[灌水]Subconscious may bias sex

Subconscious may bias sex of babies
Helen Pilcher
Mothers who expect lengthy lives tend to produce sons.

Mothers who think they have longer to live are more likely to give birth to boys than girls, a survey of British women shows. The finding backs up the long-held theory that women may unwittingly be able to influence the sex of their unborn child.

Sarah Johns from the University of Kent asked 609 first-time mothers, who had already given birth, to guess when they thought they would die. By subtracting the mother's age, she then calculated the number of years each woman thought she had left to live. The results are reported in the Proceedings of the Royal Society1.

As the number of perceived years left rose, so too did the chance that they had had a son. Every extra year on the clock increased the odds of producing a male by 1%.

The finding backs up a 30-year-old hypothesis2 that suggests women can bias the sex of their unborn babies, to enhance the chances of their genes being passed on to future generations.

Boys need more looking after than girls, the theory says. So when food is scarce and resources are low, females preferentially give birth to girls because they are more likely to live through the hard times. But boys are able to produce more offspring, so when resources are plentiful, mothers should be more likely to give birth to boys, to maximise the number of potential grandchildren.

When women guess their age of death, they may unwittingly be assessing these factors, says Johns. "erceived life expectancy may be the observable product of an evolved, subconscious psychological mechanism that assesses environmental and physical conditions," she says.

Cost analysis

"It's not something I've ever come across in practice," says midwife Sue Jarman from South Norwood Medical Centre, London, but she says the idea that unborn sons are biologically more costly makes sense. Male babies are more likely to spontaneously abort than female babies, so women may need to be on top form to carry a son to term.

Once born, males may also require more parental investment than females. Women tend to breast-feed sons longer than daughters, says Jarman.

The research accords with other human and animal studies. Mhairi Gibson of University College London showed that rural Ethiopian women with low levels of nutrition are more likely to give birth to girls3.

But the current study is one of the first to show the effect in a society where there is no obvious lack of resources. "There may be a physiological mechanism that influences psychological factors," says Gibson.

Environmental and physical factors may affect testosterone levels, which may make women more likely to produce sons, says Johns.

我们的确有分歧。

你举的例子都可以研究，例如藏书和生男生女有关的话，数据上能反映出来。如果无关，应该结果接近50％。

在没有研究之前，你不能100％肯定两者无关。虽然无关的概率很大。

关于这个研究有没有意义，是否是：“无用的”，我认为很难说。（这一点也没有进一步讨论的必要）

文中的解释可以讨论。（这个我们是一致的，都保留怀疑，不必再说。）但文中的统计方法我认为基本合格，当然针对这个统计的解释可以争论。为什么我认为这个方法基本合格？因为1）样本足够多。609，你认为够不够？2）可重复性。问题足够简单，很容易验证。

我认为，这类研究的关键问题是：很容易失败，不是没有结论，而是结论说：接近50％，两者基本无关，功夫百费了。（btw，最成功的就是：如果a，那么一定b。）

“Every extra year on the clock increased the odds of producing a male by 1%. ”文中这个结果还是很成功的。（当然，我假设研究者的统计方法正确）




［此帖子已被 remind_me 在 2004-8-16 14:17:20 编辑过］

2)可重复性 是值得怀疑的！比较难以重复的！
样本量还可以，但很有可能是为了得到统计意义而使用如此样本量的数据，呵呵！

我认为实验可重复，结果当然有不同的可能。但如果数据真实，统计方法正确，结果也应该相差不大。退一步来说，即使有问题，别人很容易查出来。

数据造假，一旦发现，学术生涯应该结束了，统计方法有误，通过论文别人可以看出。我不认为这两者概率很大。

“我也试过这种社会调查的研究方法”，能否简单介绍一下？

呵呵，有趣。不过是否是 subconscius 起作用值得怀疑。

其实本身这个研究方法就值得商榷！
几乎可以“考问”出任何你想要的结论！
呵呵，我也试过这种社会调查的研究方法

“几乎可以“考问”出任何你想要的结论！”

“asked 609 first-time mothers, who had already given birth, to guess when they thought they would die. By subtracting the mother's age, she then calculated the number of years each woman thought she had left to live”

现象是“预计生存时间与生男生女有关”，结论是文中后面的解释。

我的意思是现象是客观的，而原因可能是复杂的。

请问楼上，你是对这个现象有疑问，还是对结论有疑问？

呵呵，你可以试一下母亲的某种爱好（music etc.)和生男生女的关系，可以试一下母亲藏书多少和生男生女的关系，母亲饮食和生男生女的关系，母亲对教育的态度和生男生女之间的关系等等等等等等
我是说你设计一定的问题，就可以得出几乎任何你需要的结论，当然解释起来又是另外一回事了。
我怀疑的这种研究方法，而不是现象或是结论。

btw：作者的解释勉强说的过去（真的很勉强）。nutrition还比较可信一些，毕竟通常意义上男孩比女孩的养育成本要高一些。

如果，实验方法本身有问题，我对结论当然有理由持怀疑态度。呵呵

ps：统计的时候会得出很多“无用的”结果，虽然你可以得出貌似的结论。