Visualizing
Data
Basic Types of ways to represent data…
about Tom Brady
[Link]
Side-by-Side
Bar Graph
[Link]
[Link]
s/photos/a.153360416772/101584528
16356773/?type=3&theater
Stacked
Bar Graph
[Link]
[Link]
[Link]
Line Graph
[Link]
Pie Chart
This must be old, as there have been 52 Superbowls and Tom Brady has won 7, which is 7/52 or about
13%
Table
[Link]
[Link]
res/tom-brady-retires-with-all-th
e-records/
[Link]
es/is-tom-brady-the-goatiest-of-a
ll-the-goats/
Other
[Link]
ments/1muf9z/tom_brady_passin
g_chart_the_first_two_weeks/
[Link]
/Patriots/comments/5su2
ol/graph_of_ne_patriots_i
n_the_brady_era/
[Link]
87557401070997508/photo/1
“Tom Brady” google trends
Interest over time
Numbers represent search interest relative to the highest point on the chart for the given region and time. A value of 100 is the peak
popularity for the term. A value of 50 means that the term is half as popular. A score of 0 means there was not enough data for this
term.
Animated Graphs
Other fun websites about Tom Brady
[Link]
nships-tampa-lightning
[Link]
riots-dynasty
[Link]
[Link]
qb-stats-chart-tom-bradys-greatness-in-a-mesmerizing-way
Confirmation Bias
+ CRAAP test
[Link]
aluate-research-sources-a2755126b6b2
Log Scales
Log Scales
Log scales show relative values instead of absolute ones.
Log scales don’t care about the fact that 101 minus 100 equals the same as 2
minus 1. Instead, they are concerned with percentages: between 100 and 101
is a 1% increase, while between 1 and 2 is a 100% increase.
So on a log scale, the distance between 100 and 101 is roughly 1% of the
distance between 1 and 2.
Log Scale Example
[Link]
Log Scales
What is true if we wanted to
compare how the number of newly
HIV infected increased between
Nigeria and Swaziland?
A. Both countries are increasing by
the about same total number
amount.
B. Both countries are increasing by
about the same percentage
increase.
C. Both of the statements above are
correct.
Linear Scale
We see that the number of deaths from HIV
increased much more in Nigeria than in
Swaziland. This means that we can tell from the We can tell from the graph that a very
graph that an individual in Nigeria who contracts small percentage of people who live in
HIV is significantly more likely to die than an Swaziland died from HIV between 1990
individual in Swaziland who contracts HIV. and 2000.
○ TRUE ○ FALSE ○ TRUE ○ FALSE
Linear vs Log
If the axis is multiplying by a constant amount, it is a log scale.
If the axis is adding by a constant amount, it is a linear scale.
Other Graphs
When are pie charts appropriate to use?
● To represent parts of a whole
○ If in percent - should add up to 100%.
● These parts do not change over time or represent at a certain instant.
… what if things change over time?
E.g. world energy consumption (consisting of oil, gas, coal, etc.)?
World Energy Consumption, 1965-2016
Population Pyramid
A population pyramid illustrates the The male and female populations
age and sex structure of a country's are broken down into 5-year age
population and may provide groups represented as horizontal
insights about political and social bars along the vertical axis, with the
stability, as well as economic youngest age groups at the bottom
development. The population is and the oldest at the top. The
distributed along the horizontal shape of the population pyramid
axis, with males shown on the left gradually evolves over time
and females on the right. based on fertility, mortality, and
international migration trends.
Q6
About how many females
are there in Russia age 19
and under?
A. ~ 4 million
B. ~ 28 million
C. ~3 million
D. ~ 14 million
Misleading
Graphs
Problems?
[Link]
[Link]/the-worst-
covid-19-misleading-gra
phs/
[Link]
misleading/
Common tactics used to mislead
● Truncated axis (omitting the baseline)
● Manipulating the y-axis
● Disproportional Sizing
● Cherry Picking Data
● Using the wrong graph type
● Going against convention
Also, graphs can just be confusing, but not intentionally misleading.