Example (SAS program strtxmpl3.sas): using start values to fit a model with risk factors.

 

This example shows how to create starting values to fit a trajectory model that incorporates risk factors influencing the probability of group membership.

 

We use criminal offense data (0/1) from the Cambridge Study.

 

We start with a model with flat and cubic trajectories.

 


Step 1: Fit the trajectory model.

 

Three-group model, flat and cubic trajectories.

 

Model results.

 

Parameter estimates in the log for cut and paste.

 

 


Step 2: Create start values from the first model results.

 

We're going to add risk factors to the initial model.

 

The parameter estimates from the initial model are copied from the log.

 

Since this example uses risk factors, the first set of parameter estimates are used.

 

Start values need to be specified for the risk factors (zeros are ok.)

 

Three-group model, flat and cubic trajectories with risk parameters.

 

Final model results.

 

strtxmpl3.sas