CPANEL Procedure

The TEST statement performs Wald, Lagrange multiplier, and likelihood ratio tests of linear hypotheses about the regression parameters in the preceding MODEL statement. Like RESTRICT statements, TEST statements before the first MODEL statement are automatically associated with the first MODEL statement, in addition to any TEST statements that follow it but precede subsequent MODEL statements. Each equation specifies a linear hypothesis to be tested. All hypotheses in one TEST statement are tested jointly. Variable names in the equations must correspond to regressors in the preceding MODEL statement, and each name represents the coefficient of the corresponding regressor. The INTERCEPT keyword refers to the coefficient of the intercept.

You can specify the following options after a slash (/):

The Wald test is performed by default.

The following statements illustrate the use of the TEST statement:

proc cpanel data =  mylib.a;
   id csid tsid;
   model y = x1 x2 x3;
   test x1 = 0, x2 * .5 + 2 * x3 = 0 / all;
   test intercept = 0, x3 = 0;
run;

The first test investigates the joint hypothesis that and .

Currently, only linear equality restrictions and tests are permitted in PROC CPANEL. Tests and restriction expressions can be composed only of algebraic operations that involve the addition symbol (+), subtraction symbol (–), and multiplication symbol (*).

If you are fitting a dynamic panel model, you can perform tests on lags of the dependent variable by referencing the name of the dependent variable, followed by an underscore and the lag order. For example:

proc cpanel data =  mylib.a;
   model sales = price / dyndiff;
   test sales_1 = 0.5 / wald;
run;

When you use the TEST statement along with the MODEL statement options AMACURDY, DYNDIFF, DYNSYS, HTAYLOR, IVBTWNG, IVFIXONE, IVFIXONETIME, IVFIXTWO, IVG2SLS, IVPOOLED, IVRANONE, and IVRANTWO, only the Wald test is available.