Here’s a quick memory challenge: think of the last bar chart from a presentation that you remember. Now compare that to a map of the United States that details the results of the 2016 presidential election. Obviously, you likely had a more personal investment in the election, but do you think the image would have stuck in your head had it been displayed as a drab 50-state bar chart? The same principle can be applied to the visualizations from your spreadsheet, your big data analysis, and your business intelligence stats.
In this article, I’ll go over some quick data visualization tips pulled from around the web. These tips can be used to give your presentations some kick and make your visualizations memorable. After all, what’s the point of presenting something if people are going to quickly forget it?
Use the right type of chart or graph
First and foremost, it’s important to choose the right type of visualization. And when I say “visualization”, I’m referring to any visual representation of your quantitative data. There are visualizations that venture more into the realm of a newsy infographic – and there’s nothing wrong with them – but I’d like to stick situations where you might need to present something from your business dashboard. So, let’s go for a quick refresher over some traditional types of data visualizations:
- Pie: A chart that displays quantity based on the size of the area. To put it more simply: it’s like a pizza (pie) and a big slice means a greater value. They’ve also been around for a long time (see image below).Great for: giving an overview of information that is easy to understand. Showing different parts of a greater whole.Not great for: getting into specifics of data (e.g. picking out the difference between 16 and 19 percent). Pie charts can get overwhelmed by lots of smaller amounts; it’s difficult to show 10 different slices that all account for 1-5 percent of the chart. Showing change over time.
- Bar/Line: A chart with an x and y axis that displays multiple values (horizontally or vertically), with higher values being represented by longer bars. Great for: displaying changes over time and showing change via the x and y axis.Not great for: showing small changes over time, data that has a large gap between some of the higher and lower numbers.
- Scatter plot: A graph that heavily utilizes the x and y axis to plot points across the graph. It emphasizes how much one variable is affected by another. These points can literally be scattered across the graph.Great for: Getting a more precise look at the correlation between data, finding outliers, and handling multiple data points.Not great for: displaying simple and easy-to-follow graphs.
Those are some classic examples of ways of displaying info and what you’re most likely to see out in the wild. However, there are plenty more types of visualizations that you probably come across on a regular basis, such as bubble chart, radial tree, or even the NYC subway map. If you’re interested in seeing more visualizations, check out this guide from the Duke University library.
Information is beautiful
David McCandless is an expert in data visualization and runs Information is Beautiful. The site is worth spending time on and displays visualizations – usually interactive – on a variety of topics that include the gender pay gap, cooking oils compared, and the effectiveness of natural supplements. If you are looking for inspiration for a visual you want to put up, Information is Beautiful is a great option.
Color is important
If you’re more of a textual person like me, it’s easy to overlook the importance that color plays in your visualizations. I’m not an expert on color, but I can give you a few pointers to keep in mind, plus some resources that can your answers questions better than I can.
- Pick appropriate colors: if you are giving a presentation on declining traffic numbers, it’s not a good idea to use the color green to represent the decline in numbers. Why? Green is typically shown to be a color that indicates growth, while red is used to show decline or debt (Also see: stop signs, red alerts).
- Use a variety of clearly-differentiated shades: It’s fine to use different shades of the same color on a visualization. Different shades of color can be used to show intensity (e.g. dark red indicates a larger the number, light red/pink a small number) and they also work well on visualizations such as maps. If the difference in shades aren’t clear to your reader, you’ll run into some problems.
- Limit the number of colors you use: Although it can be tempting to use a rainbow color (formerly known as “spectral color”) scheme, it’s actually not a great idea. As mentioned above, using shades of a color allows for showing nuance as well as showing a gradual change in data.
- Be mindful of the color blind: Color blindness is frequent enough to affect 1 in 200 women. That can be a high number when you have a population in the millions, but it becomes prolific when you factor in men: 1 in 8 have some form of colorblindness. To put that into perspective: at least 4.5 percent of the US population is colorblind. Vischeck is a tool that will test your images for any problems in regards to color blindness.
Don’t go wild with your new features
Although it can be tempting to create an interactive 3D scatter point chart that features 30 vibrant shades of green, it’s not the best option to show off all the features you have available. It’s the same as watching a PowerPoint presentation that includes a flashy animation and an accompanying sound for every slide transition; it would get distracting fast. Instead, focus on making your results clear. If you want to show consumer purchasing power across Europe, try overlaying different the figures over each country. Take it a step further and use a yellow to red scale and use shading to denote lower/higher purchasing power.
Tableau recommends checking every visual with the “5-second rule.” They write, “Research shows that the modern attention span for looking at things online is, on average, less than five seconds. So if you can’t grab attention within five seconds, you’ve likely lost your viewer.Be sure your viz includes clear titles and instructions. Tell people succinctly what the visualization shows and how to interact with it.”
Data Visualization software
There are many ways to create a visualization, but the most common way is using a spreadsheet program like Excel. However, many visualization features are found on business intelligence solutions. Below are some examples of software that is either dedicated to visualization or is part of a larger suite.
Helpful links and sources
This article only scratches the surface when it comes to data visualization. There’s a ton of great information out there that is worth bookmarking for future reference. Check out the links below for more info:
- Information is Beautiful
- Top Ten Dos and Don’ts for Charts and Graphs – Duke
- Finding the Right Color Palettes for Data Visualizations – Graphiq
- 5 Expert Data Visualization Tips – Venngage
- 5 data-viz tips to let your data speak for itself – Tableau
- How to Choose Data Visualization Software: a Handy Checklist – GetApp
- User reviews of the top data visualization software on GetApp