Introduction to Statistics: Inference

Statistics 2 at Berkeley is an introductory class taken by about 1,000 students each year. Stat2.3x is the last in a sequence of three courses that make up Stat2x, the online equivalent of Berkeley's Stat 2. The focus of Stat2.3x is on statistical inference: how to make valid conclusions based on data from random samples. At the heart of the main problem addressed by the course will be a population (which you can imagine for now as a set of people) connected with which there is a numerical quantity of interest (which you can imagine for now as the average number of MOOCs the people have taken). If you could talk to each member of the population, you could calculate that number exactly. But what if the population is so large that your resources will not stretch to interviewing every member? What if you can only reach a subset of the population?

Stat 2.3x will discuss good ways to select the subset (yes, at random); how to estimate the numerical quantity of interest, based on what you see in your sample; and ways to test hypotheses about numerical or probabilistic aspects of the problem.

The methods that will be covered are among the most commonly used of all statistical techniques. If you have ever read an article that claimed, "The margin of error in such surveys is about three percentage points," or, "Researchers at the University of California at Berkeley have discovered a highly significant link between ...," then you should expect that by the end of Stat 2.3x you will have a pretty good idea of what that means. Examples will range all the way from a little girl's school science project (seriously – she did a great job and her results were published in a major journal) to rulings by the U.S. Supreme Court.

The fundamental approach of the series was provided in the description of Stat2.1x and appears here again: There will be no mindless memorization of formulas and methods. Throughout the course, the emphasis will be on understanding the reasoning behind the calculations, the assumptions under which they are valid, and the correct interpretation of results.