Messing Around Systematically - The Science of Experimentation

Most of us have tried microwaved popcorns (e.g. those sold by the Boy Scouts) at one time or another. The package would normally include some general instructions on the power level to set and time to cook to get the best results. If you read the small prints, the instructions will also say that the time and power level given are just a guide and some experimenting is required to get the best result for your particular oven. Obviously, if you are crazy about microwaved popcorns, you would like to carry out the experiment to see how you may get the best tasting popcorn and perhaps have as few "bullets" of unpopped popcorns as possible with your microwave oven. While you are at it, you might also want to experiment with different brands of popcorns as well. So you have three factors to consider (Time, Power Level, and Brand) and two outcomes of interest (Taste and Number of Bullets). What is the best way to carry out this experiment so that the optimum combination of the factors can be determined? What is the scientific and systematic approach? What approach would require the fewest runs yet give us the most information? In this talk, various approaches of experimentation will be discussed and the correct, scientific, and most efficient approach will be revealed. The approach to be presented has been in existence for decades and has considerable record of success over the last 50 years especially in the manufacturing and process industries. It has been found to allow cost reduction in product development, increase efficiency of experimentation, and often reveal the essential nature of a process. The technique can also be used for efficient conduct of computer experiments for parametric analysis, sensitivity analysis, and for simplification of complex equations and models aside from popping popcorns! The talk will be illustrated using several "toys" and it will be of interest to engineers of all disciplines who has an interest in experimentation.

Biography - Dr. Leonard M. Lye

Dr. Lye is Professor and Past Chair of Civil Engineering at Memorial University of Newfoundland. He holds a first-class honours degree in civil engineering and a PhD specializing in the areas of statistical and stochastic hydrology. His current areas of interests include hydrology and hydraulics, environmental statistics, and the design and analysis of experiments with particular applications in civil engineering. Dr. Lye is a highly rated and entertaining teacher was the winner of the Association of Professional Engineers and Geoscientists of Newfoundland award for excellence in teaching for 2001.

Further details on this workshop will be posted as they become available.

---

Last Updated February 23, 2002 1:46:27 PM