Essential Do's and Don'ts for Data Visualization
Key Points
- Data visualization, when used thoughtfully, helps turn abundant data into understandable insights, but it isn’t a universal solution and must be matched to the data type and audience.
- Keep visualizations simple and digestible—avoid unnecessary complexity, excess colors, shapes, or variables—to make it easy for viewers to draw the intended conclusions.
- Know your audience and provide appropriate context; tailor the level of detail and sophistication so both data‑savvy users and novices can grasp the message without feeling left out.
- Use clear, descriptive titles, axis labels, and legends, and never rely on ambiguity or visual tricks that could mislead the audience.
Sections
- Do's and Don'ts of Visualization - The speaker stresses that abundant data calls for visual representation, but effective visualization hinges on simple, audience‑focused designs and avoiding needless complexity.
- Clarity and Honesty in Charts - The speaker stresses the need for explicit titles, axes, legends, and footnotes—and appropriate scaling—to avoid misleading viewers, illustrated by how adjusting a basketball players' points‑per‑game y‑axis can distort perceived performance.
Full Transcript
# Essential Do's and Don'ts for Data Visualization **Source:** [https://www.youtube.com/watch?v=QGDhKyZiPAo](https://www.youtube.com/watch?v=QGDhKyZiPAo) **Duration:** 00:05:26 ## Summary - Data visualization, when used thoughtfully, helps turn abundant data into understandable insights, but it isn’t a universal solution and must be matched to the data type and audience. - Keep visualizations simple and digestible—avoid unnecessary complexity, excess colors, shapes, or variables—to make it easy for viewers to draw the intended conclusions. - Know your audience and provide appropriate context; tailor the level of detail and sophistication so both data‑savvy users and novices can grasp the message without feeling left out. - Use clear, descriptive titles, axis labels, and legends, and never rely on ambiguity or visual tricks that could mislead the audience. ## Sections - [00:00:00](https://www.youtube.com/watch?v=QGDhKyZiPAo&t=0s) **Do's and Don'ts of Visualization** - The speaker stresses that abundant data calls for visual representation, but effective visualization hinges on simple, audience‑focused designs and avoiding needless complexity. - [00:03:03](https://www.youtube.com/watch?v=QGDhKyZiPAo&t=183s) **Clarity and Honesty in Charts** - The speaker stresses the need for explicit titles, axes, legends, and footnotes—and appropriate scaling—to avoid misleading viewers, illustrated by how adjusting a basketball players' points‑per‑game y‑axis can distort perceived performance. ## Full Transcript
Data, data, data.
Today, we have access to more data than ever before.
So if that's true, what are we supposed to do with all of it?
And I don't have an end-all be-all answer for you,
but one way we can better understand our data
is with a technique that has been around for a long time and that you've probably heard of before, data visualization.
Data visualization is when we display information in formats like graphs or charts
so that we can more easily understand it.
But before we jump into the specifics, let's acknowledge something.
Data visualization is not always the answer.
The best practices for data visualization are going to depend on what
kind of data you're working with and who you might be sharing it with.
But with the right strategies and intentional planning,
data visualization can be an incredible way to share and understand data or augment an existing analysis.
understanding the data that is available to us is vital for setting ourselves up for success.
So let's talk about three sets of do's and don'ts for data visualization
to make sure that we avoid critical errors and that we are getting the most out of our data.
So do keep your visuals simple and digestible.
and don't make them needlessly complex and intricate just because you can.
There's a time and a place for complex, intricate visuals,
but I would wager that most of the time, less is going to be more when it comes to data visualization.
And remember, the point of a visualization is to help bring your audience to a conclusion.
So making it as easy as possible to bring them there is gonna be your best bet.
Some examples of what this might look like could be using abbreviations where appropriate,
making sure that you're not using too many colors and shapes,
making sure that you get rid of any unnecessary variables, or exclude any data where appropriate.
A second do is to know your audience and include context.
And a don't is don't assume that your audience is going to have the same data expertise as you.
One of the hardest parts of working with data is making sense of it for both the most data savvy among us,
and for those who just don't care about the nitty gritty of working with it.
And oftentimes you're gonna be making visuals for people who fall in both of those buckets and everywhere in between.
So it's crucial to consider this beforehand to make sure that you're getting the most out of your data visuals.
Where your data nerds might really understand advanced visualization techniques, your data novices,
might just want to see a bar graph with a couple of colors on it to get the point across.
It's crucial to consider this beforehand to the best of your ability to make sure that you're not leaving anyone behind,
but also that you're not leaving any insights on the table.
A third do is to be clear with titles, axes, and legends,
and don't be misleading or leave anything up to the audience's imagination.
It can be easy to be misleading if you aren't careful about what you're doing.
Oftentimes, visuals that you create are gonna be viewed in a vacuum somewhere down the line,
meaning that someone's gonna be looking at it, but you are not gonna be there to explain what you were doing.
So if you're doing any significant filtering of the data or doing anything strange with it,
consider adding that in a title, an axis, somewhere in a footnote
so that whoever's viewing your visual can be sure of what you were doing at the time of making it.
And speaking of axes, be very intentional about the scales that you're using so as to not mislead your audience.
As a quick example, let's talk basketball for a moment.
Let's say I'm trying to decide who's better between Michael Jordan and LeBron James,
and I'm using points per game to decide.
If I put this information into a bar graph, I can modify the y-axis to make it look a lot different than it actually is.
So, if I modify the Y axis here and make it go 26 to 31,
I can make it look like Michael Jordan is way better than LeBron James,
but if I'm honest with my audience, and down here I'll use a zero to 31 axis,
we can see that in reality, it's a little bit of a closer battle
than it would look like above.
Maybe we should look elsewhere to continue this argument,
like finals victories.
Just kidding, I'm biased, I'll admit it.
When creating data visualizations, simplicity is key.
Keep it simple and be sure to know your audience and not to assume that they have the same data expertise as you.
And keep it clear with your titles, with your axes, with any legends or footnotes that you might include.
This is a non-exhaustive list, but these are just a few best practices to
make sure that you're getting the most value out of your data.