Alexander König
Carla Märkl
They say “data is the new oil”, since whether in the IoT, in the smart factory, or in autonomous driving, gigantic amounts of data are constantly generated. This data is now the most important resource for innovations of the 21st century. But while machine learning and AI systems know how to deal with these huge amounts of data, for us humans they are just an incomprehensible collection of numbers and incoherent facts. Only when data is visualized graphically the underlying information, correlations, and patterns are revealed to us, which we can then understand and interpret.
A useful tool for presenting the data chaos in a clear and understandable way is data visualization, in which the data is usually presented in various diagrams or charts. In this blog post, you’ll learn what you should keep in mind and when which chart is the right one for your purposes.
the benefits of data visualization.
The purpose of data visualization is to process abstract data and put it into a graphical and visually understandable form that conveys meaning. This involves encoding the raw data by position, shape, size, symbols, and color. This encoded representation is then decoded again by the human visual system, creating an understanding of the data. The following list outlines the benefits of such data visualization:
- Understandability: Data visualizations present information in a way that is easier to grasp than mere numbers and letters.
- Transparency: In addition, data visualizations reveal relationships and trends that would otherwise be extremely difficult to see.
- Attention: Well-designed and appealing visualizations easily capture the viewer's interest and can sustain it for a longer period of time.
- Memorability: Simplified and elegant representations of data are far more memorable than raw data.
Data can be visualized in countless ways. The most common ways are charts for quantitative data, maps for geographic data, as well as flowcharts and tree diagrams to show relationships, processes, and interconnections.
Below, we focus exclusively on quantitative data and the most common chart types. Here, data of three different value types will be connected and subsequently visualized:
- Nominal values refer to a category without order, such as (Nevada, Ohio, Florida, ...) or (green, orange, blue, ...)
- Ordinal values refer to a category with a natural order, such as (small, medium, large, ...), (cold, warm, hot, ...), (Monday, Tuesday, Wednesday, ...)
- Quantitative values are precise numerical values, such as (1,2,3,4,5...) or (temperature values: -10°, ..., 10°) and (floating point values: 5.32; 10.36).
before we begin.
Before you start designing a data visualization, the following questions should be answered:
- What is the goal of the data visualization?
- Who is the target audience of the data visualization?
- Which prior knowledge do the viewers bring with them?
- What questions should be answered by the data visualization?
- What data is available?
- Which media do you want to use? Print or digital?
- How do you want to visualize the data?
- Which presentation possibilities are suitable?
general design and structure of diagrams.
Once these questions have been clarified, it’s time to start designing. When designing charts and data visualizations, it is important to keep the design basics in mind. Our blog post on dashboard design basics provides you with an overview. These are:
- An easy to read typography
- Conscious use of colors with high contrast
- Sufficient white space
- Hierarchies and groupings of the elements
- Use of a design grid
this is what a diagram consists of.
Title: The title can be either neutral or interpretative. A neutral title describes factually what the diagram is about and leaves it to the viewer to interpret the visualization and find out the core message. An interpretative title, on the other hand, already gives initial clues regarding the core message, which lends it a more journalistic character. For example, “Sales figures from 2015 to 2021” is a neutral title, while “Sales figures have risen sharply since 2015” points to the core statement and is thus more interpretative. You can also use both titles. In this case, the neutral title should be placed first and the interpretative title should possibly be displayed slightly smaller underneath.
Labels: The axes and values of a diagram are labeled to help the viewer identify the information in the diagram. Direct labels, i.e. those that are placed very closely to the respective elements, are preferable to indirect labels. Indirect labels are, for example, labels that are visually connected to the respective elements by lines or explanations in the legend of the diagram.
Legend: If the labels are not enough to explain the visual coding within a visualization, such as the meaning of the different colors in a pie chart, a legend has to be used. It is often found at the bottom right edge or above the diagram.
Diagram: The core of the visualization is of course the diagram itself. There are countless different ways to create a diagram. In the following, the design, the use, and possible sources of error of the different diagrams are explained.
types of charts – an overview.
Depending on the content you want to communicate, different chart types are more or less suited. We have summarized the most popular types of diagrams with their advantages and areas of application.
circle or pie chart: How do parts of a whole compare with each other and with the whole?
donut diagram: How do the parts of a whole relate to each other and to the whole?
bar and column chart: How do the absolute amounts of the individual items compare? In which order of priority do the items rank?
divided and stacked bar chart: How do parts of a whole compare with each other and with the whole? How do items compare with each other and accumulate?
line chart: How do one or more items vary with time or another continuous variable?
area chart: How items that vary with time accumulate?
which chart works for your purposes?
You now know what a chart is, what it consists of and how it makes data visualizations more understandable. We also gave you a first overview of different chart types in this blog post. However, how do you know when to use which chart to illustrate your data?
In the second part of this blog post, we’ll go into more detail about the different chart types and explain exactly how they differ and when they should be used. Subscribe to the Peakboard Blog to stay tuned for part two!
