Is there a reproducibility crisis in science? - Matt Anticole

翻译人员: Cissy Yun 校对人员: Lin Ye

In 2011, a team of physicists reported a startling discovery:

在2011年，一支物理学家团队发表了一项新发现：

中微子的传播速度大于光速

neutrinos traveled faster than the speed of light

by 60 billionths of a second

中微子走完经日内瓦到意大利探测站 全长730千米的路途

in their 730 kilometer trip from Geneva to a detector in Italy.

大约快了600亿分之一秒

Despite six months of double checking, the bizarre discovery refused to yield.

尽管用了六个月的时间再三确认，这一奇异的发现依旧如此

But rather than celebrating a physics revolution,

研究人员并没有庆祝这一物理大发现

the researchers published a cautious paper

他们反而在报纸上刊登了一篇警示性的文章

arguing for continued research in an effort to explain the observed anomaly.

主张继续研究，从而解释观察到的异常现象

In time, the error was tracked to a single incorrectly connected fiber optic cable.

与此同时，有人指出，造成此异象的原因是 研究者追踪了一条错误连接的纤维光缆

This example reminds us that real science is more than static textbooks.

这个例子告诉我们，真正的科学远不止不变的教科书

Instead, researchers around the world are continuously publishing

全世界的研究者仍持续不断地发表着

their latest discoveries

他们的最新研究成果

with each paper adding to the scientific conversation.

每篇新文章都参与到科学交流中

Published studies can motivate future research,

发表的研究成果可以促进未来的研究

inspire new products,

启发新产品的制造

and inform government policy.

并提示政府政策上的革新

So it's important that we have confidence in the published results.

所以我们对于结果发表的正确性拥有信心是很重要的

If their conclusions are wrong,

如果它们的结论是错误的

we risk time,

我们就冒着浪费时间

resources,

资源

and even our health in the pursuit of false leads.

甚至是健康的风险去跟随错误的引导

当有重大发现时

When findings are significant,

they are frequently double-checked by other researchers,

这些成果通常都会被其他研究者再三确认

或者重新分析数据

either by reanalyzing the data

or by redoing the entire experiment.

或者将整个实验重新再做一遍

For example, it took repeated investigation of the CERN data

比如，欧洲核子研究委员会对实验数据 进行了重复性调查

before the timing error was tracked down.

才追踪到计时误差

不幸的是，当下既没有资源也没有专业设备

Unfortunately, there are currently neither the resources nor professional incentives

to double check the more than 1 million scientific papers published annually.

去重新检验每年超过一百万的新发表成果

即使论文受到质疑，这样的结果也无法让人安心

Even when papers are challenged, the results are not reassuring.

最近一项研究检验了很多已发表的药学论文

Recent studies that examined dozens of published pharmaceutical papers

论文中的结果可被复制的不到百分之二十五

managed to replicate the results of less than 25% of them.

And similar results have been found in other scientific disciplines.

其他科学验证中也出现了相同的问题

论文中的结果无法被重现的原因有许多

There are a variety of sources for irreproducible results.

实验的原设计、执行、或是关于数据的分析 都有可能藏有错误

Errors could hide in their original design, execution, or analysis of the data.

还有未知因素

Unknown factors,

such as patients' undisclosed condition in a medical study,

譬如说在药物研究中，患者的未公开隐疾

就会使得实验的结果无法被复制

can produce results that are not repeatable in new test subjects.

And sometimes, the second research group can't reproduce the original results

而且有时，第二实验小组不能复制先前实验结果

simply because they don't know exactly what the original group did.

只是因为他们不知道先前的实验小组做了什么

然而，一些问题可能源于

However, some problems might stem from systematic decisions

人类研究科学的体系

in how we do science.

研究者

Researchers,

the institutions that employ them,

研究学院

and the scientific journals that publish findings

和发表研究成果的科学期刊

are expected to produce big results frequently.

都被要求迅速地产生新的重要成果

一篇重要的论文可以帮助事业的发展

Important papers can advance careers,

聚集媒体目光

generate media interest,

and secure essential funding,

和保证重要的研究资金

所以，研究者一般不会自己质疑自己的研究成果

so there's slim motivation for researchers to challenge their own exciting results.

而且，很少会有鼓励研究者

In addition, little incentive exists

to publish results unsupportive of the expected hypothesis.

发表不支持预计假设的研究成果 的情况

That results in a deluge of agreement between what was expected

这造成了如今泛滥的

与期望的一样的研究结果

and what was found.

In rare occasions, this can even lead to deliberate fabrication,

在极端情况下，这还可造成捏造成果的现象

such as in 2013, when a researcher spiked rabbit blood with human blood

在2013年，一位研究者把兔子的血液 加入人的血液中

to give false evidence that his HIV vaccine was working.

从而给出错误的证据表明他的艾滋疫苗是有用的

The publish or perish mindset

这种不是“发表“就是”消亡“的心态

can also compromise academic journals' traditional peer-review processes

还危及到了学术期刊传统的同行评审过程

这是一种安全检查

which are safety checks

where experts examine submitted papers for potential shortcomings.

专家会检测上交的论文中出现的潜在漏洞

如今的系统

The current system,

which might involve only one or two reviewers,

只会有一到两个评审

这可以被看做是无效的

can be woefully ineffective.

这被1998年的一场研究给证实

That was demonstrated in a 1998 study

where eight weaknesses were deliberately inserted into papers,

人们把八项漏洞放在了论文中

but only around 25% were caught upon review.

然而只有四分之一的漏洞在评审过程中被找出

很多科学正在努力提高其研究领域的可重复性

Many scientists are working toward improving reproducibility in their fields.

There's a push to make researchers raw data,

人们正在促进研究者的原始数据

experimental procedures,

实验过程

and analytical techniques more openly available

和分析方法更加开放、共有

in order to ease replication efforts.

这样可以减少重证结果的难度

The peer review process can also be strengthened

评审过程也可以被加强

to more efficiently weed out weak papers prior to publication.

在论文发表之前更有效地找出他们的漏洞

我们还可以通过发表更多与预计假设不符的研究成果

And we could temper the pressure to find big results

来缓解寻找重大研究成果的压力

by publishing more papers that fail to confirm the original hypothesis,

而结果和预计假设不符的情况 远远比现在被发表的结果所展示的多

an event that happens far more than current scientific literature suggests.

科学，永远有，也永远会遇到一些错误的开始

Science always has, and always will, encounter some false starts

as part of the collective acquisition of new knowledge.

这是获取新知识的过程的一部分

找到一种可以提高我们研究成果可重现性的方法

Finding ways to improve the reproducibility of our results

can help us weed out those false starts more effectively,

可以帮助我们更有效地揪出那些错误的开始

keeping us moving steadily toward exciting new discoveries.

让我们更顺利地去找寻新的发现