Admissible Run-Contingency Type Tests for Independence and Markov Dependence.

Admissible statistical tests of fixed level are derived for the hypotheses (a) of statistical independence and (b) of Markov dependence, against the alternative of a higher-order Markov dependence. The tests, which are related to the run test and to contingency tests, are conditional procedures given a complete sufficient statistic. Comparisons are made with chi-squared tests for independence using computer simulation. [ABSTRACT FROM AUTHOR]