The research involves the study of interblock rank tests in statistics for univariate and multivariate factorial designs in which all of the dependent data is ranked without regard to block or variate membership. The study will include a comparison of the small sample properties, efficiency, and robustness properties of certain interblock rank statistics with the corresponding properties of aligned rank tests, rank transform procedures and intrablock rank tests. This will be extended to the asymptotic properties of linear rank statistics with dependent data under general alternatives. Research will also be conducted on the extension of certain approaches of generating intrablock rank test from metrics on permutations to include new interblock rank tests.
