Data Visualization in Environmental Public Health: From Data to Insight
Building a Foundation for Visualization These principles provide a foundation for all forms of visualization, from simple charts and reports to interactive dashboards and Geographic Information Systems (GIS) applications. Regardless of the tool or platform used, visualizations should communicate information clearly and quickly, support accurate interpretation, and remain useful to the intended audience for informed decisions. With these principles in mind, the next step is to understand the question, audience, and underlying data before selecting a visualization approach. BEFORE YOU VISUALIZE: START WITH THE QUESTION Start With the Question Strong visualization begins before a chart, map, or dashboard is created. It starts with the question: What does the audience need to understand, decide, or do? Once that need is clear, consider the available data and the visualization approach together. The data determine what you can accurately show, while the intended visualization can reveal what data you need, what is missing, or what requires additional preparation. The audience could initially describe a desired product rather than the underlying need. For example, a program manager may request a dashboard when the real need is to identify overdue inspections, compare workloads across districts, or monitor progress toward a performance target. Through discussion, the visualization developer can clarify the intended audience, key questions, required level of detail, update frequency, and measures of success. These conversations help ensure that the final product supports a meaningful operational or public health need rather than simply displaying available data. Connect the Visualization to the Data Once the need is understood, determine how the data will be prepared and managed to address it. This step involves identifying data sources, defining calculations, standardizing categories, establishing update processes, and determining how data will move from collection systems into reports, maps, or dashboards. Preparation includes extraction, transformation, and loading (ETL), which is the process of retrieving data from their original sources, preparing and standardizing them, and moving them into a system for analysis or visualization.
Additional data cleaning and modeling might also be needed, as discussed in more detail in the data modernization guide. The questions the visualization is intended to answer should guide these activities. Consider the data structure and visualization requirements together from the beginning. For example, a dashboard designed to show inspection timeliness could require consistent inspection dates, status fields, program categories, facility identifiers, and clear rules for determining whether an inspection is overdue. As the dashboard takes shape, gaps or inconsistencies in the underlying data can become apparent. Addressing these needs early reduces costly rework and produces information that better supports decision-making.
This planning should also determine whether the visualization will use counts, percentages, rates, or other standardized measures. Meaningful comparisons often require an appropriate denominator. For example, comparing the number of private wells with elevated arsenic levels across counties could be misleading if counties tested different numbers of wells. Reporting the percentage of tested wells that exceeded the health-based standard provides a more meaningful comparison because it accounts for both the number of wells exceeding the standard and the total number tested. Identifying the denominator early helps ensure that the necessary data are collected.
4
Powered by FlippingBook