- Real, often very messy, data
- Using the popular R programming language
- Both data analysis and mathematical exercises
- Modern principles of model selection and model building, such as information criteria and data splitting, and hypothesis testing, as well as a complete approach to model selection, independent of the criteria
- Graphical displays of the data

A few years ago I was asked to teach a class on time series. The students in the class would be undergraduate statistics majors, graduate mathematics majors, and graduate students from a variety of science departments around the campus. My major concern in the course was that by the end of the course, the students be able to actually analyze data. In fact, I am of the belief that any applied statistics course, upon completion, in which the best students cannot independently analyze data, is a waste of the student's time and money.

Although this sounds like a reasonable goal, it is surprising how inadequate most time series books look, given this task. Most time series books fall into two categories: books about mathematical models with data sets used to illustrate a mathematical idea, and books on forecasting with recipes for making short-term predictions.

The only presentation close to what I wanted was Chapter 15 of *The Statistical Sleuth* (2002) by Ramsey and Schafer. However, this chapter is limited to what we will come to denote the AR(1) model and lacks the mathematical details ...

Start Free Trial

No credit card required