Skip to main content

Data Visualization

What is Data Visualization?

Data visualization is a general term that describes any process that converts data sources into a visual representation such as charts, graphs, maps, and tables.  There are three main categories: 

  1. Scientific visualization: the visualization of three-dimensional phenomenon (architectural, meteorological, medical, etc.), emphasizing realistic renderings of volumes, surfaces, flows. Scientific Visualization. (n.d.). In Wikipedia. Retrieved January 16, 2018, from https://en.wikipedia.org/wiki/Scientific_visualization. 
  2. Information visualization: visualization of numerical and non-numerical data via charts, maps and other visual renderings to reinforce understanding.  Information Visualization. (n.d.). In Wikipedia. Retrieved January 16, 2018, from https://en.wikipedia.org/wiki/Information_visualization. For a comprehensive list of visualizations in this category go to the Data Visualization Catalogue. 
  3. Infographics: combine various statistics and visualizations with a story. Infographic. (n.d.). In Wikipedia. Retrieved January 16, 2018 from https://en.wikipedia.org/wiki/Infographic. 

Reasons to visualize my data

A good data visualization can uncover (exploratory visualization) and communicate (explanatory visualization) relationships and patterns otherwise hidden in your data.
  1. Picture superiority: 

A well designed visualization is easier to remember than a table of numbers.  The visual impact of the Income in Canada infographic is easier to recall than if one were to simple read numbers from a table.
 
Infographic that highlights the trends in the income data from the previous data table on Income from Census Canada.
Figure 1: Income in Canada Infographic to show how data can be presented in a visualization.
Statistics Canada. (2016). Income in Canada, 2016 Census of Population. Retrieved on January 16, 2018. 
Open Government Licence – Canada

2. Reveal patterns: 

 Anscombe’s Quartet shows how four datasets with identical summary statistics, when displayed on a scatter plot reveal four distinct patterns.  By visualizing the same sets of data it is clear that the data is in fact different. Some insights can only be discovered visually. 
Four scatter plots showing distinct patterns
Figure 2: Four datasets defined by Francis Anscombe for which some of the usual statistical properties (mean, variance, correlation and regression line) are the same, even though the datasets are different.
Shultz. (2010). Anscombes Quartet. Retrieved on January 16, 2018.
CC BY-SA 3.0

3. To add legitimacy or credibility:  

Research shows that data are more persuasive when shown in graphs even if they do not contain any new insights beyond what already exists. See Evergreen, Stephanie. Effective Data Visualization: The Right Chart for the Right Data. Los Angeles: Sage Publications, Inc, 2017.

4. To comprehend big data:  

They enable us to make sense of information that would be otherwise impossible to understand.  For example, See the visualization of flight patterns in the US by Aaron Koblin, part of the Celestial Mechanics project at UCLA.   
The paths of air traffic over North America visualized in color and form.
Figure 3: Screen capture of the paths of air traffic over North America.
Koblin, Aaron. (2015, July 2). Flight Patterns. Retrieved January 16, 2018.  
CC BY
 

Resources to Help with Data Research

Guide: Find data and StatisticsGuide: Choose the Best InfoVideo: Thinking Critically About Data

Suggest an edit to this guide

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.