You’re learning data visualization design. What are the most important principles to master?

At this point everyone has heard “know your data, know your audience, choose a good chart type.” At my company we try to go beyond those platitudes and teach data visualization principles such as expressiveness and effectiveness. Expressiveness means that the data is encoded in visual form in a way that expresses what’s in the data, and doesn’t express what’s not in the data. For example categories with order (e.g. gold/silver/bronze) are shown to have order. Effectiveness means that the data visualization achieves its objective, whatever that may be. If the ability to make accurate guesses about the proportions in the data is critical, then encodings like position or length are going to be more effective than angle or color saturation.

Mastering data visualization design requires clarity, accuracy, and relevance. Simplify complex data sets, ensuring clear representation without distortion. Maintain consistency in design elements for a cohesive look and establish hierarchy to guide attention effectively. Engage viewers with interactive features and compelling storytelling. Ensure accessibility for all users and provide context to enhance understanding. Seek feedback to refine designs and continuously learn about emerging trends and best practices. By adhering to these principles, you can create impactful visualizations that communicate insights, engage audiences, and drive informed decision-making.

All data is important, but all data is not equally important! Focusing on the most valuable data, in a way that the business can understand, helps to eliminate the 'so what?' that business stakeholders sometimes feel when presented with a load of data dashboards....

Different data types and messages require different types of visualizations. Selecting the right chart or graph type (e.g., bar chart, line graph, scatter plot) is crucial to accurately representing your data and ensuring that your message is understood. So it's critical to understand your audience, the message you want to communicate, and the data available for the assignment.

Understanding our data is fundamental. we need to comprehend the structure, relationships, and limitations of your dataset to effectively represent it visually.

Understanding your audience is critical to creating the correct visualisation. What is appropriate for a highly data literate consumer compared to a novice can be vastly different, but the same key concept will apply to both, if the consumer has to work too hard to understand your visual then you've missed the mark.

The second step after understanding data is to know and understand your audience. Who are the audience, What they exactly want, What is their level of expertise, and so on. This will help you in knowing what exactly you need to focus on and explain them such that they can understand what you are explaining.

Knowing your audience is another essential aspect in data visualization design. Before creating visualizations, it's crucial to understand who the end users will be and what their needs, knowledge, and expectations are. What is their level of technical expertise? What information are they seeking? How do they prefer to consume information? These questions will help tailor your visualizations to be more effective and relevant for your specific audience. Whether it's executive presentations or detailed reports, adapting the style and content of your visualizations to your audience will ensure that your message is conveyed clearly and convincingly.

Choosing the right visual elements is crucial in data visualization design. Each type of data and each story you want to tell requires a unique visual approach. From bar and line charts to maps and scatter plots, selecting visual elements should align with communication goals and data characteristics. Are you showcasing trends over time? A line chart might be the best choice. Are you comparing different categories? A bar chart or scatter plot could be more effective. By choosing the right visual elements, you can highlight key information clearly and effectively, facilitating your audience's understanding and interpretation of the data.

Never use pie charts. The human brain has a tough time error perceiving sizes in a non square ratio. Pie charts are heritage from the past, and unfortunately still used. Sometimes in a new jacket called 'donut chart', but the message still stands. There are simpler alternatives: A bar chart when comparing sizes of the individual members, and a treemap to compare all members and the size of the total population.

Selecting appropriate visual elements such as charts, graphs, maps, and diagrams is crucial. Each type of visualization has strengths and weaknesses, so chooseing the right visual element that effectively communicates our data insights is crucial. Visual elements play a pivotal role in conveying the story hidden within our data.

Selecting and choosing the appropriate visual is critical to better understand to the audience. Make sure you focus on formatting as well, to avoid confusion and distraction. Every chart has its own use case, like what fields must be chosen for the right data visualization. Add labels, titles and legends to help the audience in better understanding of the data.

I always start my design approach with "Why is this not a bar chart?" they are the simplest and easiest visual to understand, and with refinement and formatting can create compelling visuals, but if there is a genuine reason a bar chart won't work that will inform the journey to the visual that will work. Using reference guides like the graph continuum or Visual vocabulary will give you ideas and options to explore.

Na minha trajetória de estudos e práticas o importante em visualização de dados incluem também: Simplicidade e clareza, evitar o excesso de informações que possam sobrecarregar o espectador. Escolher os elementos visuais que comunicam a mensagem de forma direta e eficaz. Acessibilidade, certificar-se de que sua visualização seja acessível para todos os tipos de público, incluindo pessoas com deficiências visuais ou cognitivas. Utilizando técnicas como contraste adequado, legendas descritivas e opções de zoom para garantir que todos possam compreender a informação apresentada.

Data is only as valuable as it is presented. In my experience, it is essential to create charts, tables etc. whose main points can be easily interpreted without outside commentary. This is an especially important aspect to master for scientific conferences and poster sessions as your data needs to present itself at times when you may not be around to answer questions. So, how does one do this? Make a visual line through whatever kind of medium you are using to showcase your data and keep it consistent. Just like knowing where to start reading a book or an article, it should be evident where your story starts and ends. Overall, while you are the expert on the data, make sure it always has an active role in its presentation.

Mastering the fundamentals of data warehousing, including fact tables, dimension tables, joins, and cardinality, is essential for data visualisation. This strong foundation not only aids in overcoming career challenges but also enhances analytical capabilities. Strengthening these core concepts is paramount for success in the field of data analytics. 📊💡 #DataFoundations #AnalyticsSkills #CareerGrowth