pathdoc - stock.adobe.com

Tip

8 steps to improve data visualization literacy

Data visualization literacy is a crucial element of analytics -- it helps you communicate findings. These eight steps can help you improve your DVL.

Creating and understanding visualizations that communicate data and analysis insights requires data visualization literacy.

Visualizations provide an effective tool for communicating patterns and trends in massive data sets. They help users grasp complex concepts and relationships through visual cues. To get the most out of your data, you need the necessary data visualization literacy (DVL) skills to create effective visualizations.

DVL refers to the set of skills necessary to understand, evaluate and create visualizations that clearly and effectively communicate data insights. Some resources define DVL as simply the ability to interpret visualizations, but visualization literacy is much more inclusive. It incorporates comprehension and design, like how language literacy refers to both reading and writing.

An organization that supports and nurtures a comprehensive DVL effort can benefit from improved communication and better decision-making. DVL enhances and promotes the critical appraisal of information consumption. Users can rely on more accurate, concise and meaningful visualizations that use data to tell a story. Visualization designers with strong DVL skills understand how to incorporate data elements effectively to communicate that story.

DVL is sometimes incorrectly treated like data literacy. Data literacy is concerned with the data itself. It considers factors such as how to collect, store and transform data, while keeping in mind the organizational need for effective decision-making. Data literacy makes it possible to understand and work with the data in any form and use it when communicating with others.

In contrast, DVL centers on how to effectively present the data. It focuses on choosing chart types, labels, colors and other visual elements that help tell a story. Although DVL requires an understanding of many of the same factors as data literacy, it is much narrower in scope, with a focus on how to translate data to visual images.

If you want to improve your visualization literacy, you should identify the gaps in your DVL skills. Assess how well you and your team can understand, analyze and create visualizations that effectively communicate salient information. Not everyone needs to be a data scientist to be visually literate, but they should all possess the skills necessary to use data in their jobs.

Evaluating DVL skills needs to be an agreed-upon standard of what the skills should be. For example, it might be a certain level of competency in statistics, data cleansing or design layout. Your organization should then invest in the training and education necessary to achieve the DVL level each role needs.

How to improve data visualization literacy

Not everyone starts at the same level of visualization literacy. Some users are more comfortable working with data and images than others. Others might have better knowledge of audiences and their needs. It takes a mix of technical and communication skills to create the best data visualizations. You should consider the following seven guidelines to enhance your DVL skills.

1. Build a foundation in data literacy

If you want to improve your DVL skills, you should have a strong foundation in data literacy. You can use data literacy skills to explore data, derive meaningful information and communicate that information to others. Data literacy skills are often regarded as highly technical and associated with data scientists. However, they can encompass a wide range of technical and nontechnical skills:

  • Analyze data and use statistical methodologies.
  • Interpret and create data visualizations.
  • Familiarity with data analytics tools and techniques.
  • Familiarity with programming languages.
  • Wrangle and manipulate data.
  • Communicate data and storytelling.
  • Apply critical thinking to data-driven information.
  • Understand how data is structured, collected and stored.

Not everyone has or needs every skill. You don't need to possess the same level of ability for a particular skill as someone else. For example, some people are adept at creating a data visualization, but most should be able to understand its message.

2. Improve your visualization skills

DVL requires knowing how to create, interpret and critique visualizations. Visualization skills are a core component of data literacy and, by extension, visualization literacy. Building effective visualizations requires seven skills:

  1. Statistical data analysis. Possesses a strong foundation in statistics, data analytics and their underlying mathematical principles.
  2. Storytelling. Capable of using design elements to demonstrate why the data is important and what insights it reveals.
  3. Visualization tools. Proficient in data visualization tools, including one or two of the leading products, such as Tableau and Microsoft Power BI.
  4. Visual design. Strong graphic design skills help create intuitive, informative and aesthetically appealing visualizations.
  5. Collaboration/communication. Capable of creating visualizations that communicate core concepts and establish a sense of collaboration with others.
  6. Critical thinking. Approaches each visualization project critically and methodically, regardless of the underlying data ecosystem.
  7. Feedback, evaluation, improvement. Uses feedback from others to refine and improve visualizations on a continuous basis.

You should have a good understanding of visualization basics, such as knowing how to incorporate colors and labels into their designs and which visualization types to use in specific situations. You should also study existing visualizations to better understand which ones most effectively communicate the essential concepts they're trying to convey.

3. Identify visualization biases

When creating visualizations, you might introduce biases that affect how users interpret the information. To help prevent this, you should learn to identify biases in other visualizations so you can recognize and mitigate your own biases.

A visualization built on biased data is bias. For example, the source data might not be representative of the larger population. Consider a visualization that attempts to show the number of people in the general population who eat dessert. If the data comes from randomly surveying individuals as they come out of ice cream parlors, the responses will likely skew in that direction.

Bias can also occur because of design choices. For example, a line chart might provide statistics on two unrelated events, such as monthly influenza rates and average monthly stock prices. People who view the visualization might assume a causal relationship between the two and draw the conclusion that flu rates affect stock prices, or vice versa. You can also introduce biases by manipulating a chart's scaling or choosing colors that lead to inaccurate assumptions, which influence how the audience interprets the data.

You must proactively eliminate any biases. You should ensure that the data is complete and representative, and be open about how you source, collect and process data. You should communicate any concerns about the data to the audience and point out the potential for bias. You should also understand design principles so that graphics, colors, labels and other visual elements don't promote bias.

4. Understand your audience

An important component of DVL is the ability to visually communicate information about the underlying data. To do so effectively, you must fully understand your audience and what information they need to clearly understand the underlying data and derive actionable insights.

You should learn as much as you can about who your audiences are, how they will interact with the visualizations and how they will use the available data. Keep in mind that people learn and absorb information differently. When possible, tailor visualizations to the people who will use them. For example, a visualization for an audience of data analysts might be quite different than one for an audience of product consumers.

Audiences can vary significantly in terms of their domain knowledge, level of data literacy and ability to comprehend different types of visualizations. People with strong data literacy skills can typically handle more sophisticated visualizations. The average user might require simpler visualizations with accompanying explanations.

Audiences can also vary in terms of how much time they can spend interacting with the data. For example, some users might want to drill more deeply into the underlying data, while others might prefer basic visualizations that provide snapshots of the essential information.

5. Understand the business context

When creating visualizations, you should have a good sense of the business context that drives your design projects. The business context provides a big-picture perspective of the organization's goals and objectives, along with the environment in which the organization operates. You can create visualizations that align with the organization's priorities and concerns, while supporting stakeholder initiatives and informed decision-making.

Your visualization should reflect the business context whether it appears in reports, presentations, web pages, dashboards or other places. It should provide stakeholders and decision-makers with the information they need when they need it, so they can quickly grasp current trends, the status of their operations and the effectiveness of their business strategies.

Business context is directly related to understanding the audience; the audience determines the context. For example, IT managers might focus on performance anomalies, while sales managers might want to know the success of their marketing campaigns. In both cases, the managers need information within a specific business context, so they can make the necessary adjustments to their strategies and plan future strategies.

6. Evaluate your data sources

The quality of the data used in a visualization determines its reliability and value. Only complete, accurate and relevant data can create a quality visualization. You should know where the data comes from and how it was prepared. You should also be familiar with the different types of data and which types best serve your visualization.

The currency, relevance, authority, accuracy and purpose (CRAAP) test is one popular tool for evaluating data quality. It's a framework for determining the value of the source data, based on five criteria:

  • Currency. Timeliness of the information.
  • Relevance. Importance of the information for the specific needs.
  • Authority. Source of the information.
  • Accuracy. Reliability, truthfulness and correctness of the content.
  • Purpose. Reason the information exists.

You should know if the data is complete and representative and recognize potential outliers. The goal is to work with data sets that do not have duplications, inconsistencies or biases.

7. Implement a feedback loop

A good way to improve your DVL skills is to seek out feedback from stakeholders and members of the target audience and other designers. Feedback helps you improve your current visualizations and create future visualizations that better serve your target audience. Feedback helps you better understand what works and what doesn't, which can lead to better storytelling, greater clarity and more effective designs.

In some cases, you can solicit feedback early in the development process by creating prototypes that stakeholders can review. One approach is to use rapid prototyping; create visual mockups that employ different design strategies and gathers input on which works best.

Feedback is useful at any stage in the visualization lifecycle, even when the visualizations are in production. You should think of a visualization project as an iterative process that continuously improves based on ongoing feedback.

You might also seek out feedback through collaboration to gain new ideas and perspectives. You also receive immediate feedback on your designs, which can help you learn new techniques and improve your skills. Collaboration is particularly useful when working with designers who have extensive experience in creating visualizations.

8. Create a data visualization style guide

Organizations can support data visualization literacy by developing a data visualization style guide. A style guide provides a set of standards and guidelines for creating visualizations across the organization, like how many organizations create editorial style guides. A visualization style guide might include details about labeling, color usage, font types, line thickness and chart types.

A visualization style guide should fit neatly with the organization's branding and editorial guidelines. It should help you create visualizations more efficiently and provide visual consistency across the organization. The style guide should enforce a common visual language that everyone in the organization uses. Visual consistency can help audiences grasp complex concepts more quickly.

Putting together a visualization style guide can be a significant undertaking. Review style guides created by other organizations, many of which are available online. One popular style guide example is available through the Urban Institute, which publishes the guide to benefit nonprofit and research organizations, as well as the data visualization community. It provides a good example of the types of information that an organization might include in its own style guide.

Robert Sheldon is a freelance technology writer. He has written numerous books, articles and training materials on a wide range of topics, including big data, generative AI, 5D memory crystals, the dark web and the 11th dimension.

Dig Deeper on Data visualization