|
||
|
Associative Forecasting Exercise
Flybai-Knight Industries (FKI) is a local firm that specializes in a variety of residential remodelling projects, and currently their best selling service is the installation of security systems. FKI is having a hard time staffing the firm, since demand for security systems exhibits a large amount of variability. They do, however, feel that recent burglaries can be used as a leading indicator for predicting demand. Demand data and burglary data for the past 12 months are given below. Plot each observed monthly demand versus the corresponding number of burglaries. Then, using regression (via SPSS, Excel or other software package), determine the straight line forecasting model for predicting demand as a function of previous burglaries. Use the Syx value to comment on the usefulness of this particular model. [ Warning: This is not a time series analysis; it is instead a forecasting of demand (dependent variable) as a function of some factor (in this case, burglaries). Hence, it is a correlation study or, in the terminology used in forecasting, an associative approach to demand forecasting. You already did this sort of thing when scatterplots and regression analysis were introduced; now you are to apply it to a forecasting situation. ]
| Month | Demand | Burglaries |
| November | 552 | 1328 |
| December | 198 | 1100 |
| January | 150 | 1772 |
| February | 251 | 1935 |
| March | 246 | 1562 |
| April | 243 | 593 |
| May | 73 | 800 |
| June | 80 | 2333 |
| July | 577 | 2280 |
| August | 350 | 2096 |
| September | 300 | 1461 |
| October | 198 | 1989 |
| November | 279 | 1688 |
| December |
Please submit any comments, corrections, etc. about this
document to John Seydel |