Positive Approaches Journal, Volume 9, Issue 3

1. Similar to the five principles of Gestalt,⁴ quality visualizations should contain clear explanations of data elements and labels, carefully selected icons/graphics, and appropriate placement of graphs and charts. It should be packaged in a way that makes it easy for the reader to interpret without much additional explanation. Poorly designed visualizations filled with too much clutter, unclear labeling, or unique representations of the data unfamiliar to the reader will not have great impact.

2. Often, there are multiple ways to represent the same data to allow you the opportunity to interpret the information in different ways. Check out the Gallery on Makeover Monday⁵ to see how others take one data source and create multiple visualizations from their unique perspectives.

1. To read a visualization you need the right technology. Programs including Microsoft Excel or Tableau Reader⁶ provide the reader the ability to manipulate the information from a variety of devices. Many of these programs also have mobile interfaces and can be embedded into social media with just a click of a button.

2. Read a visualization like a book. Start at the top and move left to right. Pay attention to the title and look for captions or summary statements near each visualization; Icons or other tool tips,⁷ including legends, axes, and directions from the author may also provide you with additional guidance for how to interact with the data.

3. Hover, click, and explore your way through each element in the visualization. Often what you see on the surface is only the beginning. Authors may add icons or arrows to tell users where and how to click and hover. For example, the World Shape visualization,⁸ allows you to compare countries to the world by manually dragging the slider across years or pushing play to automatically view the change over time. This visualization also pulls specific facts about each country you select within the same visualization.

4. Filters are a feature within the visualization software that are designed to narrow your focus on specific elements of a visualization that are of interest to you. Use filters to compare elements (e.g. dates, counties) to an entire data set or to sort the data (e.g. population, sales) in a different way. Tableau Public provides many examples of visualizations with a variety of filters. One dashboard in particular highlights Pennsylvania Home and Community-Based Services⁹ where several filters including service category, fiscal year, county, and enrolled service users can be applied to the graphs within the visualization, resulting in immediate change as you explore the data.

5. Pay attention to color schemes as they may provide you with critical information to draw your attention. Eva Murray¹⁰ explains the importance of color in data visualizations to create an impact through our emotions and to create different associations with the information being shared. As you apply different filters, notice how colors change the view.

6. For additional tips on how to understand and interpret a visualization, check out Diana Aleman’s¹¹ six tips on reading a visualization, where she breaks down the components to help visualizations feel less daunting and intimidating.

1. Share with others! Creating a visualization using the best practices described above is only the first step. Dissemination is key to encourage others to explore the information. While some visualizations can be emailed, others may require specific software. For example, if you receive a visualization created in Tableau Desktop, you will need Tableau Reader⁶, on your device. Tableau Reader is a free application that allows you to present views as a slideshow, filter, drill down, export or print views of the data, and/or publish as PDF file.¹²

2. Data should make you ask more questions. Reach out to the author or creator of the visualization with questions or takeaways. It may open up conversations to explore the data in other ways.

3. Check back frequently. While some visualizations are an extract, or a one-time snapshot of data, many visualizations are updated at regular intervals. Be sure to check back on the visualizations to see the change over time to the visualization.