Summarizing a Large Quantity of Data: Tables and Charts
To yield useful information, a mass of data must first be summarized and reduced to understandable proportions. Imagine sending a large crew of interviewers to ask each family in the United States the size of its annual income. Each income could be recorded, perhaps together with the name of the family, on a 3 × 5 card, and the cards assembled in a central location. Some idea of the mass of this data can be had from the fact that 45 million 3 × 5 cards, one for each family in the United States, would solidly fill three rooms in a typical house.
Imagine this mountain of cards neatly stacked in a big warehouse. The warehouse contains all the information there is on the subject of United States family incomes, but the unorganized mountain of data is too vast to be informative. The first task of statistics is to reduce the mountain of raw data to a form that can yield useful information. This can be done in three ways: (1) The data can be reduced to a table; (2) they can be plotted as a chart; and (3) averages and other descriptive key numbers, called statistics, can be calculated. This chapter is devoted to the first two of these: the construction and use of frequency tables and charts.
The first step in the summarization of data is to reduce the mountain of cards to a frequency table showing the 45 million families classified by income brackets. The entire range of income is partitioned into a number of convenient subdivisions, and the families in each bracket counted.