4 Principles of Excellent Data Visualization

From the courses I've taken and the challenges I've hitherto participated in within data analysis and data visualization, I've learned a lot of important lessons for how to conduct great analysis and to make stunning visualizations. The lessons from the challenges, however, have not only been limited to the practice I've gotten myself, but from witnessing the fantastic work of people with way more practical experience in the field than myself.

Based on the challenges I've participated in during the last month by Maven Analytics (Unicorn Company Dataset) and Onyx Data (Nobel Prize Dataset), I wanted to present my three favorite entries in these challenges as examples of what I think are some of the most important principles in conducting excellent data visualization that really stands out: Mina Saad, Özge Deniz YILDIZ and Stephanie French.

1. Clear insight and narrative are the most important parts of a visualization. Visualizations shouldn't be meant for just showing numbers and graphs, but rather for telling a story, and the numbers, graphs, tables, etc. are tools for presenting that. Özge's visualization, for instance, explains clear insights briefly for each section with clear highlights of the important parts, and Stephanie's visualization, although being larger and having more text, allures the reader with a gripping story and clear narrative about the insights that can be drawn from the dataset with her graphs, making it more appealing to read them through to learn more, as well as clearly highlighting with color for emphasis.
2. Real estate management is key. The saying "Keep it Simple Stupid" is extremely relevant in data visualization, and the mistake many analysts make is that they believe that more is better when it comes to including graphs and numbers to visualize. The reality is the opposite: Less is more. There were many entries in the challenges with plenty of graphs and numbers, but a reason I thought Özge's stood out in the Unicorn challenge was that it was simple and clean, whereas other entries were way too messy. Though there may have been more insights to take out of the dataset, this visualization was better at prioritizing the use of space. In the Nobel Prize challenge, Stephanie's entry also performed excellent real estate management. The visualization was bigger with more items included, but what's great about this one is that there isn't a single thing I would advise to remove from it. Each plays an important part in presenting the narrative, and this shows that she's been thorough in analyzing and understanding the dataset, as well as researching to present an outstanding presentation, which I frankly think is so good that it would deserve to be an official poster/visualization for the Nobel Committee.
3. Follow psychological teachings of visual appeal. Gestalt psychology presents several principles for perception, and these can be leveraged to improve your visualizations. The principles of similarity, proximity, continuity, and figure/ground are here especially relevant. In Stephanie's visualization, for instance, the highlighted parts of the graphs are in a bright yellow, whereas the background is lighter or grey to lie more in the background. The pictures of double Nobel Prize winners next to each other also follow the principle of proximity giving an immediate first impression that they have something in common, and the consistent use of colors between yellow, black and grey through the visualization gives a red thread through the visualization by the principles of similarity and continuity, making it appealing to read through all the text too for the special insights.
4. Excellence lies in the details. Although I can make some decent visualizations after a couple months of learning data analysis and visualizations, there are some things that clearly distinguishes my work from the experts with years of professional experience. Mina's visualization is a good example of this, as the custom background design, iconography, coloring, and font really makes it stand out. Furthermore, Stephanie's visualization also shows excellent focus on detail, and it's clear that there has been thorough effort in the analysis process given aspects like the use of icons and being able to identify and present in a clear and interesting narrative the share of immigrant winners, collaborations, and the detailed breakdown of the representation of women among Nobel Prize winners.

I hope these tips can be helpful for reflecting on how to improve your data analysis and visualization abilities. These visualizations have at least been inspiring for me to understand better how I can take the next step from good to great in terms of my visualizations. Comment below if you disagree with any points or would like to add other tips in the matter, and if you want more guidance for improving your analyses and visualizations, feel free to connect and message me on LinkedIn.