Selected Works of E. L. Lehmann pp 391-402 | Cite as

# The Efficiency of Some Nonparametric Competitors of the t-Test

## Summary

Consider samples from continuous distributions F(x) and F(x — θ). We may test the hypothesis θ = 0 by using the two-sample Wilcoxon test. We show in Section 1 that its asymptotic Pitman efficiency, relative to the f-test, never falls below 0.864. This result also holds for the Kruskal-Wallis test compared with the jF-test, and for testing the location parameter of a single symmetric distribution.

A number of alternative notions of asymptotic efficiency are compared in Section 2. In this connection, certain difficulties arise because power is not necessarily a convex function of sample size. As an alternative to the Pitman notion of asymptotic efficiency, we consider in Section 3 one based on the speed with which power at a fixed alternative tends to 1. In particular we obtain, for the sign test relative to the t in normal populations, the limit as n —► ∞ of the sequence of power efficiency functions. It is noted that certain interchanges of limit passages are not always possible.

### Keywords

Summing### References

- E. J. G. Pitman, Lecture notes on nonparametric statistical inference, Columbia University, 1949.Google Scholar
- G. E. Noether, “On a theorem of Pitman,” Ann. Math. Stat., Vol. 26 (1955), pp. 64–68.MathSciNetMATHCrossRefGoogle Scholar
- W. Hoeffding and J. Rosenblatt, “The efficiency of tests,” Ann. Math. Stat., Vol. 26 (1955), pp. 52–63.MathSciNetMATHCrossRefGoogle Scholar
- A. Mood, “On the asymptotic efficiency of certain non-parametric two-sample tests” Ann. Math. Stat., Vol. 25 (1954), pp. 514–522.MathSciNetMATHCrossRefGoogle Scholar
- F. C. Andrews, “Asymptotic behavior of some rank tests for analysis of variance,” Ann: Math. Stat., Vol. 25 (1&54), pp. 724–735.Google Scholar
- S. Saks, Theory of the Integral, Warsaw-Lvov, 1937.Google Scholar
- W. Hoeffding, “Optimum* nonparametric tests,” Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability, University of California Press, 1951.Google Scholar
- M. E. Terry, “Some rank order tests which are most powerful against specific par-ametric alternatives,” Ann. Math. Stat., Vol. 23 (1944), pp. 346–366.MathSciNetCrossRefGoogle Scholar
- M. Dwass, “The large sample power of rank order tests in the two-sample problem,” Ann. Math. Stat., this issue.Google Scholar
- B. L. van der Waekden, “Ein neuer Test ftir das Problem der zwei Stich proben,” Math. Ann. Vol. 126 (1953), pp. 93–107.MathSciNetCrossRefGoogle Scholar
- W. H. Kruskal and W. A. Wallis, “Use of ranks in one-criterion analysis of var-iance,” Jour. Am. Stat. Assoc., Vol. 47 (1952), pp. 583–621.MATHCrossRefGoogle Scholar
- J. L. Hodges, Jr. and E. L. Lehmann, “Matching in paired comparisons,” Ann. Math. Stat., Vol. 25 (1954), pp. 787–791.MathSciNetMATHCrossRefGoogle Scholar
- W. J. Dixon, “Power functions of the sign test and power efficiency for normal alternatives,” Ann. Math. Stat., Vol. 24 (1953), pp. 467–473.MATHCrossRefGoogle Scholar
- W. J. Dixon, “Power under normality of several nonparametric tests,” Ann. Math. Stat., Vol. 25, (1954), pp. 610–614.MATHCrossRefGoogle Scholar
- H. Chernoff, “A measure of asymptotic efficiency for tests of a hypothesis based on the sum of observations,” Ann. Math. Stat., Vol. 23 (1952), pp. 493–507.MathSciNetMATHCrossRefGoogle Scholar
- J. V. Uspensky, Introduction to Mathematical Probability, McGraw-Hill Book Comp-any, Inc., 1937, New York.MATHGoogle Scholar
- W. L. Nicholson, “A computing formula for the power of the analysis of variance test,” Ann. Math. Stat., Vol. 25 (1954), pp. 607–610.MathSciNetMATHCrossRefGoogle Scholar