Think Stats, 2nd Edition

Chapter 14. Analytic Methods

This book has focused on computational methods like simulation and resampling, but some of the problems we solved have analytic solutions that can be much faster.

I present some of these methods in this chapter, and explain how they work. At the end of the chapter, I make suggestions for integrating computational and analytic methods for exploratory data analysis.

The code in this chapter is in normal.py. For information about downloading and working with this code, see Using the Code.

Normal Distributions

As a motivating example, let’s review the problem from Sampling Distributions:

Suppose you are a scientist studying gorillas in a wildlife preserve. Having weighed 9 gorillas, you find sample mean kg and sample standard deviation, kg. If you use to estimate the population mean, what is the standard error of the estimate?

To answer that question, we need the sampling distribution of . In Sampling Distributions we approximated this distribution by simulating the experiment (weighing 9 gorillas), computing for each simulated experiment, and accumulating the distribution ...