To run
a Generalized Linear Models analysis, from the menus choose:
Analyze > Generalized Linear Models > Generalized Linear Models...
Figure 1. Generalized Linear Models Type of Model tab
Select Custom as the type of
model.
Select Gamma as the response distribution.
Select Power as the link function and type -1 as the exponent of the power function.
This is an inverse link.
Click the Response tab. Figure 2. Generalized Linear
Models Response tab
On the Response tab, select Average
cost of claims as the dependent variable.
Select Number of claims as the
scale weight variable.
Click the Predictors tab. Figure 3. Predictors tab
On the Predictors
tab, select Policyholder age through Vehicle age (when variables are listed in
file order) as factors.
Click Options. Figure 4. Generalized Linear Models
Options dialog
In the Options dialog,
select Descending as the category
order for factors. This indicates that the first category of each
factor will be its reference category; the effect of this selection
on the model is in the interpretation of parameter estimates.
Click Continue, then
click the Model tab. Figure 5. Predictors
tab
On the Model
tab, select holderage through vehicleage (when variables are listed in file
order) as model terms.
Click the Estimation tab. Figure 6. Estimation tab
On the Estimation tab, select Pearson chi-square from the Scale Parameter
Method drop-down list in the Parameter Estimation group. This is the
method used by McCullagh and Nelder, so we follow it here in order
to replicate their results.
Click the EM Means tab. Figure 7. EM Means tab
Select holderage (Policyholder age) as a term to
display means for and select Repeated as the contrast.
Select vehiclegroup (Vehicle group) as a term to display means
for and select Pairwise as the
contrast.
Select vehicleage
(Vehicle age) as a term to display means for and select Repeated as the contrast.
Select Sequential Sidak as the
adjustment method.
Click the Save tab. Figure 8. Save tab
On the Save
tab, select Predicted value of linear predictor and Standardized deviance residual. These values are saved to the active dataset and can help you diagnose
any problems with the model fit.