd, Ordinary

Stephen Diamond — Thu, 23 May 2013 13:32:17 -0000

No, quite the opposite is the case. If conspiracy theories were all false, then non-conspirativists would be expressing their personalities through their beliefs less than conspirativists express their personalities, since they would also be expressing their epistemic superiority. ( http://tinyurl.com/6kamrjs ) The absence of conspiracies would vitiate the parallel that Hanson is drawing between the two stances, each more or less equally sensitive to truth.

Re: Sex At Dusk v. Dawn

Alexey Sunly — Thu, 23 May 2013 11:48:37 -0000

There is no complete story, Lucas :-) Science can not give you one, no one can give you one. Neither Ryan nor Saxon. Just read Barash's book I mentioned, and I think you'll see what I mean :-)

Re: Best To Mix Odd, Ordinary

Philo — Thu, 23 May 2013 11:43:09 -0000

“Some people just like to be odd.” And some people just *are* odd, whether or not they like to be so.

Re: Best To Mix Odd, Ordinary

gwern — Thu, 23 May 2013 11:16:23 -0000

Of course it does.

Re: Best To Mix Odd, Ordinary

Alexei Sadeski — Thu, 23 May 2013 11:08:46 -0000

Actually, it does not.

Re: Best To Mix Odd, Ordinary

IMASBA — Thu, 23 May 2013 08:43:02 -0000

The number of people who worked on mkultra, and the cost of the program were relatively small, although it eventually was revealed in the 1970s, also, the KGB probably knew about the program but didn't reveal it because they had a similar program (which is why a 9/11 truther conspiracy is so unlikely: the Chinese would have nothing to loose and much to gain from revealing such a conspiracy to the world). That there were Soviet spies in the US was no secret, everyone knew this, they just didn't know the identities of the agents, just like the Soviets didn't know the identities of Western agents in the USSR, but knew that they were there.

Re: Best To Mix Odd, Ordinary

dmytryl — Thu, 23 May 2013 08:20:53 -0000

Yet, as a matter of historical fact, there was a sizeable number of ridiculous conspiracies involving quite large number of people, which did not surface for decades or longer. E.g. mkultra, or US human irradiation experiments, just to name two.

From the other side, there was a fairly big number of commie spies everywhere including upper echelons in the US. Note that presence of commie spies in the secret US conspiracies didn't unearth those either because USSR ain't going to tell you what it knows.

(The wealth of intelligence may well have been the reason world war 3 didn't ever happen, though).

Re: Best To Mix Odd, Ordinary

dmytryl — Thu, 23 May 2013 01:58:40 -0000

Ordinary or not, your predictions have only an incredibly low probability of working, due to reliance on multitude of mostly unstated assumptions. (For the ordinary assumptions, it is perhaps more explicit, and so those who want to be persuasive pick extraordinary assumptions, which, even though even less likely to hold, look like there was more effort and less assuming. )

Predictions are almost always very unlikely to be true even without any evidence towards any contrary prediction, through the simple matter of it being a wild guess. If you had extreme inference ability, there are many areas (e.g. history where you can infer backwards then find confirmation) where you could get recognition for it, and far future societies are not among those.

You're not using that inference ability on such topics, because you want to win some status guaranteed, and because somewhere inside something in your head counts all those assumptions and knows that your probability of being correct is an epsilon. (It may be a larger epsilon than that of other guessers, true, but it is still an epsilon).

On the other hand, of course the non insane speculations are socially useful right now, even if incredibly unlikely to be correct, simply to provide some more samples and dilute influence of people who privilege ideas for fun and profit.

edit: an illustration on assumptions:

Suppose that you are to predict 10 digits from a pseudorandom number generator which outputs 0 with higher probability than any other digit. (E.g. it's a 20 sided die with 0 on 5 of the faces, 1 on 4 of the faces, 2 on 3 of the faces, 3 on 2 of the faces, and other digits on 1 face). You predict 0 0 0 0 0 0 0 0 0 0 . Your guess is the best. But those who understand your guess know you have about 1 in a million chance of being right. No one is going to be very certain in your guess.

Meanwhile this other guy, he tries to look like a living crystal ball with some superior powers of guessing, and he predicts something like 2 3 0 2 1 0 0 4 3 6 which looks like something that a crystal ball might tell, and so is more persuasive in some sense, even though, not being a working future-telling crystal ball, less likely to be correct. Some % people will give that guy's claims probability well above one in a million, and given enough reach, he'll have a larger number of believers. Not because his guess is better, and not because his argument is better, but because it is more deceptive. (Likewise, predicting 9 9 9 9 9 9 9 9 9 9 will perversely persuade more people in >2^-20 probability than your 0 0 0 0 0 0 0 0 0 0 , because former looks like secret knowledge and latter does not).

edit:
Let's give it a try. We are guessing about a PRNG, e.g. sha1 hash of this sentence converted to base 20 then used in place of 20-sided die described above. Let's suppose I say that the first digits will be 2 3 0 2 1 0 0 4 3 6 . You say, the digits will be 0 0 0 0 0 0 0 0 0 0 . Maybe I calculated the correct digits in my head somehow, maybe I actually calculated them on computer, who knows? You clearly didn't. So my number would persuade some people much better than your number would persuade anyone. (People tend to do this trick subconsciously, and it gets described as a fallacy).

Re: Sex At Dusk v. Dawn

Lucas — Thu, 23 May 2013 01:29:01 -0000

Saxon gave Brash's book 5 out of 5 stars. And from what I understand both their books are pretty much in agreement in their conclusions.

But you may be misunderstanding my position in this, I don't really care what the conclusions are, what I care about is how dishonest Ryan is in his book. Both Saxon and Barash are in agreement about that as well.

Of course facts can be be cherry-picked, that's what I'm saying, that's what Ryan is doing, but it's dishonest and no it doesn't show the truth, it shows an incomplete story - it misleads.

You seem to be ok with the fact that Ryan cherry-picked the data to support his view. If that's true, fine. We will just have to agree to disagree on that.

In any case, I'm exhausted and probably not being as clear as I'd like. So I wish you a goodnight.

Re: Sex At Dusk v. Dawn

Alexey Sunly — Thu, 23 May 2013 01:05:56 -0000

I am aware of Barash's opinions, the question is what does Saxon has to say about Barash' work... ;-) You thanked me for my thoughts, Lucas, but you are not listening. All facts can be "cherry picked" to support any truth, as long as you know what you are looking for, but you would know that if you were an academic yourself. If Barash is so committed to the "truth," how is that the facts he "picks" contradict Saxon's conclusions :-)

Re: Sex At Dusk v. Dawn

Lucas — Thu, 23 May 2013 00:22:31 -0000

Here is a review of Sex at Dawn by David Barash you might find interesting:

http://chronicle.com/blogs/bra...

Haven't read his book, would like to. From what I understand he is much more committed to truth/facts than Ryan is.

Re: Sex At Dusk v. Dawn

Lucas — Thu, 23 May 2013 00:11:22 -0000

Sorry Alexey, I have to disagree.

Ryan show a 'fact' by only showing the part he wants, that seems (in its incomplete form) to support his claim. Saxon shows the fact in its complete form.

Here's a made-up example:

Ryan says – "The research shows that villagers wore black, so wearing black is natural. You can wear other colors (like red or blue) but it's not natural, it's gonna be an uphill struggle."

Saxon would point out – "The original research shows that black hats were worn, and blue shirts and red pants were also worn."

FACT from original researcher – "The villagers wore black hats, blue shirts, and red pants."

See what I'm saying? That happens over and over in 'Sex at Dawn'. Ryan is not simply showing the evidence in a different manner, he is cherry-picking. Again, it's dishonest.

Out of curiosity though, have you read 'Dusk'?

Re: Sex At Dusk v. Dawn

Alexey Sunly — Thu, 23 May 2013 00:06:12 -0000

The Myth of Monogamy: Fidelity and Infidelity in Animals and People, by David P. Barash and Judith Eve Lipton

Why don't you review this particular book as well for us, Lucas, and tell us how it "misrepresents" facts as opposed to the two already mentioned :-)

Re: Sex At Dusk v. Dawn

Alexey Sunly — Wed, 22 May 2013 23:27:13 -0000

Sex at Dawn do represent the facts, Lucas. Only in manner different from the way Sex at Dusk does it. They both represent facts as it suits their authors. Science can only examine theories and "evidence." The conclusions are made by people like you. There is not right or wrong, only choices. Choices that make you happy and choices that make others either happy or unhappy. That's all. And you are welcome for my thoughts :-)

Re: Sex At Dusk v. Dawn

Lucas — Wed, 22 May 2013 23:11:45 -0000

Yes Alexey, and if wishes were horses, beggars could ride (or something like that).

This has nothing to do with what makes me happy - I'm perfectly happy, but thanks for the concern :P

I do make up my own mind – but I base it on evidence. That's how reason works. And that's how science works, as opposed to religion which doesn't use evidence – that's the difference.

'Sex at Dawn' doesn't show facts, it misrepresents them. It's dishonest.

Thanks for your thoughts though.

Re: Best To Mix Odd, Ordinary

gwern — Wed, 22 May 2013 22:00:21 -0000

Pointing out that there are many true conspiracy theories casts doubt on any meta-level discussion of conspiracy theorizing and inferences therefrom, is all.

Re: Best To Mix Odd, Ordinary

RobinHanson — Wed, 22 May 2013 21:54:54 -0000

How is what I said in conflict with that article?

Re: Best To Mix Odd, Ordinary

Stephen Diamond — Wed, 22 May 2013 21:35:15 -0000

My recent work has been on a very unusual topic: the social implications of brain emulations. To avoid the above mentioned biases, I thus try to make ordinary assumptions, and to use ordinary methods and sources.

I contend this is the wrong way to deal with bias, and it produces another bias: overcompensation. (See "The deeper solution to the mystery of moralism—Morality and free will are hazardous to your mental health" — http://tinyurl.com/9exlxlk .)
Added 5/23 A better reference for overcompensation bias is "Overcompensation for persuasion" — http://tinyurl.com/26w4k4j .

Re: Best To Mix Odd, Ordinary

IMASBA — Wed, 22 May 2013 20:13:54 -0000

Even a conservative watch is right twice a day...