[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:07.81,0:00:10.84,Default,,0000,0000,0000,,A toothpaste brand claims\Ntheir product will destroy more plaque
Dialogue: 0,0:00:10.84,0:00:12.91,Default,,0000,0000,0000,,than any product ever made.
Dialogue: 0,0:00:12.91,0:00:16.41,Default,,0000,0000,0000,,A politician tells you their plan\Nwill create the most jobs.
Dialogue: 0,0:00:16.41,0:00:18.95,Default,,0000,0000,0000,,We're so used to hearing these\Nkinds of exaggerations
Dialogue: 0,0:00:18.95,0:00:20.85,Default,,0000,0000,0000,,in advertising and politics
Dialogue: 0,0:00:20.85,0:00:23.13,Default,,0000,0000,0000,,that we might not even bat an eye.
Dialogue: 0,0:00:23.13,0:00:26.11,Default,,0000,0000,0000,,But what about when the claim\Nis accompanied by a graph?
Dialogue: 0,0:00:26.11,0:00:28.47,Default,,0000,0000,0000,,Afterall, a graph isn't an opinion.
Dialogue: 0,0:00:28.47,0:00:32.61,Default,,0000,0000,0000,,It represents cold, hard numbers,\Nand who can argue with those?
Dialogue: 0,0:00:32.61,0:00:36.40,Default,,0000,0000,0000,,Yet, as it turns out, there are plenty\Nof ways graphs can mislead
Dialogue: 0,0:00:36.40,0:00:38.19,Default,,0000,0000,0000,,and outright manipulate.
Dialogue: 0,0:00:38.19,0:00:40.74,Default,,0000,0000,0000,,Here are some things to look out for.
Dialogue: 0,0:00:40.74,0:00:45.76,Default,,0000,0000,0000,,In this 1992 ad, Chevy claimed to make\Nthe most reliable trucks in America
Dialogue: 0,0:00:45.76,0:00:47.51,Default,,0000,0000,0000,,using this graph.
Dialogue: 0,0:00:47.51,0:00:51.96,Default,,0000,0000,0000,,Not only does it show that 98% of all\NChevy trucks sold in the last ten years
Dialogue: 0,0:00:51.96,0:00:53.59,Default,,0000,0000,0000,,are still on the road,
Dialogue: 0,0:00:53.59,0:00:57.34,Default,,0000,0000,0000,,but it looks like they're twice\Nas dependable as Toyota trucks.
Dialogue: 0,0:00:57.34,0:01:00.63,Default,,0000,0000,0000,,That is, until you take a closer look\Nat the numbers on the left
Dialogue: 0,0:01:00.63,0:01:05.47,Default,,0000,0000,0000,,and see that the figure for Toyota\Nis about 96.5%.
Dialogue: 0,0:01:05.47,0:01:09.31,Default,,0000,0000,0000,,The scale only goes between 95 and 100%.
Dialogue: 0,0:01:09.31,0:01:12.96,Default,,0000,0000,0000,,If it went from 0 to 100,\Nit would look like this.
Dialogue: 0,0:01:12.96,0:01:16.24,Default,,0000,0000,0000,,This is one of the most common\Nways graphs misrepresent data,
Dialogue: 0,0:01:16.24,0:01:18.33,Default,,0000,0000,0000,,by distorting the scale.
Dialogue: 0,0:01:18.33,0:01:20.80,Default,,0000,0000,0000,,Zooming in on a small portion\Nof the y-axis
Dialogue: 0,0:01:20.80,0:01:25.70,Default,,0000,0000,0000,,exaggerates a barely detectable difference\Nbetween the things being compared.
Dialogue: 0,0:01:25.70,0:01:27.97,Default,,0000,0000,0000,,And it's especially misleading \Nwith bar graphs
Dialogue: 0,0:01:27.97,0:01:31.02,Default,,0000,0000,0000,,since we assume the difference\Nin the size of the bars
Dialogue: 0,0:01:31.02,0:01:33.23,Default,,0000,0000,0000,,is proportional to the values.
Dialogue: 0,0:01:33.23,0:01:36.12,Default,,0000,0000,0000,,But the scale can also be distorted\Nalong the x-axis,
Dialogue: 0,0:01:36.12,0:01:40.41,Default,,0000,0000,0000,,usually in line graphs\Nshowing something changing over time.
Dialogue: 0,0:01:40.41,0:01:44.75,Default,,0000,0000,0000,,This chart showing the rise\Nin American unemployment from 2008 to 2010
Dialogue: 0,0:01:44.75,0:01:47.100,Default,,0000,0000,0000,,manipulates the x-axis in two ways.
Dialogue: 0,0:01:47.100,0:01:50.40,Default,,0000,0000,0000,,First of all, the scale is inconsistent,
Dialogue: 0,0:01:50.40,0:01:53.42,Default,,0000,0000,0000,,compressing the 15-month span\Nafter March 2009
Dialogue: 0,0:01:53.42,0:01:56.76,Default,,0000,0000,0000,,to look shorter than \Nthe preceding six months.
Dialogue: 0,0:01:56.76,0:02:00.11,Default,,0000,0000,0000,,Using more consistent data points\Ngives a different picture
Dialogue: 0,0:02:00.11,0:02:03.70,Default,,0000,0000,0000,,with job losses tapering off\Nby the end of 2009.
Dialogue: 0,0:02:03.70,0:02:06.68,Default,,0000,0000,0000,,And if you wonder why\Nthey were increasing in the first place,
Dialogue: 0,0:02:06.68,0:02:10.62,Default,,0000,0000,0000,,the timeline starts immediately after\Nthe U.S.'s biggest financial collapse
Dialogue: 0,0:02:10.62,0:02:12.63,Default,,0000,0000,0000,,since the Great Depression.
Dialogue: 0,0:02:12.63,0:02:15.22,Default,,0000,0000,0000,,These techniques are known as\Ncherry picking.
Dialogue: 0,0:02:15.22,0:02:18.87,Default,,0000,0000,0000,,A time range can be carefully chosen\Nto exclude the impact of a major event
Dialogue: 0,0:02:18.87,0:02:20.65,Default,,0000,0000,0000,,right outside it.
Dialogue: 0,0:02:20.65,0:02:24.76,Default,,0000,0000,0000,,And picking specific data points\Ncan hide important changes in between.
Dialogue: 0,0:02:24.76,0:02:27.36,Default,,0000,0000,0000,,Even when there's nothing wrong\Nwith the graph itself,
Dialogue: 0,0:02:27.36,0:02:30.94,Default,,0000,0000,0000,,leaving out relevant data can give\Na misleading impression.
Dialogue: 0,0:02:30.94,0:02:33.100,Default,,0000,0000,0000,,This chart of how many people watch\Nthe Super Bowl each year
Dialogue: 0,0:02:33.100,0:02:37.63,Default,,0000,0000,0000,,makes it look like the event's\Npopularity is exploding.
Dialogue: 0,0:02:37.63,0:02:40.20,Default,,0000,0000,0000,,But it's not accounting \Nfor population growth.
Dialogue: 0,0:02:40.20,0:02:41.97,Default,,0000,0000,0000,,The ratings have actually held steady
Dialogue: 0,0:02:41.97,0:02:45.11,Default,,0000,0000,0000,,because while the number \Nof football fans has increased,
Dialogue: 0,0:02:45.11,0:02:47.96,Default,,0000,0000,0000,,their share of overall viewership has not.
Dialogue: 0,0:02:47.96,0:02:49.89,Default,,0000,0000,0000,,Finally, a graph can't tell you much
Dialogue: 0,0:02:49.89,0:02:53.32,Default,,0000,0000,0000,,if you don't know the full significance \Nof what's being presented.
Dialogue: 0,0:02:53.32,0:02:56.46,Default,,0000,0000,0000,,Both of the following graphs\Nuse the same ocean temperature data
Dialogue: 0,0:02:56.46,0:02:59.72,Default,,0000,0000,0000,,from the National Centers \Nfor Environmental Information.
Dialogue: 0,0:02:59.72,0:03:02.49,Default,,0000,0000,0000,,So why do they seem to give\Nopposite impressions?
Dialogue: 0,0:03:02.49,0:03:05.28,Default,,0000,0000,0000,,The first graph plots the average\Nannual ocean temperature
Dialogue: 0,0:03:05.28,0:03:07.99,Default,,0000,0000,0000,,from 1880 to 2016,
Dialogue: 0,0:03:07.99,0:03:10.15,Default,,0000,0000,0000,,making the change look insignificant.
Dialogue: 0,0:03:10.15,0:03:12.88,Default,,0000,0000,0000,,But in fact, a rise of even\Nhalf a degree Celsius
Dialogue: 0,0:03:12.88,0:03:15.80,Default,,0000,0000,0000,,can cause massive ecological disruption.
Dialogue: 0,0:03:15.80,0:03:17.22,Default,,0000,0000,0000,,This is why the second graph,
Dialogue: 0,0:03:17.22,0:03:19.86,Default,,0000,0000,0000,,which show the average temperature\Nvariation each year,
Dialogue: 0,0:03:19.86,0:03:22.39,Default,,0000,0000,0000,,is far more significant.
Dialogue: 0,0:03:22.39,0:03:27.38,Default,,0000,0000,0000,,When they're used well, graphs can\Nhelp us intuitively grasp complex data.
Dialogue: 0,0:03:27.38,0:03:31.18,Default,,0000,0000,0000,,But as visual software has enabled\Nmore usage of graphs throughout all media,
Dialogue: 0,0:03:31.18,0:03:35.90,Default,,0000,0000,0000,,it's also made them easier to use\Nin a careless or dishonest way.
Dialogue: 0,0:03:35.90,0:03:39.56,Default,,0000,0000,0000,,So the next time you see a graph,\Ndon't be swayed by the lines and curves.
Dialogue: 0,0:03:39.56,0:03:40.88,Default,,0000,0000,0000,,Look at the labels,
Dialogue: 0,0:03:40.88,0:03:42.13,Default,,0000,0000,0000,,the numbers,
Dialogue: 0,0:03:42.13,0:03:43.05,Default,,0000,0000,0000,,the scale,
Dialogue: 0,0:03:43.05,0:03:44.36,Default,,0000,0000,0000,,and the context,
Dialogue: 0,0:03:44.36,0:03:46.78,Default,,0000,0000,0000,,and ask what story the picture\Nis trying to tell.