11.1 Introduction

According to the SAS Institute, statistical analysis is “the science of collecting, exploring and presenting large amounts of data to discover underlying patterns and trends.”1 Statistical analysis is an important element of business intelligence and of making successful strategic decisions. In fact, we can define two primary types of statistical analysis: descriptive analysis and predictive analysis or modeling.

With descriptive statistical analysis, the objective is to discover and understand patterns and behaviors of existing business processes and operations. Another term used to describe predictive analysis is modeling. With predictive statistical analysis, we build models and test hypotheses to forecast future possibilities. With an understanding of existing patterns, strategic decisions can then be made based on the modeling results, which can influence future performance and turn possibilities into probabilities or desired outcomes.

There are many different approaches to performing statistical analysis of data. Depending both on the available data and on the desired intelligence, different techniques are used. One approach, as defined by TechTarget, consists of a five-step process:2

1. Describe and understand the data and its patterns

2. Explore the relationship between the data sample and the underlying population

3. Create appropriate models to explain the data and the population behavior

4. Test the validity of the models

5. Use predictive analytics (simulation) to forecast future outcomes

In the previous lesson, we explained the concepts relating to the first two steps in this process—understanding the data and exploring the relationship between samples and populations. We also began the third step and described many distinct probability distributions. In this lesson, we continue with these steps by explaining various types of models that enable the prediction of possible outcomes and their likelihoods or probabilities.