Descriptive Statistics

The first step in any analytical process is to describe the data under consideration. Descriptive statistics are methods, such as the average, median, range, standard deviation, and so on, that summarize a dataset. Business intelligence and exploratory analysis are the most common uses of descriptive statistics.

Descriptive statistics are the rear-view mirror of management. Most performance reports consist of large volumes of tables and graphs that report on what happened in the previous reporting period. Business reporting is often viewed as a lowly aspect of data science because you need to look through the windshield to drive a car, and not just look through the rear-view mirror. More crudely, some say that traditional performance reporting is like counting the dead after the battle has been fought.

Although this analogy might seem a compelling argument to ditch the traditional business report, it does not negate the importance of a rear-view mirror. The car analogy...