How many visualizations are there in the tableau

Learn from data journalists: visualize and communicate data effectively

Today's knowledge worker not only has to find, edit and analyze data, but increasingly also has to be able to process it graphically in order to effectively communicate his point of view with data visualizations. How to do this can be learned from data journalists who have made data visualization a true art form. Using examples, it is argued that effective data visualization as seen in the media achieves one or more of the following three goals: 1) it attracts the reader's attention, 2) it helps explain complex issues, 3) it makes the reader remember the content of the text.

Today’s knowledge workers not only have to be able to find, process and analyze data, but increasingly have to be able to present them in graphical form, so as to effectively communicate their point with data visualizations. One can learn from data journalists how to do this, who have taken data visualization to a form of art. With the help of examples, this paper argues that effective data visualization, as it can be found in the media, tries to achieve one or more of the following aims: 1) it attracts the attention of the reader, 2) it helps explain a fact, and 3) it creates lasting memories of the text's content.

Dans notre société de l’information, les professionnels doivent non seulement être capables de trouver, traiter et analyzer des données, mais également de les mettre en forme graphiquement afin de communiquer efficacement leurs idées. Les data journalistes sont parmi les meilleurs maîtres pour se former à la visualization de données tant ils ont érigé cette discipline en art. Au moyen d'exemples, cet article démontre que la visualization efficace de données, telle que pratiquée dans les médias, cherche à atteindre au moins un des trois objectifs suivants: 1) attirer l'attention du lecteur, 2) aider à expliquer un fait , 3) laisser une impression durable du contenu du texte.


Today's knowledge worker not only has to find, process and analyze data, but increasingly also has to be able to process it graphically in order to communicate his point of view effectively with data visualizations; be it on a personal blog, on the company website or on social media. Telling stories with data is the new trend.

Data journalists therefore occupy a top position. Using examples from the media, this text tries to show how good data visualizations can be implemented in practice.

Data journalists have made data visualization a true art form. This means that the works are largely based on the artistic intuition of the respective designer. Nonetheless, there is also increasing scientific research into the effectiveness of different types of visualization. If possible, reference is made in this text to the relevant sources.

What Makes Data Visualization Effective? Successful data visualization achieves one or more of the following three goals: 1) it draws the reader's attention to the article, 2) it promotes understanding of complex issues, 3) it makes the reader remember the content of the text. In the following, these three goals will be examined in more detail separately.

1 arousing interest

Love it or hate it, but the reality is that the media is actively vying for the reader's attention. In the age of dwindling print runs and an increasing selection of news channels on the Internet, they try to get as many clicks as possible by all means. To achieve this, almost every article now has a preview image, not just on the homepage, but - almost more importantly - on Twitter, Facebook, Bendle or one of the many other apps and social networks that are now replacing traditional news channels. The reason is simple: as Twitter found out in a study, a tweet with a preview image gets four times as much attention as one that contains only text. And a tweet with a series of images is clicked more often than one with just one image.[1]

And more and more often you can also see a data visualization as a preview image instead of a photo, such as For example, in Figure 1. This may be because no suitable photo was found for the story, that the author wants to stand out from the crowd, or that the dates themselves play the main role in the story.

In most cases, the data visualization was designed in such a way that it is suitable as a preview image, preferably in a size that fits on a smartphone screen. The majority of readers now consume their messages via mobile devices.

However, there is no guarantee that the reader will read the text after opening the article. In the age of tight attention spans, a common strategy is to break up longer texts with multiple multimedia elements, including data visualizations.

Some articles, such as the one from the Süddeutsche Zeitung on the anniversary of the first ascent of the Matterhorn z. B., go one step further (Figure 2): You integrate text and visualization in such a way that the displayed graphic adapts to the reading progress. In this example, the corresponding intermediate stations of the mountain ascent are displayed on a map. In this way, the reader is not only informed, but playfully encouraged to read on.

2 Promote understanding

Data visualizations are - if they are well done - a powerful tool for communicating numerical facts. This is because visual elements act on the pre-attentive perception. This means that visual patterns are processed by the brain without great mental effort.

In order for data visualization to achieve this, three things must be given. First, the type of graphic should match the message you want to communicate. Second, the visualization should be free of distracting graphic elements such as superfluous 3D effects, shading or z. B. be bold grid lines.[2] Thirdly, it helps if the lettering, choice of color, and notes make it clear to the reader what he should see.

The data visualization in an article on in Figure 3 is a good example. The authors had researched how many football players who have accrued for the national team since reunification come from the new federal states. The graph shows the number of goals scored by national players since reunification. You can see immediately that players from the old federal states scored three times as many goals as the players from the new federal states.

The type of data visualization selected here - namely horizontal bars - is a popular choice nowadays, which is always available when you have a quantitative number (e.g. here the number of goals) over various subgroups (e.g. here the origin the player) wants to compare across the board.

Research studies have shown that the brain can easily compare the length, horizontal and vertical position of graphic elements.[3] In the example above, you can immediately see that the red bar is three times as long as the black.

The studies also showed that other graphic properties such as B. Angle, area size, or color saturation and shade are less suitable for comparing quantitative measures.[4] For this reason, one also sees pie charts - here the brain has to compare the angles and areas of different pieces of cake - not quite as often as it was a few years ago. Pie charts are only suitable if you want to show the relationship of one (at most two) segments in relation to the big picture.

In contrast to vertical columns, horizontal bars also have the advantage that the labels are easier to read because you don't have to turn your head. You do not need a color legend for this, as the bars can be labeled directly. That makes it easier to see what to see right away.

The choice of color also helps. In Figure 3, the contrast between the black and red bars makes it immediately clear what to compare. While the color is less suitable to represent quantitative differences, it can be used to highlight categorical differences, as seen in this example.

Another trick is to sort the bars according to size, such as: B. in Figure 1. Here the view is immediately directed to the longest bar. But, if the story is about number two or three, then you have to highlight the bars in a different way (arrows, lettering, colored highlighting, etc.), such as. For example, in Figure 4, a graphic from an article discussing the controversial third runway for Heathrow Airport.

Figure 5 serves as a counterexample: At first glance you don't know what to look at. The graph shows various causes of death in the United States. The headline says more people are killed by firearms than terrorist attacks. But you don't immediately see that in the graphic. The reason is that while the color scheme looks very nice, it is not effective in keeping the eye on the essentials. In addition, it remains unclear whether one should compare the values ​​for the individual years or the various categories. Ultimately, you have to rely on a color legend, so you have to look back and forth a few times until you know which bars represent which cause of death.

In comparison, Figure 6 manages to convey the same story better, even if it simplifies the data a bit and focuses on two causes of death. Here you can see immediately that each year significantly more Americans die from firearms than from terrorist attacks. The time trend (which is of secondary importance here) was shown here with a line diagram. The white space between the lines and the contrasting choice of colors (blue and red) make it immediately clear which differences are relevant here.

In Figure 6, the problem has been solved by keeping the dataset as simple as possible; unlike in Figure 5, only two time series are compared with one another. Keeping data visualization to the essentials is often the best way to deal with complex information.

Of course, that doesn't always work. Reality is not always black or white. Sometimes you don't just want to depict simple opposites, but rather illustrate differences in differences. There are different ways. For example, you can build up the graphic in stages and explain each step to the reader. Or you can distribute the data to different visualizations. It is often advisable to arrange several similar graphics in a grid (so-called "small multiples"), as in Figure 7.

You can also try another type of data visualization to illustrate what you want to show. A so-called dot plot is suitable e.g. B. often to show differences in differences, as Figure 8 shows. This graphic is about Borussia Dortmund's race to catch up during the 2014/2015 season and the question of whether a team has caught up 11 places in the table in the history of the Bundesliga. You can quickly see - by comparing the distances between the points, i.e. the length of the connecting lines - that in addition to Borussia Dortmund, Hamburger SV has also managed a similar race to catch up.

Create 3 memories

However, a clean, easy-to-understand data visualization is no guarantee that it will reach the reader or that he will internalize the core message in the long term.

The memorability of data visualizations was little studied until recently. There are only a few scientific studies that deal with this.[5] In an experiment it was shown that artistic equipment of the graphic can help to better remember the content of the visualization.

Decorating data visualizations with icons and other graphic elements was popular in the 1980s and 1990s; nowadays we are more likely to see a more minimalist design that aims for intelligibility, as explained above, and as suggested by data visualization gurus like Edward Tufte[6] or Stephen Few[7] is propagated.

Even so, there are many good examples of memorable data visualizations these days. Data journalism experts like Alberto Cairo[8] or Cole Nussbaumer Knaflic[9] would say the most memorable data visualizations are those that tell a story.

How do you achieve this storytelling effect? Like any good story, you need an introduction that explains the context, a body that discusses the problem in detail, and an ending that resolves the problem.

For the "main part" you usually need differences - all good stories are based on opposites between the main characters. Likewise in data visualization. So you should try to find them in the data - rich versus poor, liberal versus conservative, what was versus what could have been, etc. - and work it out visually using the aids described in the previous section or the reader the possibility to give to explore them yourself.

Furthermore, a good story manages to pick up the reader at the point he is already familiar with. The less the reader knows the world of the story (such as a fable world), the more important the introduction that describes it becomes.

Sometimes you can do that on the same graph, like in Figure 6. As a layperson, you don't necessarily have an idea of ​​how many people die from various causes of death in the United States. However, most of us can still remember September 11th. The number of victims of terrorist attacks is represented by the “hill” at the beginning of the blue line - the value for the year 2001. This allows the numbers to be compared with something that we have a concrete idea of.

In the case of more complex issues, you first have to show another graphic as an "introduction", for example in Figure 9 and Figure 10. These are from an article about the performance curves of the fastest 100m runners in the world. With the help of the first graphic, which shows the annual peak times of the individual runners in relation to age, the authors show that most athletes run slower times again after a certain age. According to the dashed trend line, the average performance peak is around 22 years (Figure 9).

In the article, the careers of several athletes are discussed individually. Figure 10 shows e.g. B. Usain Bolt's and Justin Gatlin's performance curves. The former ran its best time so far - the current world record - at the age of 22. Since then, things have been going downhill for Bolt. Will he be able to set a new record again? The form curve of Gatlin, who ran his best time at the age of 33, gives hope.

This data journalistic arc of tension was achieved here through the use of individual diagrams. But you could also have used an annotated animation or a “slide show” with different slides. It is only important that the reader is not flooded with information, but slowly introduced to the story. Then the chances are good that he will also take something away from it.


It has been suggested that data journalists have three different intentions when using data visualizations: to generate interest, promote understanding, and create memories.

While this is a helpful framework for examining good examples, it should also be understood that these goals are not always clearly separable. Characteristics that make a graphic understandable can also help to keep it in mind longer.

It may also be legitimate to aim for only one or two of the three purposes. Especially since it sometimes seems to be the case that the categories contradict each other. It can happen that someone gives up easy understanding in order to achieve a high attention effect with a very unusual visualization.

The fine art of data journalism knows how to achieve all three goals with data visualization.

It is hoped that this framework will make it easier for knowledge workers who wish to communicate their data to learn from the many examples found in today's media.


Bateman, S. Mandryk, R. L., Gutwin, C., Genest, A., McDine, D., Brooks, C. (2010). Useful junk? The Effects of Visual Embellishment on Comprehension and Memorability of Charts. In: ACM Conference on Human Factors in Computing Systems. Atlanta. 2573-2582. Search in Google Scholar

Borkin, M. A., Member, S., Vo, A. A, Bylinskii, Z., Isola, P., Sunkavalli, S., Oliva, A., Pfister, H. (2013). What Makes a Visualization Memorable? In: IEEE Transactions on Visualization & Computer Graphics 19, No. 12. 2306-2315. Search in Google Scholar

Borkin, M. A., Bylinskii, Z., Kim, N. W., Bainbridge, C. M., Yeh, C., Borkin, D., Pfister, H., Oliva, A. (2015). Beyond Memorability: Visualization Recognition and Recall IEEE Transactions on Visualization & Computer Graphics 22, No. 1. 519-528. Search in Google Scholar

Cairo, A. (2012). The Functional Art: An introduction to information graphics and visualization.San Francisco: New Riders. Search in Google Scholar

Cleveland, W.S., McGill, R. (1984). Graphical perception: Theory, experimentation, and application to the development of graphical methods. In: Journal of the American Statistical Association 79, No. 387, 531-554. Search in Google Scholar

Few, S. (2004) Show me the numbers: Designing tables and graphs to enlighten. Burlingame: Analytics Press.Search in Google Scholar

Knaflic, C. (2015). Storytelling with Data: A Data Visualization Guide for Business Professionals. Hoboken, NJ: Wiley-Intersience.Search in Google Scholar

Robbins, N. (2004). Creating more effective graphs. Hoboken, NJ: Wiley-Intersience.Search in Google Scholar

Tufte, E., Graves-Morris, P. (1983). The visual display of quantitative information Cheshire: Graphics Press. Search in Google Scholar

Twitter, “#TweetSmarter”. in Google Scholar

Florian Ramseger is a product specialist at Tableau Software. There he helps users of Tableau Public to visualize their data. Before that he worked for the World Economic Forum and the International Red Cross in the field of data analysis and data visualization. He holds a Masters in Economics from the London School of Economics.

Published online: 2016-2-5
In print: 2016-2-1

© 2016 Walter de Gruyter GmbH, Berlin / Boston