Sampling, Monte Carlo and Problem Solving: How Analyzing Statistics Helps us Improve  –  Edward Rothman
Political candidates drop out of elections for the U.S. Senate and New York Governorship because their poll numbers are low, while Congress fights over whether and how much statistics can be used to establish the meaning of the U.S. Census for a host of apportionment purposes influencing all our lives. Suppose we need to learn reliably how many people have a certain disease? How safe is a new drug? How much air pollution is given off by industry in a certain area? Or even how many works someone knows? For any of these problems we need to develop ways of figuring out how many or how much of something will be found without actually tallying results from a large portion of the population, which means we have to understand how to count this by sampling. And Monte Carlo? Sorry, no field trips to the famous casino! But, Monte Carlo refers to a way of setting up random experiments to which one can compare real data: does the data tell us anything significant, or it is just random noise? Did you know you can even calculate the number ? This way?!?!The survey and sampling techniques that allow us to draw meaningful conclusions about the whole, but based on the analysis of a small part, will be the main subject of this course.