Poisson Regression |
For this model and this set of data, there does not appear to be sufficient explanatory power in the YEAR*PERIOD effect to include it in the model.
Click on YEAR*PERIOD in the fit window. |
Choose Edit:Delete from the menu. |
Figure 17.8: Modified Fit Model
Follow the previous steps to remove the other two interaction terms from the model. The resulting main effects model is shown in Figure 17.9.
Figure 17.9: Main Effects Model
The estimate of the dispersion parameter suggests that overdispersion exists in the model. Type III (Wald) Tests table shows that all of the main effects are significant.
Copyright © 2007 by SAS Institute Inc., Cary, NC, USA. All rights reserved.