Are You Really As Good at Something As You Think? | Robin Kramer | TED

翻译人员: Joyce Lim 校对人员: Yip Yan Yeung

I don't mean to brag,

我无意吹嘘，

but there are lots of things that I'm pretty average at.

但在很多方面我都相当平庸。

From playing table tennis,

从打乒乓球、

cooking risotto, finding countries on a map,

做意大利烩饭、查看地图，

just to name a few.

仅举几例。

Now, in our everyday lives,

在我们的日常生活中，

we're not typically assessed on our skills and abilities,

我们的能力和技能通常不被外界评估，

so we're forced to rely on our own judgments.

因此我们被迫依靠自己的判断。

I may think I'm pretty decent with Italian cuisine,

我可能认为自己在意大利料理方面不算太差，

but how accurate is my assessment?

但我的评估有多准确呢？

Now, what we’re talking about here is metacognition:

我想讲的是“元认知”：

our insight into our own thought processes.

我们对自己思维过程的见解。

If I have good metacognitive insight,

如果我具备良好的元认知洞察力，

then how good I think I am at a particular task

那么我预期的表现

should line up pretty well with how good I actually am.

应该与我的实际表现吻合。

Of course, in the real world this is often not the case.

当然，在现实世界中往往与之相反。

And indeed, we probably all know someone

实际上，我们可能都认识

who thinks they're great at navigating maps,

一个自认为很擅长看地图的人，

when in fact the reality is often the opposite.

而实际情况却恰恰相反。

Not to name any names, of course, but still.

不指名道姓，但事实就是如此。

Perhaps you think this applies to other people

也许你认为这只适用于其他人，

and that you, yourself, wouldn't make this sort of mistake.

而你自己绝对不会犯这样的错误。

So let's try a quick experiment.

让我们来进行一个快速的实验。

I want you to think about how you would rate yourself

我想让你评估一下

in terms of your driving ability.

自己的驾驶能力。

Would you rate yourself as below average, average

你认为自己低于平均水平、处于平均水平

or perhaps even above average?

又或是高于平均水平呢？

So most people rate themselves as above average,

大多数人都认为自己高于平均水平，

which, of course, is mathematically impossible,

而这从数学的角度是不可能的，

and something that we call the "better than average" effect.

我们称之为“优于平均效应”。

This is just one of a number of cognitive biases that we see

这只是人们在判断自身能力时

when people judge their own abilities.

所出现的众多认知偏差之一。

Today, I'm going to focus on a related bias,

今天，我将重点讨论一个相关的认知偏差，

the Dunning-Kruger effect.

即邓宁-克鲁格效应。

So back in 1999,

早在 1999 年，

two psychologists at Cornell University, Dunning and Kruger,

两位来自康奈尔大学的心理学家， 邓宁和克鲁格，

described the mistakes people make when estimating their own abilities.

描述了人们在自我能力评估时所犯的错误。

So if we take a sample of people and we divide them into four groups

如果我们随机选人然后

based on their scores on a test,

根据他们测试的分数将他们分为四组，

and order those groups from lowest to highest.

再从低到高对这些组进行排序。

If we plot those scores on a graph along with their self-estimates,

如果将这些分数与他们的自我评估数值 一起绘制在图表上，

so how well they thought they did on the test,

我们所看到的这个图形

this is the pattern that we see.

就是他们对这项测试的预期表现。

So the red line is a steep slope representing their actual scores.

红色的陡坡代表的是他们实际分数。

As it must be, since we ordered the groups

这是必然的，毕竟我们原先就是

based on their scores in the first place.

根据小组分数进行排序的。

Now what's interesting is the blue shallower line.

真正有趣的是相对平缓的蓝线。

This represents their self-estimates.

这代表他们的自我评估数值。

So, how good they thought they did on the test.

那么，他们认为自己在测试中的表现多好呢。

Now the Dunning-Kruger effect describes how the weakest performers

邓宁-克鲁格效应描述了表现最差的人如何

significantly overestimate their performance,

过于高估自己的表现，

shown here in the green oval.

就如这个绿色椭圆所示。

The explanation for this, according to Dunning and Kruger,

根据邓宁和克鲁格的说法，对此的解释

is that insight and ability rely on the same thing.

是洞察力和能力相辅相成。

So if I'm poor at a task,

如果我做事表现不佳，

I also lack the metacognitive insight to accurately assess my ability.

我也会因此缺乏元认知洞察力 而无法准确评估自身能力。

Now this pattern has been seen again and again

这种趋势在许多领域中

across a number of domains,

反复出现，

from driving skill to exam-taking, even chess-playing.

从驾驶技术到考试，甚至是下棋。

However, in recent years,

然而近年来，

a number of criticisms have been leveled at this approach,

这种方法受到了许多批评，

and we now have reason to believe

但我们现在有理由相信

that this pattern results is virtually unavoidable.

这种模式几乎是不可避免的。

One reason for this is the statistical effect,

其中一个原因是统计效应，

regression to the mean.

即回归均值。

Now this is something that comes about

当我们有两个相互关联

when we have two measures that are related but not perfectly so.

但并不一致的衡量标准时， 就会出现这种情况。

So imagine we have a sample of people

试想我们随机抽人

and we measure their heights and their weights.

并测量他们的身高和体重。

Now height and weight are related,

身高和体重是有关联的，

tall people are typically heavier,

高的人通常更重，

but the relationship is far from perfect.

但是这个说法并非无懈可击。

So unlike in the figure at the top here,

与上图红线所示不同，

the shortest people in red won't all be the lightest people.

在红色区块的人虽然最矮但未必最轻。

Some of them will be overweight or particularly muscular, for example.

例如，有些人会超重或肌肉特别发达。

Similarly at the top end,

在另一端，同样的，

the tallest people in blue won't all be the heaviest people.

蓝色区块的人虽然最高但未必最重。

Some of them will be underweight, and so on.

他们当中有些人会体重不足，依此类推。

Now as a result, on average,

因此，平均而言，

the shortest people will rank higher for weight than they do for height,

比起身高来说身材矮小的人 会在体重上排名更高，

and the tallest people will rank lower for weight than they do for height

而最高的人比起身高来说 体重排名更低

producing this blue line here

因而产生这条蓝线

and the crossover pattern you're now becoming familiar with.

和这个两线之间的交点。

Now, some people might put forward a spurious explanation

有些人可能会试图提出一些谬论，

for why short people are relatively overweight

说明为什么个子矮小的人相对超重，

or tall people relatively underweight,

或者个子高大的人相对过瘦，

when in fact no explanation is needed.

但这实际上并不需要任何解释。

Perhaps more compelling a reason to doubt the Dunning-Kruger effect

或许质疑邓宁-克鲁格效应更合理的原因

is that we can produce the same pattern in our data

在于我们可以用已经损坏的数据

when our data is entirely meaningless.

复刻出相同的图形。

So if we collect people's test scores

因此如果我们收集人们的测试分数

along with their self-estimates of those scores,

以及他们的自我评估数值，

but then we shuffle those self-estimates

随后将这些自我评估数值随机打乱

and then analyze as before,

再像之前一样进行分析，

then we still find that same pattern in the data.

我们仍然会发现同样的规律。

Of course, any effect that we can find

当然，在随机打乱的数据中

with shuffled or randomized data

发现到的的任何差异

is one that we should surely be suspicious of.

都值得被关注。

So, given these and other issues with the Dunning-Kruger approach,

考虑到邓宁-克鲁格效应 存在的种种问题，

I was saddened and disappointed

当我在发现到这个趋势逐渐

and, frankly, a little annoyed to discover

适用于我所在的领域，

that the same approach was now being applied

即面部配对时，

in my field of expertise,

我感到伤心和失望，

which is face-matching.

坦白说也有点烦躁。

Now, this is a task where we're showing two images of faces

上图显示的工作便是对比两张脸部图像

or an image and a live person,

或是对比一张图像和一个活人，

and we're asked to decide whether they show the same person

并被要求判断究竟来自同一个人

or two different people.

又或是来自不同的人。

Now, we've all stood in line at passport control,

我们则在边检柜台排队，

anxiously awaiting the passport officer's decision

焦急地等待官员决定

as to whether our ID photos look sufficiently like us or not.

我们的样子是否与证件照片吻合。

Indeed, I've included at the top here

事实上，我在上方列出了

some examples of ID images from my own life,

一些我自己生活中的证件照片为例，

just to illustrate some variability.

只是为了说明一些可变性。

Some proud moments in photographic history,

我相信你也会同意这些照片

I'm sure you'll agree.

展现着一些摄影史上令人骄傲的时刻。

And so what I'd like to do now

所以我现在想做的

is first see how well you might perform as passport officers.

就是先测试你们 作为护照官员能表现得多好。

So here are four pairs of images,

这里有四组照片，

some students’ ID images and some student photos.

一些学生证件照和学生自己的照片。

For each pair, I'd like you to decide whether it's a match,

我希望你们来判断每一组照片是匹配的，

so two images of the same person,

即两张照片均来自同一个人，

or a mismatch,

或是不匹配的，

two images of different people.

即两张照片均来自不同的人。

Some of you might be surprised to hear that the top two pairs are matches,

你们可能有人会惊讶于 上方两对其实是匹配的，

so images of the same people,

均来自同一个人，

and the bottom two pairs show mismatches,

以及下方两对是不匹配的，

so two different people.

即来自不同的人。

Now we know this task is particularly difficult

我们现在知道了分辨陌生的脸庞

when the images show identities that we're unfamiliar with.

这项任务其实特别困难。

This is because it's hard to take into account the changes

这是因为我们很难去辨别

that can happen to the face across time,

那些随着时间而产生的面部变化，

as well as over different situations,

又或是一些其他的变数，

so changes in facial expression or lighting, for instance.

例如面部表情又或是光线的变化。

We know this task is difficult for passport officers as well,

这项任务对护照官员来说也很困难，

and they also make mistakes.

并且他们也会犯错。

So this is why I thought it would be particularly interesting

这就是为什么我认为从这种安保角度下

to look at the relationship between insight and ability

去观察元认知洞察力 和真实能力之间的关系

in this important security context.

将会特别有趣。

So given the issues we’ve described already

考虑到我们在看待整体分数和自我评估时

with looking at overall scores and people’s self-estimates,

所描述的问题，

I instead decided to focus on individual decision making.

我决定专注于个人决策。

So over a series of experiments,

在接下来一系列实验中，

I asked people to look at pairs of images

我的要求是观察一对照片

and decide whether they were a match or a mismatch.

并决定它们是否匹配。

But I also asked people to provide a rating of confidence in each decision.

我同时要求参与者 对自己的自信程度进行评估。

Now a good metacognitive insight

良好的元认知洞察力

would be reflected in people being much more confident

将会反映在那些做出正确选择

in decisions that turned out to be correct

且对自己的答案极为确信的人身上，

and much less confident in decisions that turned out to be incorrect.

反之错误的决定 往往伴随的是自信心的缺失。

So let's have a look at how people did.

让我们一起来看看他们的表现。

Now I think this pattern is particularly fascinating,

我认为这个图形特别有趣，

but also fairly intuitive.

但也相当直观。

Let's start with the red line,

从红线开始看，

which represents people's confidence in their incorrect responses.

它代表的是做出错误决策时 人们的自信程度。

So as you can see,

正如你所看到的，

it doesn't matter how good people were at the test overall,

无论大家整体上做得多好

represented by the score on the X-axis at the bottom there;

（下方X轴上的分数作代表）；

people were approximately the same

答错的人对自己的信心程度

in terms of their confidence when they were incorrect.

大致相同。

Now what's interesting is the blue line,

现在有趣的是蓝线，

which represents confidence when people were correct in their decisions.

它代表的是做出正确决策时人们的自信程度。

As you can see, the best performers on the test

如你所见，比起答错的人，

were much more confident in their correct responses

那些表现优越的人

in comparison with their incorrect ones.

对自己所做出的决策要自信的多。

So shows good metacognitive insight.

这显示了良好的元认知洞察力。

The weakest performers, on the other hand,

另一方面，表现最差的人

were no different in their confidence

对自己的答案

for their correct and incorrect responses,

所具有的自信程度大抵一致，

shown here in the green circle.

正如绿圈所示。

And so they show poor metacognitive insight.

这说明他们不具备良好的元认知洞察力。

So what might be going on with these weak performers?

这些表现较差的人到底是怎么一回事呢？

Now it could be the case that they have some sense

有可能他们感知到了

they tend to perform poorly on tests in general,

自己在测试中表现得较差，

and so they're just less confident overall in their responses.

因此对自己的决策较为不自信。

However, I didn't find that pattern of lower confidence in my data,

然而我并没有在我的数据中 观察到自信心较低的规律，

at least with individual decision-making.

至少在个人决策方面是如此。

Instead, it's more likely that they were more confident in their correct responses

这更像是比起错误答案，

in comparison with their incorrect ones.

他们对自己的正确答案更有自信。

But this was simply unrelated to their accuracy on each trial

但这与他们在每项试验中的准确率完全无关，

because they had poor insight.

因为他们的洞察力很差。

So how does this all fit in with the Dunning-Kruger effect?

那么邓宁-克鲁格效应如何适用于这一切呢？

So Dunning and Kruger argued that the weakest performers

邓宁和克鲁格认为表现最差的人

show the least amount of insight

具备较弱的洞察力

and they overestimated their performance.

并往往高估自己的表现。

And that's implied that they had greater confidence.

这意味着他们更有信心。

Now, we didn't see that here in our data.

我们的数据没有体现这一点。

The weakest performers didn't seem to be overly confident.

表现最差的人并没有过度自信。

However, the Dunning-Kruger effect

但是，邓宁-克鲁格效应

also describes how insight depends on ability.

也描述了洞察力如何取决于个人能力。

And so the weakest performers showed the least amount of insight,

表现最差的人表现出的洞察力最弱，

overestimating their performance in their case.

因此往往高估了自己的表现。

As we've just seen,

正如我们刚才所见，

the weakest performers do seem to show the least amount of insight.

表现最差的群体体现出的洞察力似乎最弱。

Here, they couldn't differentiate between their correct and incorrect responses.

他们无法区分正确和错误的决策。

So insight does appear to depend on ability,

因此，洞察力似乎确实取决于能力，

but not in the way that Dunning and Kruger originally thought.

但并非像邓宁和克鲁格所想的那样。

So if there are two things I'd like you to remember from this talk

如果我想让你从这次演讲中记住两件事

and take home, think about afterwards,

并带回家思考，

they are: first, more broadly, science is always updating.

这两件事就是：首先，更广泛地说， 科学一直在更新。

Research comes along, new evidence that may contradict

随着不停的研究， 新的证据可能与以前的研究相矛盾

or even disprove previous work.

甚至是推翻以前的研究成果。

In this case, the Dunning-Kruger effect may well not be a thing,

在这种情况下，邓宁-克鲁格效应 很可能不适用。

despite the fact that it's so prevalent in popular culture.

尽管它在流行文化中非常普遍。

Second, insight depends on ability.

其次，洞察力取决于能力。

For the weakest performers,

对于表现最差的人来说，

there's no difference between their confidence

他们对于正确和错误回答

for correct and incorrect responses.

所具有的信心没有区别。

They have poor insight, they can't tell the difference.

他们的洞察力很差， 他们无法分辨出区别。

For strong performers,

对于表现优异的人来说，

when they're giving a correct answer,

当他们给出正确答案时，

they're much more confident.

他们会更加自信。

Of course, the inverse isn't always true.

当然，反过来却不一定是正确的。

Being more confident doesn't mean that you're right.

更自信并不意味着你是对的。

You might be wrong and simply have poor insight.

你有可能只是因为洞察力弱。

So in our everyday lives,

因此在我们的日常生活中，

you should think about who it is that you ask the opinions of.

你应该考虑谁更适合征求意见。

If someone is an expert in their field,

如果他是自己所在领域里的专家，

then if they're more confident, they're probably right,

那么如果他们更自信， 他们可能是对的，

but if they're unsure,

但是如果他们不确定，

this is also informative and tells us something useful.

这也告诉我们一些有用的资讯。

It's much more sensible to find someone that we know is knowledgeable in an area,

找一个在特定领域更专业的人要明智得多，

rather than someone who is simply confident in their opinion,

而不是找一个单纯对自己观点 充满信心的人，

because confidence is easily misplaced.

因为信心是很容易被错用的。

And finally,

最后，

for those of you who are still wondering how good my risotto actually is,

对于那些想知道我的意大利烩饭 究竟有多好吃的人，

that may have to wait for a future talk.

这可能得等到下一次讲座了。

Thank you.

谢谢。

(Applause)

（掌声）