International Operations Management
Download 340.14 Kb. View original pdf
|
final-questionnaires
| Exhibit 11.5 Framework for Forecasting and Planning Monitor outcomes Implement plans Yes No Forecasts Forecasting methods Plans Planning process Data bank Environment Are forecasted results satisfactory ? Source: J. Scott Armstrong (2001), Principle of Forecasting—A Handbook for Researchers and Practitioners. Kluwer Academic Publishers, p. period over which a forecast will be made is known as the time horizon of the forecast. Depending on their time horizon, forecasts can be Short term (up to one year ahead Medium term (from one to three years ahead Long term (for more than three years ahead). A large number of forecasting methods exist today and the advent of computers has greatly facilitated their acceptance by management. The further development of forecasting methods and their practical application has enabled not only forecasting experts but also managers to understand and use these methods. 16 The main classification of forecasting methods is between statistical and judgmental. Statistical forecasting methods are based on measurement. In other words, they concern data that either exist or can be obtained through experimentation. These methods vary from a simple extension of the past into the future to more complicated and elaborate procedures. The advent of computers has greatly facilitated the acceptance of statistical methods in business. Judgmental forecasting methods, on the other hand, are based on personal judgment and/or expert opinion. Because these methods do not use any numerical data, they are highly subjective. Generally speaking, judgmental forecasting methods are used more for longer-term forecasts, whereas statistical forecasting methods are used more for intermediate and shorter-term forecasts. ............................................................................. ............................................................................. Elsevier US Job code KIB Chapter: Ch11-H7983 6-12-2006 9:22 p.m. Page:372 Trimsize:7.25 in in Fonts used Sabon & Frutiger Margins:Top:36 pt Gutter:66 pt Font Size pt Text Width PC Depth:43 Lines 372 International Business When we forecast we should always keep in mind that the future is by definition unpredictable. No forecasting method can therefore guarantee that it will produce perfectly accurate forecasts. Even the most accurate forecasts will never be 100% accurate. A good forecast is therefore one that will be close enough to what will actually happen. The results of a classic competition (known as the M-competition) regarding the accuracy of a large number of forecasting methods when applied to different time series data sets across different industries showed that simple forecasting approaches, such as exponential smoothing, produced accurate results. Regression methods, on the other hand, were not found to be as robust. 17 Although no reasonable forecaster would identify the best method from the various forecasting competitions and adopt that method for his or her specific forecasting problem, competitions on the accuracy of different forecasting methods have been helpful at improving the actual practice of forecasting in industry. 1819 The following example illustrates how the method of exponential smoothing—a simple numerical approach to forecasting—can be used to help managers make better decisions regarding the operations of their organization. Example 1 A small Japanese manufacturer of electrical appliances exports its products to a number of European destinations. The operations manager of the company wants to predict the number of component parts needed for the production of a particular model. The following data have been collected over a six-week period: Week Download 340.14 Kb. Share with your friends:
|