“Clutter and confusion are not attributes of data – they are flaws of design.” -Edward Taft
Data visualization: to represent the data information after analysis and processing in the form of some kind of schematic diagram.
In other words, it’s a way to visually convey some information about the content of the data. Depending on its data properties, data can be represented in many different ways, such as line charts, bar charts, pie charts, scatter charts, or maps.
Of course, for data analysts and graphic designers, data visualization is not simply the use of graphics to represent data, and it is very important to observe the means of data visualization to best present the results of data analysis.
Not just visually appealing, but never misleading. Especially when dealing with very large data sets, developing a uniform format is critical to creating visualizations that are both useful and appealing.
Why use data visualization?
According to IBM, the world creates 2.5 terabytes of data every day.
Andrew McAfee, a research scientist at M.I.T., and Erik Brynjolfsson, a professor at M.I.T., noted that “more data is passing through the Internet per second than was stored in the entire Internet 20 years ago.”
As the world becomes connected to more and more electronic devices, the volume of data will continue to grow exponentially. Scientists predict that there will be 163 ZB (163 trillion GB) of data by 2025.
It’s hard for the human brain to make sense of all this data, in fact, it’s hard for the human brain to make sense of numbers greater than five without some kind of analogy or abstraction. Data visualization designers can play a crucial role in creating these abstractions.
After all, big data is useless if it can’t be understood and used in useful ways. That’s why data visualization plays an important role in fields ranging from economics to science and technology, healthcare and human services. By turning data and other information into charts, content becomes easier to understand and use.
When to use data visualization?
Past data analysis reports have made it difficult to understand the information behind large amounts of data in a quick and unambiguous way. Data visualization can clearly and effectively convey the important information of all the data generated in the process of enterprise operation in the way of visual charts. Many enterprise managers have seen the value of data visualization from practice, which enables decision makers to solve the problem that it is difficult to quickly understand data analysis reports and understand data with data visualization mode, so as to make better decisions for the enterprise.
Whether in business, technology, science, or other fields, you need to understand big data sets to make informed decisions. Clear data visualizations make complex data easier to grasp and therefore easier to take action on.
Iii. What are the principles to be followed in the design and production of data visualization?
1. Define project objectives
Data visualization should answer important strategic questions, provide real value, and help solve real problems.
For example, it can be used to track performance, monitor customer behavior, and evaluate the effectiveness of processes. At the beginning of the data visualization project, it should be clear how much time is needed, clear the project purpose and the priority of data analysis and presentation, and the final data visualization effect should be useful, avoiding wasting time to create unnecessary visual effects.
2. Know your audience
Data visualization is useless if it is not designed with clear communication with the target audience in mind.
It should be compatible with the audience’s expertise and allow them to view and process data easily and quickly, taking into account their familiarity with the basic principles of data presentation, whether they are likely to have a background in data visualization and whether they need to review charts regularly.
3. Use the right data chart
There are many kinds of charts, and choosing which type of data is best for visualization is an art in itself. The right chart will not only make the data easier to understand, but will also display it in the most accurate way. In order to make the right choice, you must fully consider what type of data you want to transmit and to whom you want to transmit it.
4. The most popular types of data visualization charts
1) Line chart
Line charts are used to compare values over time and are useful for showing both large and small changes, and they can be used to compare changes to more than one set of data.
2) Bar chart
Bar charts should be used to compare several types of quantitative data. They can also be used to track changes over time, but they are best used only when those changes are significant.
3) Scatter diagram
A scatter plot is used to display the values of two variables in a set of data. They are perfect for exploring the relationship between two groups.
Source: Wall Street Journal/US unemployment Statistics
The pie chart 4)
Pie charts are used to show a portion of the whole. They can’t display things like changes over time.
Figure source: network
Fourth, the data visualization design should maintain a certain degree of coherence
Consistency is especially important when compiling large data sets into visual charts. A coherent design will effectively fade into the background, making it easy for users to process information. Good visualizations help viewers draw conclusions about the data being presented.
When creating a data hierarchy, there are ways to show decision-makers individual data points that need to be highlighted. For example, you can order them from highest to lowest to emphasize the highest value, or highlight the categories that are more important to the user.
Even changing the order in which the data is displayed, the colors used (such as brighter colors for the most important points, or gray for baseline data) and the size of the individual elements of the chart (such as extending some slices of the pie chart to the regular border of the chart) can help users interpret the data more easily.
5. Chart color selection in data visualization design
Color is widely used as a way of representing and distinguishing information. According to a recent study conducted, it is also a key factor in determining user decisions.
Some researchers have analyzed people’s reactions to different color combinations used in charts and found that they prefer palettes with subtle color changes because they are more aesthetically appealing to people.
However, scholars have found that although the color palette with subtle color changes is very attractive, the use of delicate colors in visual charts will make the charts direct and difficult to distinguish, leading to the failure of effective data analysis and insight, which will completely violate the purpose of creating visual display data.
Therefore, we should pay attention to using some skills to improve the readability of graphs in the color selection of data visualization charts:
- Use high-contrast colors
- Use colors with patterns or textures to convey different types of information
- Mark elements with text or ICONS
Filling the map plates of different countries with high-contrast colors can make people clearly understand the classification of different regions at a glance, and convey the information clearly.
Never distort data in data visualization
Good data visualization should tell the story clearly, avoid distortion, and avoid visual representations that do not accurately represent the data set, such as pie charts in 3D.
A 3D pie chart like this makes it difficult to actually show the scale of each slice.
Data visualization should be used to allow viewers to draw certain conclusions without distorting the data itself, a principle that is particularly useful when designing things like infographics for public consumption.
Data visualization diagrams are often created to support specific conclusions rather than simply convey data. Therefore, designers can use things like color selection and specifying specific data points to emphasize key data rather than a misleading chart form.
Bad examples of data visualization design
Not starting the Y-axis from zero can make the data appear to have a larger gain than it actually does, which makes the visualization misleading and fails to clarify the displayed data.
Another example of a chart where the Y-axis does not start from zero, thus distorting how the results are displayed.
The bar chart of this major brand is misleading in scale because there is no Y-axis. Even a small difference (less than 1%) can magnify the oversized blue bar out of proportion.
Good examples of data visualization design
Bar charts like this are a great way to show differences between data sets, although the enhanced color contrast will make the image more accessible to visually impaired users.
The plant transport data visualization uses several different visualizations to show the relevant data in an easy-to-understand format at a glance. The data also has good labels, which can better show the operation status of the factory.
Combining a clean, uncluttered design with easy-to-interpret data visualizations and simple diagrams provides a great user experience.
Interactive data analysis visualizations also do an excellent job of making the data easy to understand.
Nine, conclusion
Good data visualizations convey data sets clearly and effectively through the use of graphics, and the best visualizations make it easy to understand the data at a glance. Visualization breaks down complex data information to make it easy for the target audience to understand and make decisions based on it.
“The basic criterion of design is how much it helps to understand the content, not how fashionable it looks.” Data visualization, in particular, should follow this idea, with the goal of enhancing data through design rather than calling attention to the design itself.
Keeping these data visualization design points in mind will help you create data visualization infographics that are truly useful to your audience.