| | CSIS 405 |
| | Chapter 4: Introuduction to Forecasting with Regression Trend Models |
| | 1. Homework: Exercises 8, 9, 10, 11, 14 and 18 |
| | | Exercise 4: |
| | | | a. t-ratio=coefficient/standard error |
| | | | The rule of thumb for significance testing is if the absolute value of t-ratio is greater than 2 |
| | | | then, the coefficient is statisically significant which means the coefficient is significantly different from zero. |
| | | | For example, t-ratio for "Constant" is 3.04 which is greater than 2. Therefore, the coefficient of 20,720 is |
| | | | significantly different from zero. |
| | | | b. It is R-squared right below the table |
| | | | c. Substitute 20 into EXP in the following equation |
| | | | Salary = 20,720 + 805 (EXP) for point estimate |
| | | | the approximate 95 percent confidence interval estimates would be |
| | | | point estimate ± 2 (SEE) |
| | | | SEE is the standard erro of the estimate which is right under R-squared value |
| | | Exercise 10: |
| | | | a. The equation is |
| | | | Booking = b0 + b1(Income) |
| | | | Booking is the dependent variable and Income is the independent variable |
| | | To obtian the regression eq;uation using Forecast X: |
| | | | First, highlight the data for "Bookings" and "Income." Do not highlight "Location" |
| | Open Forecast X > Forecast Method > Choose "Multiple Regression" under "Forecast Technique" > Report > Audit>Finish |
| | | Exercise 18: |
| | | | a.For trend forecat, open Forecast X > Forecast Method > Choose "Trend (Linear) Regression" |
| | | | b. To reseasonalize the data, time the value with the corresponding seasonal index |
| | | | For example, for Jan-17, actual shoe store sales SSS=SASSS x 0.74 = 2,2422 x 0.74=1,792 |
Zola_17
Hints_for_Exercise_41.xls
