WEBVTT

00:00:00.107 --> 00:00:03.926
♪ [音乐] ♪

00:00:20.880 --> 00:00:22.077
- [Thomas Stratmann] 大家好！

00:00:22.077 --> 00:00:24.268
在接下来的一系列视频中

00:00:24.268 --> 00:00:26.858
我们将向你介绍一个
炫酷的新工具

00:00:26.858 --> 00:00:30.414
来帮助你理解数据

00:00:30.414 --> 00:00:31.981
那就是“线性回归”

00:00:32.885 --> 00:00:34.668
假设你有这么一种理论

00:00:34.668 --> 00:00:37.249
你发现外貌出众的人

00:00:37.249 --> 00:00:39.067
好像总能得到特殊的优待

00:00:39.642 --> 00:00:40.878
你在想

00:00:40.878 --> 00:00:43.798
“还有什么地方
也能看到这种现象呢？”

00:00:44.132 --> 00:00:45.637
对教师来说
这种现象也存在吗？

00:00:45.637 --> 00:00:48.259
有没有可能
外貌出众的老师

00:00:48.259 --> 00:00:50.010
也会得到特别优待呢？

00:00:50.350 --> 00:00:53.899
学生们会不会
对这些老师更好

00:00:53.899 --> 00:00:57.209
给他们打更高的学生评价分？

00:00:57.866 --> 00:01:00.467
如果确实如此
外貌对评价分的影响

00:01:00.467 --> 00:01:03.573
是很大还是很小呢？

00:01:04.159 --> 00:01:07.519
假设有位教师刚刚开始
到一所大学上班

00:01:07.519 --> 00:01:08.759
- [男人] 老兄，早啊

00:01:08.759 --> 00:01:11.810
- 仅从他的外貌
我们能对他的学生评价分

00:01:11.810 --> 00:01:13.371
做出怎样的预测？

00:01:13.940 --> 00:01:17.216
由于评价分能影响加薪

00:01:17.671 --> 00:01:21.709
如果这个理论属实
老师们可能会

00:01:21.709 --> 00:01:24.519
采取一些令人惊讶的手段
来提高他们的得分

00:01:24.519 --> 00:01:25.731
- [Lloyd Christmas] 耶！

00:01:25.731 --> 00:01:27.461
- 如果你想弄清楚

00:01:27.461 --> 00:01:30.801
更出众的外貌是否真的会
带来更高的评价分

00:01:31.441 --> 00:01:34.450
你会怎样检验这个假说呢？

00:01:34.956 --> 00:01:36.552
你可以收集数据

00:01:36.761 --> 00:01:40.025
首先，让学生从1到10

00:01:40.025 --> 00:01:42.076
给老师的外貌打分

00:01:42.076 --> 00:01:44.807
由此你可以得出
这位老师的颜值平均分

00:01:45.229 --> 00:01:48.552
然后你可以从25名学生处

00:01:48.552 --> 00:01:50.421
收集这位老师的教学评价分

00:01:50.421 --> 00:01:53.273
我们通过散点图

00:01:53.273 --> 00:01:54.738
来同时查看这两个变量

00:01:54.981 --> 00:01:57.419
我们用横轴表示外貌

00:01:57.852 --> 00:02:00.589
纵轴表示教学评价分

00:02:01.223 --> 00:02:04.903
例如，这一点代表着 Peate 教授

00:02:04.903 --> 00:02:06.423
- [Bib Fortuna] De wana wanga.

00:02:06.423 --> 00:02:08.811
- 他得到了3分的外貌分

00:02:08.811 --> 00:02:11.866
8.425的教学评价分

00:02:12.084 --> 00:02:14.958
这边的是 Helmchen 教授

00:02:14.958 --> 00:02:16.797
- [Ben Stiller, "Zoolander"]
帅到不像话！

00:02:16.797 --> 00:02:18.721
- 他的外貌得分非常高

00:02:18.721 --> 00:02:20.872
但评价分没那么高

00:02:21.101 --> 00:02:22.283
你能看出规律吗？

00:02:22.283 --> 00:02:25.533
当我们沿x轴从左向右移动

00:02:25.533 --> 00:02:27.963
从难看向好看移动

00:02:27.963 --> 00:02:31.186
评价分呈现出上升趋势

00:02:31.870 --> 00:02:35.174
对了，我们在这个系列视频中
使用的数据

00:02:35.174 --> 00:02:38.923
不是编造出来的
而是来自于

00:02:38.923 --> 00:02:40.897
在德克萨斯大学做过的
真实研究

00:02:41.337 --> 00:02:46.023
另外你可能不知道
“pulchritude”只不过是

00:02:46.023 --> 00:02:47.880
“颜值”的另一种
比较高端、学术的说法

00:02:48.405 --> 00:02:51.474
有些时候

00:02:51.474 --> 00:02:55.594
用散点图很难判断出
两个变量之间的确切关系

00:02:55.594 --> 00:02:59.104
尤其是随着我们
从左向右移动

00:02:59.104 --> 00:03:01.318
数值的波动很大的时候

00:03:02.000 --> 00:03:04.908
处理这种波动的一种方法是

00:03:04.908 --> 00:03:08.144
画一条直线
穿过这团数据

00:03:08.144 --> 00:03:10.775
让这条直线

00:03:10.775 --> 00:03:12.613
尽可能贴切地概括这些数据

00:03:13.295 --> 00:03:17.181


00:03:17.669 --> 00:03:20.888


00:03:20.888 --> 00:03:24.278


00:03:24.278 --> 00:03:26.456


00:03:27.087 --> 00:03:29.536


00:03:30.067 --> 00:03:32.596


00:03:32.596 --> 00:03:35.358


00:03:36.107 --> 00:03:39.827


00:03:40.794 --> 00:03:43.807


00:03:43.807 --> 00:03:45.587


00:03:46.070 --> 00:03:50.237


00:03:50.237 --> 00:03:53.026


00:03:53.544 --> 00:03:55.907


00:03:55.907 --> 00:03:59.469


00:03:59.768 --> 00:04:03.939


00:04:04.377 --> 00:04:07.420


00:04:07.857 --> 00:04:10.764


00:04:11.158 --> 00:04:13.903


00:04:14.389 --> 00:04:17.304


00:04:17.770 --> 00:04:21.262


00:04:21.579 --> 00:04:25.569


00:04:25.569 --> 00:04:28.429


00:04:28.429 --> 00:04:30.229


00:04:30.229 --> 00:04:31.297


00:04:31.297 --> 00:04:34.109


00:04:34.683 --> 00:04:36.749


00:04:37.019 --> 00:04:38.749


00:04:39.233 --> 00:04:41.665


00:04:41.665 --> 00:04:43.515


00:04:44.844 --> 00:04:47.890


00:04:47.890 --> 00:04:49.770


00:04:49.770 --> 00:04:52.039


00:04:52.838 --> 00:04:55.439


00:04:55.439 --> 00:04:58.340


00:04:58.833 --> 00:05:00.430


00:05:00.430 --> 00:05:03.639


00:05:03.639 --> 00:05:06.900


00:05:07.344 --> 00:05:09.965


00:05:09.965 --> 00:05:12.456


00:05:12.456 --> 00:05:15.645


00:05:16.052 --> 00:05:18.956


00:05:19.228 --> 00:05:22.972


00:05:23.660 --> 00:05:27.668


00:05:28.080 --> 00:05:32.095


00:05:32.095 --> 00:05:34.335


00:05:34.335 --> 00:05:37.383


00:05:37.861 --> 00:05:40.388


00:05:40.388 --> 00:05:43.537


00:05:43.537 --> 00:05:45.848


00:05:46.346 --> 00:05:49.807


00:05:50.289 --> 00:05:53.620


00:05:53.620 --> 00:05:54.900


00:05:54.900 --> 00:05:58.166


00:05:58.922 --> 00:06:02.069


00:06:02.069 --> 00:06:05.781


00:06:06.626 --> 00:06:09.575


00:06:09.846 --> 00:06:14.577


00:06:14.577 --> 00:06:18.994


00:06:19.408 --> 00:06:21.758


00:06:21.758 --> 00:06:24.477


00:06:24.477 --> 00:06:26.219


00:06:26.219 --> 00:06:28.368


00:06:28.368 --> 00:06:30.737


00:06:31.762 --> 00:06:35.509


00:06:35.509 --> 00:06:39.070


00:06:41.169 --> 00:06:42.445


00:06:42.445 --> 00:06:45.247


00:06:45.568 --> 00:06:47.139


00:06:47.139 --> 00:06:48.700


00:06:48.700 --> 00:06:50.404


00:06:50.865 --> 00:06:53.976


00:06:54.313 --> 00:06:55.364


00:06:55.598 --> 00:06:58.325


00:06:58.325 --> 00:07:01.642


00:07:01.892 --> 00:07:04.406