Heads up! To view this whole video, sign in with your Courses account or enroll in your free 7-day trial. Sign In Enroll
Preview
Start a free Courses trial
to watch this video
Practitioners of data visualization carry a lot of responsibility. While good charts can create joy, provide insight, or win an argument, bad charts have the potential to confuse, misinform, or mislead. And in the most extreme cases, a faulty visualization could incite panic.
Related Discussions
Have questions about this video? Start a discussion with the community and Treehouse staff.
Sign upRelated Discussions
Have questions about this video? Start a discussion with the community and Treehouse staff.
Sign up
We've come a long way
from our first video,
0:01
where we defined the term
data visualization and
0:03
checked out some charts from an article
called The Case for Dennis Rodman.
0:06
Let's take a moment to test our knowledge.
0:11
Here's a chart called Visualizing
the Dominance of Dennis Rodman.
0:14
This isn't part of Ben Morris's article
on Dennis Rodman we studied earlier.
0:18
It's just something I found online.
0:23
It seems to me there's more than one
problem with this visualization.
0:26
Take a moment and see if you can
identify any issues with this chart.
0:30
So what did you find?
0:38
My first concern is that
the heading doesn't match the data.
0:39
The heading reads
The Best Rebounder Of All Time, but
0:43
the statistics shown compare only the best
rebounders of the 1997/98 NBA season.
0:47
In addition, let's compare
Dennis Rodman to David Robinson,
0:54
the lowest-ranked rebounder on the list.
0:58
This bar chart exaggerates
Rodman's dominance by starting
1:01
the x-axis at 9 rebounds rather than 0.
1:06
Finally, the measure used by this chart,
rebounds per inch, is a bit strange.
1:10
The fact that Dennis Rodman collected
more rebounds than some significantly
1:16
taller players is interesting.
1:21
But it's not really relevant to
the question of who was the best rebounder
1:23
of all time.
1:27
If you were able to answer that question,
1:29
it means you've got a good grasp on the
basics of how to create an accurate chart.
1:32
And that's important because practitioners
of data visualization carry a lot of
1:37
responsibility.
1:41
While good charts can create joy,
provide insight, or win an argument,
1:43
bad charts have the potential to confuse,
to misinform, or mislead.
1:48
And in the most extreme cases,
a faulty visualization could incite panic.
1:54
In an article called
Mapping Coronavirus Responsibly,
2:00
author Kenneth Field evaluates
a choropleth map of coronavirus
2:05
cases in China as of February 24th, 2020.
2:10
The flaws of this map are not nearly as
obvious as the previous Dennis Rodman
2:14
graphic, and that's part of the problem.
2:18
A graphic like this looks reasonable and
is easily shared as is.
2:21
However, the map diagrams
raw numbers of cases.
2:26
So while we learn that Hubei Province
has over 65,000 cases,
2:30
it's hard to say whether that's a lot,
2:34
since the map doesn't indicate how
many people live in each province.
2:36
In addition, the map uses dark red
to indicate provinces with more than
2:41
1,000 coronavirus cases.
2:46
So we might assume those five provinces
are experiencing similar numbers.
2:48
However, when we check out
the data as a bar chart,
2:54
we learn that no other
province has even 2,000 cases.
2:57
Hubei Province is a gigantic outlier,
3:02
a fact that's not communicated
by the choropleth map.
3:05
Field redesigns the map to reflect the
number of cases per 100,000 people, using
3:11
a dark blue for Hubei Province and a pale
blue for the rest of China's provinces.
3:16
Here we get a much more accurate sense
of the spread of coronavirus as of late
3:23
February 2020 and how it impacts the
population of each of China's provinces.
3:28
We see that every province has less
than 3 cases per 100,000 people,
3:34
except Hubei Province,
which has 111 cases per 100,000.
3:39
If you're wondering why
the change from red to blue,
3:45
the color red is associated with danger
and death in many Western cultures.
3:49
Seeing a map of coronavirus cases in
which several provinces are colored
3:54
blood red could give users an exaggerated
sense of the current danger.
3:58
Not that the coronavirus in February
2020 wasn't dangerous, of course.
4:04
But visualizing data ethically means
portraying numbers accurately and
4:09
never understating or
overstating their importance.
4:13
In addition,
4:18
the term Red China has derogatory
connotations in the United States.
4:19
And hate crimes against
Asian Americans and
4:24
Pacific Islanders have skyrocketed since
the arrival of the coronavirus pandemic.
4:27
While it's perfectly okay to make
a chart documenting accurate coronavirus
4:34
numbers in China,
4:38
it's important to consider the cultural
associations of the colors you're using.
4:40
As well as the consequences if
your numbers look exaggerated or
4:46
your information is misunderstood.
4:49
While most graphs you make won't deal with
topics as heavy as the COVID-19 pandemic,
4:52
I've included an article
by Amanda Makulec called
4:59
Ten Considerations Before
You Create Another Chart
5:02
About COVID-19 in the
teachers' notes.
5:06
As Makulec says, whenever you're making
a chart that impacts public health,
5:10
it's critical to consider whether your
chart serves an actual information need in
5:14
the public domain.
5:19
Does it add value to the public and
uncover new information?
5:21
Do you fully understand the context
of the data you're working with?
5:26
If not, don't share it.
5:30
You made it.
5:34
We've covered a lot of information here.
5:35
But keep practicing what you've learned,
and you'll be well on your way to creating
5:38
informative, accurate, accessible,
and ethical data visualizations.
5:43
You need to sign up for Treehouse in order to download course files.
Sign upYou need to sign up for Treehouse in order to set up Workspace
Sign up