Speculation and dispatching has been a scheduling and budgetary issue. Multiple regressions are advised to be the answer. Assessment of regression coefficients allows a detailed comprehension of patterns, seasonality, and marketing influence.
Case with Lag
1. Build the regression model, starting with the independent variables in the file “harmon.xls”. Please include the regression output for your model in your report.
2. Argue that your model is good.
a. How does it compare with other possible models (if you include/remove some independent variables, etc.)?
b. Check whether the regression assumptions are satisfied (briefly summarize your conclusions from residual analysis).
3. For the next month, January 1988, the planned amount of Consumer Packs is 100 000 cases and Dealer Allowances are set at 500 000$. Use your model to predict sales in January 1988. Please report a point estimate and a 95% prediction interval.
4. Give a short interpretation for each of the regression coefficients in your model. Construct 80% confidence intervals for each coefficient.
5. If some extra budget becomes available, would you allocate it to Consumer Packs or to Dealer Allowances