The following features have been added to the QLIM procedure:
The following Bayesian analysis features have been added:
harmonic mean evaluation of the marginal likelihood to compare competing models (this estimator does not require additional simulations after the posterior samples have been obtained)
evaluation of the marginal likelihood using an importance sampling algorithm based on the cross-entropy theory (this estimator requires additional importance sampling simulations after the posterior samples have been obtained)
Bayesian analysis support for multivariate models in which the likelihood function is not available in closed form
The RANDOM statement, which enables you to estimate the random-intercept models, has been added to PROC QLIM. Any single-equation model in PROC QLIM can be expanded to a random-intercept single-equation model. The RANDOM statement enables you to use panel data in your estimations, and it offers three methods to integrate out the random intercept. These methods are Gauss-Hermite quadrature, simulation, and quasi–Monte Carlo using a Halton sequence.
The random-intercept models include the following:
random-intercept linear regression models
random-intercept discrete choice models, including binary probit, binary logit, ordinal probit, and ordinal logit models
random-intercept limited dependent variable models, including censored regression and truncated regression models
random-intercept stochastic frontier models