The TCALIS Procedure |
SIMTEST Statement |
When the estimates in a model are asymptotically multivariate-normal, continuous and differentiable functions of the estimates are also multivariate-normally distributed. In the SIMTEST statement, you can test these parametric functions simultaneously. The null hypothesis for the simultaneous tests is assumed to have the following form:
: , , ...
where is the set of model parameters (independent or dependent) in the analysis and each is a continuous and differentiable function of the model parameters.
To test parametric functions simultaneously in the SIMTEST statement, you first assign a name for the simultaneously test in test_name. Then you put the parametric functions for the simultaneous test inside a pair of parentheses: either the '{' and '}' pair, or the '[' and ']' pair. For example, if , , , and are parameters in the model and you want to test the equality of and and the equality of and simultaneously, you can use the following statements:
simtest Equality_test = [t1_t2_diff t3_t4_diff]; t1_t2_diff = theta1 - theta2; t3_t4_diff = theta3 - theta4;
In the SIMTEST statement, you test two functions t1_t2_diff and t3_t4_diff simultaneously in the test named Equality_test. The two parametric functions t1_t2_diff and t3_t4_diff are computed in the SAS programming statements as differences of some parameters in the model.
See also the TESTFUNC statement for testing parametric functions individually.
Note: This procedure is experimental.
Copyright © 2009 by SAS Institute Inc., Cary, NC, USA. All rights reserved.