See also: Hypothesis Tests and Confidence Bounds
Computes Wald-type confidence bounds for all parameter in the current model. The confidence intervals are computed as
where
zα is the appropriate normal deviate for the two-sided interval and
sβ is the asymptotic standard error of the parameter estimate. For parameters in loglinear subterms, estimates and bounds for eβ are also given. The bounds in that case are computed by exponentiation of the bounds on the original scale.
You can set the default level for these (and all subsequent) bounds and choose whether or not to print the parameter covariance or correlation matrix.
Related script commands