Statistics in a Nutshell, 2nd Edition

Chapter 4. Descriptive Statistics and Graphic Displays

Most of this book, as is the case with most statistics books, is concerned with statistical inference, meaning the practice of drawing conclusions about a population by using statistics calculated on a sample. However, another type of statistics is the concern of this chapter: descriptive statistics, meaning the use of statistical and graphic techniques to present information about the data set being studied. Nearly everyone involved in statistical work works with both types of statistics, and often, computing descriptive statistics is a preliminary step in what will ultimately be an inferential statistical analysis. In particular, it is a common practice to begin an analysis by examining graphical displays of a data set and to compute some basic descriptive statistics to get a better sense of the data to be analyzed. You can never be too familiar with your data, and time spent examining it is nearly always time well spent. Descriptive statistics and graphic displays can also be the final product of a statistical analysis. For instance, a business might want to monitor sales volumes for different locations or different sales personnel and wish to present that information using graphics, without any desire to use that information to make inferences (for instance, about other locations or other years) using the data collected.

Populations and Samples

The same data set may be considered as either a population or a sample, depending ...