What is data visualization?

Data visualization is the process of representing information using visual means, such as a chart/graph, diagram, map, or picture. There are two fundamental types of data visualization: exploratory visualization and explanatory visualization.

Exploratory data visualization is an essential part of exploratory data analysis: the process of summarizing and analyzing data to understand how it is organized, make transformations where necessary, and explore potential underlying patterns in the data. Exploratory data visualization is an important step in trying to make sense of your data by visualizing trends and relationships that may be hidden by summary statistics (the classic example of this is Anscombe's Quartet). In other words: "Exploratory analysis is the process of turning over 100 rocks to find perhaps 1 or 2 precious gemstones." (Cole Nussbaumer Knaflic 2014, Storytelling with Data).

Explanatory data visualization, on the other hand, is the process of telling a story about a particular aspect of your data. Whereas an exploratory visualization has the primary purpose of helping you understand your own data, an explanatory visualization communicates what you have determined is important about your data to an audience. An explanatory visualization can, for example, take the form of a graph with only a subset of the data highlighted or an infographic that combines one or more graphs, tables, or images with text that frames your argument and tells a more complete story than a single chart alone.

The following steps represent the general process you might take in order to create a visualization.*

1.  Know your data. It is helpful to start with data that is clean (read more about what it means for data to be "clean" in the article "Tidy Data" by Hadley Wickham, 2014). If you are not the original creator of the dataset you want to visualize, you should also start by making sure you understand what the variables in the dataset mean and ensure that the data has secure provenance (i.e. you know where it came from and how it was created).
    
2. Determine your purpose. Decide whether you are in the exploratory or explanatory stage of visualizing your data. Are you trying to find patterns, or are you trying to tell a story or support an argument with your visualization? Understanding your purpose can help you decide what data to include in your final visualization.
    
3.  Choose a chart type. Different charts are useful for different kinds of data, representing different numbers of variables, or showing different patterns and relationships among those variables. Check out the Best Practices page on this guide for interactive resources that can help you decide which chart is most useful for your purpose (Step 2) and your particular data (Step 1).
    
4. Decide on a visualization tool. There are many options available to help you create your visualization, from point-and-click online tools to desktop software or programming solutions using R, Python, or JavaScript. The Tools page on this guide provides an introduction to some of the options based on the type of data you are visualizing, noting which platforms are free/open-source and which are available online.
    
5. Refine your visualization. Once you have selected a tool and input your data, you can adjust some of the textual and visual aspects of your chart to make it more readable, by tweaking the chart axes, labels, or color palette. You should consider how your visualization will appear to people who have visual impairment or different kinds of colorblindness and try to make your visualization as understandable and accessible as possible, whether it will appear in print or on a screen. Some accessibility guidelines related to font and color can be found on the Best Practices page on this guide.
   
   * Adapted from the Data Visualization guide at GMU University Libraries.