Data visualization is an important part of the data driven industry. You can use all the newest technologies and developments to collect and analyse data, if you’re not able to show your data properly, everything else is pretty much useless.
Data visualization is one of the most important things around data analysis and at the same time it’s very hard to communicate your findings in a proper and understandable way. As a data analyst, scientist, specialist, call it whatever you want, you get familiar with the data, you know what you’re looking at, you know what it all means and you see the structure. But it’s also the other way around. Because you are the specialist, you are the data ‘nerd’. That means, all the other people (the people that have to look at your results and take decisions based on that) are not. They need to onderstand, and to make that happen, data visualization should be kept as easy as possible. Because if not, your work is pretty much useless.
If the decision maker doesn’t understand, because your graphic is too complex, then how are you going to convince them?
Easy & simple
Easy and simple comes down to minimizing and/or deleting everything that you don’t want to put any attention on. Or, maximizing where you want to put your attention on. This can be reached by adjusting background & grid, colors of the graphic, titles & axes, amount of data in one graphic.
Background & grid
Keep your background in the graphic as basic and light as possible. Because the background is not what the focus is on. Consider whether you really need a grid. What does it add to your graphic and number? By minimizing those two, you already focus more on the data. The bar is what’s it about, the line is what you want to show. Less is more. But by adjusting this, you’re not quite there yet.
Focus on what you want the graphic to tell. In every detail. Take time to think about it. Again, less is more. Do you show a line chart? What line is it that you’re gonna say something about? Make sure that one stands by colour and put the rest of the lines in a more neutral color. Do you show a scatterplot? What part of your scatterplot stands out and should be highligted. Make sure you mark that by a distinctive color. Besides randomly picking a color, think about the color used by the organization you’re working for and try not to use more than 5 different colors.
Also, most brains have an automated expectations when they see colors. Red is often seen as wrong, bad, emergency, alert level. Green is usually seen as good, positive, doing alright. Think about that when you pick colors, depending on what you want to show. I am not a designer. I want the data to show off. I’m not here to make the most, bright, complex, colorful, designer like graphic. That’s what designers are for. It is not about that beautiful colour that you’ve found online. Or trying to make the newest, most complex, 3D chart. Because it’s about the data, and the color and the type of chart is there to help show your data.
Titles and axes
Titles and axes are seen as guides to guide people to the proces of understanding the graphic shown. Make sure there is always a title that really summarizes what is shown in the graphic. Short but enough to explain what’s in the graphic. Axes are used to explain what the numbers mean. People need to understand immediately what a number means, to be able to interpret the graphic. If there is more graphics about the same topic, make sure the numbers in the axes are the same. It helps people to interpret and combine the lines and don’t give a false message.
Make also sure the numbers are evenly spread out, with the same distance in between. And if you show multiple graphic about the same topic, make sure your axes are the same on every graphic. As mentioned above about colors, keep it simple in a grey color for example. Everyone can read it and it doesn’t distract from what you want to show.
Combining data into one chart
Sometimes they ask you to present something on one sheet, even tho you have to present multiple results. Combining everything in one graph sounds like THE solution, but it’s not. One graphic with combined data is very complex for a brain to understand. A graph should explain itself by clear and only necessary input. A graphic with a line and a bar chart, (sometimes even crossing each other), with an ax on the left, right and bottom, is not gonna make it. A manager, a colleague and also you mum, will look at it for 5 second, raise their eyebrows, look up to you, not saying much, but thinking something like: what the f*** am I looking at. And if that’s the case, you lost.
Rather have a few different understandable smaller charts on one sheet, than one big one with all the data combined.
If you ask me, I could probably read it and understand quite a few charts, because I work with data every day, have seen a ton of charts and I’m quite well trained to understand. You’ll probably find a few more data scientists and analysts and a manager that have a good feeling for numbers. But that’s exactly it. You don’t want it to be luck, you want to make sure that your analysis is going to be understood, by the manager, your colleague, even your mum (or dad). That’s how you should think about data visualization if you ever have to make a graphic again.
I’m honest, it’s challenging for me sometimes to even know if I made it easy enough to understand but the more I do it, the more I learn from it as well. I still make graphics and when I look back at it I realize I could have done better. But hey, everyday is a schoolday.
Addition to the graphics & data
I split up the combined graphic to show how much more clear this becomes in two graphics compared to two. It's questionable whether you should use the line in this graphic or that a bar chart is more useful.
Note: Graphics are not based on real data.