Note from the Editor
This casebook provides a series of cases designed to enhance teaching the application of statistical tools to real business problems. The cases were selected to provide opportunities to apply statistical tools to real problems: often several tools can be usefully used for the same problem, depending on the sophistication of the user. The cases can be grouped by major statistical topic area (although there is a great deal of overlap).
Probability and Expected Value
PCB Manufacturing, Inc. presents a quality control example which appears simplistic but which raises a number of issues which provide an opportunity to explore probabilities in some depth. If students understand "likelihood functions" the analysis can proceed even more deeply.
Descriptive statistics
Direct Mail Company Inc. provides a set of data which, on careful statistical analysis (means and sequence), contain a terminal message for the firm. Bamberger's Department Store requires a good understanding of means and proportions. Firestone Canada Inc. - "Pay No Dough If It Doesn't Snow" provides the context to make informed use of means, standard deviations, and distributions (including the normal approximation). Ford Motor Company and Cruji Management Consulting (A) illustrates that a good understanding of descriptive statistics can often suggest directions for change that produce large cost savings.
Estimation
Four Star Fitness Club and Home Education, Inc. both require estimates of required financial reserves. Domglas Inc. is an example of estimation based on survey data and provides an opportunity to discuss surveys and survey methods, as well as the central limit theorem. The Morrice Collection (A) provides a good example where simple estimation techniques were used and had an impact.
Hypothesis testing
J. Walter-Thompson Company Ltd., and Warner-Lambert Canada Inc. is a complex example where designing tests appears as the major issue. The Ian Steele case can be analysed using a difference-of-means test.
Regression
Alfonso's Department Store tests both marketing basics and regression model interpretation. Ian Steele includes seasonality and indices, and provides an opportunity to build several regression models. The Morrice Collection case demonstrates the value of regression in sharpening estimates and providing confidence and prediction intervals. Brent-Harbridge Developments requires that decisions be made using expected values derived using probabilities taken from regression models. The objective in LanCo Catalogue Sales is to achieve an understanding of factors that effect returns and a number of analyses help achieve this understanding. Finally, the objective in the Northern Napa Valley Winery, Inc. case is to make a forecast of future business and a variety of statistical forecasting methods (including regression) yield useful looking projections.