What is data visualization and how do you make visualizations that work?

What is data visualization?

The exponential growth of data means that we can measure and reveal more and more. For many people, however, data is abstract. Data visualizations turn data into something tangible and readable, and help users of a dashboard to read and analyze it quickly. That can be a clear table, a tile or an interactive chart.

Data visualization is a method for converting abstract data into concrete information and possibly knowledge. You can present the information in different ways: as a report, chart, animation, map or dashboard.

The weather forecast shows several types of (real-time) data visualization, such as a map and a table with the temperatures for the coming days. Which type you use depends on the situation:

• Bar chart: when you want to compare different variables with each other.
• Line chart: when you want to visualize a trend or a development over time.
• Map: as soon as a geographical component is important.
• Table: when you want to show numbers and values in detail.
• Scatter plot: to show spread using two variables. You can also use it to visualize correlation.

Chart versus table

When do you choose a chart and when a table?

Why is data visualization important?

When you convey a story or advice, effective communication is very important. The recipient can easily misunderstand your message. Think of a message via WhatsApp. Why is a message misunderstood less often face to face? That has to do with expression, body language and sound. A sarcastic message, for example, is much easier to convey face to face than in a chat.

In short: expression and sound support the message. Our early ancestors were already visually and auditorily oriented, as a survival mechanism to assess dangerous situations, and we still are. You can convey a theory or advice in both ways, and it works best when you combine them: visualize the data and then explain the visualization in person. That is storytelling. Well-known tools for visualizing data and storytelling are Tableau, Power BI and Qlik.

Create a data culture

For many companies it is difficult to create a data-driven culture. The advantage of such a culture is that the organization can respond more flexibly, faster and smarter to events inside and outside the organization. One way to start thinking in a data-driven way is by visualizing data. Having employees actively create reports, charts and dashboards makes them think from a data perspective. You can promote this in the following ways:

• Create data visualizations yourself and use them as an example for employees and colleagues. A good example is followed.
• Have the results substantiated with data during the weekly team meeting.
• Show during a sales pitch that data sells. A good data visualization creates mutual understanding.

Three basic principles of data visualization

Start with a design on paper or a whiteboard. Making a design keeps you thinking creatively. Once you are satisfied with it, you continue in the tool. Keep the following basic principles in mind.

1. Garbage in is garbage out. If the data you use is incorrect or incomplete, the visualization gives an incorrect or incomplete picture of reality. So your data quality has to be in order.
2. Less is more. Make sure your visualization looks calm and offers as little distraction as possible. The aim is to convey a message, and you want the reader to focus on it.
3. Show, don’t tell. Make sure your target group understands the visualization. Who is it for? Shapes, format, legends and colors help.

Making an impressive infographic

An infographic is a special form of data visualization. Its purpose is to show complex information quickly and simply. A good infographic combines text and image and tells a story in a compelling way. It requires knowledge of analysis, design and storytelling. The steps:

1. Subject: think of a subject you want to visualize.
2. Analysis: do thorough research into the subject and make a mind map.
3. Conclusions: make sure you have interesting conclusions and facts.
4. Storytelling: create a storyline, an order in conclusions and facts.
5. Design: design the infographic with colors, text blocks, charts and images.
6. Visualize: create the visualization with your favorite tool.
7. Feedback: ask for feedback from people close to you and from professionals in design and marketing.

Research into effective data visualizations

In 1984 the scientists Cleveland and McGill carried out fundamental research into the speed and accuracy with which people perceive visualizations. They arrived at a logical ordering in the effectiveness of perceiving quantitative data. We perceive a change of position most accurately and quickly, typically a line chart or a map as background. Then follow length (bar chart), angle (pie chart), slope (line chart), area (bubble chart), volume (3D), density and color. Yet you still see many 3D charts in all the colors of the rainbow, although that is the least effective form of data visualization. Keep it simple.

Further reading

• Dashboards: 7 steps and 20 tips.
• Data literacy: reading and interpreting charts.
• Data analysis: what comes before the visualization.
• Data quality: garbage in, garbage out.

Sources

• Cleveland, W.S. & McGill, R. (1984): Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Methods. Journal of the American Statistical Association, 79(387).

About the author

Mark de Kort has been active in the world of data since 1994, from developing software for business intelligence and scorecarding to setting up and improving BI architectures. He still works hands-on, from data preparation and modeling to building dashboards.

About Diamond Decisions

Diamond Decisions helps organizations become better by enabling better decisions. We don’t start with data, dashboards or AI, but with the decision that needs to improve, and work back from there to KPIs, data and technology. We do that in four steps: inspiration and alignment, business modeling, analytics and data platform and data-driven decision makers. That way BI and AI don’t remain an IT project, but deliver value every day.