Often the most effective way to describe, explore, and summarize a set of numbers–even a very large set–is to look at pictures of those numbers.
Graphics visually display data by using points, lines, a coordinate system, numbers, symbols, words, shading, and color. The aim of data visualization is to establish the main features of the data and to guide in choosing the appropriate statistical techniques. Furthermore, visualization procedures help in spotting errors and unusual values in the data allowing corrective steps to be taken early in the analysis.
Theorists like Edward Tufte have defined the requirements for the attainment of excellence in statistical graphics. He says graphical displays should:
- show the data;induce the viewer to think about the substance rather than about methodology, graphic design, or the technology of graphic production;avoid distorting what the data have to say;present many numbers in a small space;make large data sets coherent;encourage the eye to compare different pieces of data;reveal the data at several levels of detail, from a broad overview to the fine structure;serve a reasonably clear purpose: description, exploration, tabulation, or decoration;be closely integrated with the statistical and verbal descriptions of a data set.
Until recently, graph design for data analysis and presentation has been largely unscientific. William Cleveland and Robert McGill applied theory to graphical perception with their study which examined the relative accuracy with which various graphical forms conveyed quantitative information. Their results showed that when graphics are “read,” position judgments about certain graphic elements such as bars in a bar chart and lines on a line chart were more accurate than length and angle judgments such as those made when viewing pieces of a pie in a pie chart.
Their conclusions called for a dismissal of some of the more popular chart forms–bar charts, divided bar charts, pie charts, and statistical maps with shading–in favor of dot charts, dot charts with grouping, and framed-rectangle charts.
Cleveland explains the vital link between graph and visual perception. When a graph is made, quantitative and categorical information is encoded, then the information is visually perceived. It is neither the information itself, the technological displays, nor the sophistication of the statistical analysis that allows information to be transferred by the a visualization, for if perceptions fail–the graph fails. It is the graphical method.
Some graphical methods lead to efficient, accurate transmission, while others lead to inefficient, inaccurate perceptions of the information. Through understanding of visual perception, informed judgments can be made about display methods.