SECTION 15.18 WALD-WOLFOWITZ TEST

The Wald-Wolfowitz test uses a run test to test the hypothesis that two sample distributions are from the same underlying theoretical distributions. See S. Siegel and N. J. Castellan, Nonparametric Statistics, 2nd ed., 1988, pp. 58-64.

Be sure to CALL DFLT before calling WALDWOLF and to include the COMMON/BPRINT/ common block. The call is

        CALL WALDWOLF(A1,A2,N1,N2,IRUN,IS,IL,ZVAL)

where

        A1    = DOUBLE PRECISION array dimensioned A1(N1) containing
                the first sample (input)
        A2    = DOUBLE PRECISION array dimensioned A2(N2) containing
                the second sample (input)
        N1,N2 = sample sizes (input)
        IRUN  = the number of runs (output)
        IS    = LOGICAL variable set to .TRUE. if small sample test
                is feasible and gives significant difference (output)
        IL    = LOGICAL variable set to .TRUE. if asymptotic test is
                performed (output), in which case
        ZVAL  = DOUBLE PRECISION scalar containing the z-value (output)

|Beginning|SHELL|