Carnegie Mellon University - Department of Statistics
Measuring the size of a population is an old statistical problem but methods for multiple lists or samples have taken on a new focus in the past decade, allowing for dependence among lists and heterogeneity. This talk presents an overview of the emerging toolkit of methods, both Bayesian and frequentist, that have found application in diverse areas. Examples in this presentation will include: the population of the US (sampling and the census), the number of Italians with diabetes, and the size of the World Wide Web.
Dr. Fienberg is president of the Institute of Mathematical Statistics, and author of "Intelligence, Genes, and Success: Scientists Respond to 'The Bell Curve'," and "Who Counts? The Politics of Census Taking in Contemporary America." His research focuses on the use of statistics in public policy and the law, and the role of statistical methods in census taking.