A Brief History of Data Visualization
Data visualization has come a long way. From simple cave drawings depicting the success of a hunt, to the intricate dashboards we have today connecting to millions of rows of data.It’s always good to know where you came from in order to better understand where you are now and where you are going. Michael Friendly, from York University, published a paper on the milestones in the history of Data Visualization, Milestones in the History of Data Visualization: A Case Study in Statistical Historiography. The paper provides an excellent breakdown of major points that have lead us to modern day data visualization.Prior to the 17th century, data visualization existed mainly in the realm of maps, displaying land markers, cities, roads, and resources. As the demand grew for more accurate mapping and physical measurement, better visualizations were needed.In 1644, Michael Florent Van Langren, a Flemmish astronomer, is believed to have provided the first visual representation of statistical data. The one dimensional line graph below shows the twelve known estimates at the time of the difference in longitude between Toledo and Rome as well as the name of each astronomer who provided the estimate. What is notable here is that while Van Langren could have provided this information in a table, it is the use of the graph that really visually displays the wide variations in estimates.The 18th century saw the beginning of thematic mapping. Attempts at thematic mapping of geologic, economic, and medical data were made near the end of the century. Abstract graphs of functions, measurement error, and collection of empirical data were introduced at this time.This period also gave us William Playfair, who's widely considered to be the inventor of many of the most popular graphs we use today (line, bar, circle, and pie charts). Many statistical chart types, including histograms, time series plots, contour plots, scatterplots, and others were invented during this period. A graph by Playfair (1821), shown below, shows the price of wheat, weekly wages, and reigning monarch over a two hundred fifty year span from 1565 to 1820.
A number of factors contributed to this “Golden Age” of statistical graphing: the industrial revolution, which created the modern business; official government statistical offices, to support an increasingly aware and global populace; and a growing recognition for the importance of numerical data in social planning, medicine, military, industrialization, commerce, and transportation. Statistical Theory also provided the means to make sense of large datasets.The growing trend towards statistical visualizations hit a small roadblock in the early 20th century. Friendly describes this era as the modern dark ages for data visualization. Statisticians were increasingly concerned with exact numbers, and considered images to be overly inaccurate. While innovation in the field may have indeed veered away from data visualizations, this period saw a growth of data visualization in the public consciousness. Charts and graphs of various sorts were rapidly becoming adopted into text books, business applications, science, and government.[Editor’s Note: While statisticians gave data visualization the cold shoulder during this era, the first half of the 20th century also brought psychology forward as a science; in particular, the development of cognitive psychology and the study of human perception has provided a much better understanding as to how the brain interprets information and recognizes patterns. This research has been instrumental in developing and refining the science of data visualization best practices.] The latter half of the 20th century is what Friendly calls the ‘rebirth of data visualization’, brought on by the emergence of computer processing. Computers gave statisticians the ability to collect and store data in increasingly larger volumes, as well as the ability to visualize the information quickly and easily. The 1960s and 1970s saw the emergence of researchers like John W. Tukey in the United States and Jacques Bertin in France, who developed the science of information visualization in the areas of statistics and cartography, respectively.. The early 80s saw the emergence of Edward Tufte, whose seminal work, The Visual Display of Quantitative Information is still used today in university courses for data visualization and statistical analysis. Tufte also introduced us to the sparkline, which gives the general shape of a trend in a small amount of space.