個案總覽

Discusses what causation is and what one can (and cannot) learn about causation from observational (nonexperimental) data.

Provides an example of regression in one of its most important roles. Relating probabilistic forecasts based on past data to decision analysis.

Flanders is a catalog merchandiser. Various decisions on catalog distribution policy, ordering and inventory policy, and catalog format design are considered. This was a final examination, and serves as a review for a number of topics in the course.

L.L. Bean must make stocking decisions on thousands of items sold through its catalogs. In many cases, orders must be placed with vendors twelve or more weeks before a catalog lands on a customer's doorstep, and commitments cannot be changed thereafter. As a result, L.L. Bean suffers annual losses of over $20 million due to stockouts or liquidations of excess inventory. Provides a context in which buying decisions that balance costs of overstocking and understocking when demand is uncertain are made and implemented on a routine basis.

Develops the idea that the relevance of costs and revenues depends on what decisions are under consideration. Revenues and costs are relevant if they would be different under one decision choice than under the other. Teaching purpose: Basic background material for decision making under certainty and under uncertainty. Provides a framework under which a variety of costs and revenues, resulting from substitution, complementarity, capacity, etc., can be identified as relevant.

An introduction to sampling and statistical inference that covers the main concepts (confidence intervals, tests of statistical significance, choice of sample size) that are needed in making inferences about a population mean or percent. Includes discussion of problems of sampling in the real world where response bias and nonrepresentativeness violate the principles on which statistical inference is based.