CNTSELECT Procedure
Displayed Output
PROC CNTSELECT produces the following displayed output.
Model Fit Summary
The "Model Fit Summary" table contains the following information:
dependent (count) variable name
number of observations used
number of missing values in data table, if any
data table name
type of model that was fit
error components model type for panel data
group id variable for panel data
number of cross sections for panel data
parameterization for the Conway-Maxwell-Poisson model
frequency variable name, if any
weight variable name, if any
offset variable name, if any
zero-inflated link function, if any
zero-inflated offset variable name, if any
log-likelihood value at solution
maximum absolute gradient at solution
number of iterations
optimization technique used
AIC value at solution (smaller value indicates better fit)
SBC value at solution (smaller value indicates better fit)
covariance matrix estimation method used
A line below the "Model Fit Summary" table indicates whether the algorithm successfully converged.
Parameter Estimates
The "Parameter Estimates" table gives the estimates of the model parameters. In zero-inflated (ZI) models, estimates are also given for the ZI intercept and ZI regressor parameters, which are labeled with the prefix "Inf_". For example, the ZI intercept is labeled "Inf_intercept". If you specify "Age" as a ZI regressor, then the "Parameter Estimates" table labels the corresponding parameter estimate "Inf_Age". If you do not list any ZI regressors, then only the ZI intercept term is estimated.
If the DISPMODEL statement is specified for the Conway-Maxwell-Poisson model, the estimates are given for the dispersion intercept, and parameters are labeled with the prefix "Dsp_". For example, the dispersion model intercept is labeled "Dsp_Intercept". If you specify "Education" as a dispersion model regressor, then the "Parameter Estimates" table labels the corresponding parameter estimate "Dsp_Education". If you do not list any dispersion regressors, then only the dispersion intercept is estimated.
"_Alpha" is the negative binomial dispersion parameter. The t statistic that is given for "_Alpha" is a test of overdispersion.
Covariance of Parameter Estimates
If you specify the COVB option in the PROC CNTSELECT or MODEL statement, the CNTSELECT procedure displays the estimated covariance matrix, which is defined as the inverse of the information matrix at the final iteration.
Correlation of Parameter Estimates
If you specify the CORRB option in the PROC CNTSELECT or MODEL statement, the CNTSELECT procedure displays the estimated correlation matrix, which is based on the Hessian matrix used at the final iteration.