How to spot a misleading graph - Lea Gaslowitz

翻译人员: HE TIANKAN 校对人员: Lipeng Chen

A toothpaste brand claims their product will destroy more plaque

一个牙膏品牌声称他们的产品 可以比以往任何的产品

than any product ever made.

除掉更多的牙菌斑。

A politician tells you their plan will create the most jobs.

一位政客告诉人们 他们的计划会制造最多的就业机会。

We're so used to hearing these kinds of exaggerations

我们已经太习惯在广告和政治宣传中

in advertising and politics

听到这些夸大其词了，

that we might not even bat an eye.

以至于有时我们根本不把这些东西放在眼里。

But what about when the claim is accompanied by a graph?

但是假如这些说法同时伴随着一个图表呢？

Afterall, a graph isn't an opinion.

毕竟，图表并不是主观意见，

It represents cold, hard numbers, and who can argue with those?

它呈现的只是冰冷确切的数字， 谁又能来质疑这些冷冰冰的数字呢？

Yet, as it turns out, there are plenty of ways graphs can mislead

然而，事实是，图表可以通过很多方法误导人们

and outright manipulate.

并完全地操纵人们的想法。

Here are some things to look out for.

下面便是一些图表误导人们的方式。

In this 1992 ad, Chevy claimed to make the most reliable trucks in America

在这则1992年的广告中， 雪弗兰利用这张图表声称

using this graph.

他们制造了全美最耐用的卡车。

Not only does it show that 98% of all Chevy trucks sold in the last ten years

这张图表不仅显示雪弗兰过去的十年间卖出的卡车

are still on the road,

有98%目前仍在使用中，

but it looks like they're twice as dependable as Toyota trucks.

图表还暗示他们的卡车比丰田要耐用一倍。

That is, until you take a closer look at the numbers on the left

事实看上去就是如此， 直到你仔细观察纵坐标上对应的数值，

and see that the figure for Toyota is about 96.5%.

你才会发现丰田（十年间卖出的）卡车的使用率为96.5%左右。

The scale only goes between 95 and 100%.

该图表的问题在于 纵坐标的范围仅仅是95到100.

If it went from 0 to 100, it would look like this.

如果范围是从0到100， 图表会是这个样子。

This is one of the most common ways graphs misrepresent data,

这便是图表误传数据的最普遍方法之一，

by distorting the scale.

也就是扭曲某一坐标的尺度。

Zooming in on a small portion of the y-axis

将y轴的某一小部分放大

exaggerates a barely detectable difference between the things being compared.

可使两个被比较的事物之间 难以察觉的差距被夸张放大。

And it's especially misleading with bar graphs

这个方法对于柱状图来说尤其起效。

since we assume the difference in the size of the bars

因为我们总是假设柱状图的长度与数值

是成比例对应的。

is proportional to the values.

But the scale can also be distorted along the x-axis,

另一方面，x轴的间距也是可以被扭曲的，

usually in line graphs showing something changing over time.

这种扭曲通常发生在 呈现某事物随着时间改变的线形图上。

This chart showing the rise in American unemployment from 2008 to 2010

这张图标呈现的是美国2008年 到2010年的失业情况，

manipulates the x-axis in two ways.

其使用了两种方法操纵x轴。

First of all, the scale is inconsistent,

首先，x轴的间距是不一致的。

compressing the 15-month span after March 2009

2009年3月之后的15个月的跨度被压缩

to look shorter than the preceding six months.

使其看起来比之前的6个月还短。

Using more consistent data points gives a different picture

如果使用一致的数据点， 我们将会看到一张截然不同的图表，

with job losses tapering off by the end of 2009.

其中的失业情况在2009年年底之后逐渐减弱。

And if you wonder why they were increasing in the first place,

如果你对图表前一部分的 失业情况为何会加重感到不解，

the timeline starts immediately after the U.S.'s biggest financial collapse

其原因是该图表中时间线的起点正是金融海啸之后

since the Great Depression.

美国开始财政崩溃之后的时刻。

These techniques are known as cherry picking.

这种技巧被称为“计划性选择”。

A time range can be carefully chosen to exclude the impact of a major event

也就是通过别有用心地选择一个时间段

right outside it.

来排除该时间段之外发生的某一事件的影响。

And picking specific data points can hide important changes in between.

而选择某些特定的数据点 可以掩盖该时间段内的重要变化。

Even when there's nothing wrong with the graph itself,

即使图表本身没有任何错误，

leaving out relevant data can give a misleading impression.

省略某些相关的数据点 也会让人留下错误的印象。

This chart of how many people watch the Super Bowl each year

统计每年观看超级碗观众人数的表格

makes it look like the event's popularity is exploding.

让人们以为超级碗的人气火爆。

But it's not accounting for population growth.

但事实上它统计的并不是观众人数的成长。

The ratings have actually held steady

事实上，超级碗的收视率是保持稳定的，

because while the number of football fans has increased,

因为虽然球迷的数量在增加，

their share of overall viewership has not.

但是每名观众的收视占有率却并没有增加。

Finally, a graph can't tell you much

最后，在不知道图表呈现的重点的情况下，

if you don't know the full significance of what's being presented.

人们很难从其中获得有用的信息。

Both of the following graphs use the same ocean temperature data

以下的两幅图利用了国家环境信息中心提供的

from the National Centers for Environmental Information.

同一组海洋温度的数据。

So why do they seem to give opposite impressions?

然而为什么它们却看起来完全相反呢？

The first graph plots the average annual ocean temperature

第一幅图描绘了1880年到2016年的

from 1880 to 2016,

年平均海洋温度。

making the change look insignificant.

虽然温度的变化看起来并不明显，

But in fact, a rise of even half a degree Celsius

但是，即使是半摄氏度的温度上升

can cause massive ecological disruption.

也可能导致严重的生态问题。

This is why the second graph,

这也就是为什么

which show the average temperature variation each year,

呈现了年平均温度波动的第二张图表

is far more significant.

重要性远大于第一张。

When they're used well, graphs can help us intuitively grasp complex data.

如果使用得当，图表可以帮助我们 更直观地了解复杂的数据。

But as visual software has enabled more usage of graphs throughout all media,

但是可视化软件一方面大大 增加了图表在各种媒体上的应用，

it's also made them easier to use in a careless or dishonest way.

另一方面粗心大意和刻意误导也变得更加频繁。

So the next time you see a graph, don't be swayed by the lines and curves.

因此，下一次在遇到图表时， 不要被直线和曲线的走向误导。

Look at the labels,

仔细看一看单位、

the numbers,

数值、

the scale,

间距

and the context,

以及该图表的背景，

and ask what story the picture is trying to tell.

问问自己这张图标到底要传递什么信息。