Visualizing Relationships - Data Points: Visualization That Means Something - page 226

In case you're wondering how to do this, it's just a bunch of bar charts,
so you could change the mfrow parameter like you did before and use the
plot() or polygon() functions. I did this using the Column Graph tool in
Illustrator though, because I happened to already have it open.
I learned a couple of things after posting this graphic. The first and most
important is that aggregates and distributions are not something every-
body deals with on a regular basis, so you need to do your best to explain
the data and take extra care in telling the story. The second is that people
love their movies, and when you say that their favorite movies of all time
are horrible, they sort of take it to heart.
Wrapping up
Looking for relationships in your data can be challenging at times and
requires more critical thinking than blindly graphing numbers, but it can
also be the most rewarding and informative. It's how your data, or rather,
how the things that your data represents relate and interact with each
other that's interesting—that's what makes for the best stories.
This chapter covered how to look for correlations between multiple vari-
ables, but explained relationships in a more general sense, too. Look at
how everything relates to each other as a whole through distributions.
Look within the distributions for outliers or patterns, and then think about
the context of what you see. Then if you find something interesting, ask
why. Think about the context of the data and possible explanations.
This is the best part about playing with data because you get to explore
what the data is about and maybe dig up something interesting. Then when
you dig enough, you can explain to readers what you find. Remember, not
everyone speaks the language of numbers, so keep it at a human level for
the general audience. Don't be afraid to turn it up to nerd level though, if
you have the right audience.