Back to search

Boston University

The treatment of ties in non-parametric tests.

Abstract

dc:description.abstract

In most of the non-parametric methods, only a few general assumptions are made concerning the underlying distribution of the population from which a certain sample is drawn. One of the most frequent of these assumptions is that of continuity, i.e., that the population from which the sample is drawn possesses a continuous distribution, and, therefore, the probability of two or more equal observations is zero. Actually, however, due to limitation of measurement, experimental data are such that they must usually be regarded as coming from discontinuous distributions and equal observations will occur. When this is the case, one speaks of the occurrence of "tied" observations, or simply "ties", in the data. [TRUNCATED] In this paper are discussed the dif£erent solutions that various statisticians have suggested in treating tied observations when applying the following non-parametric tests: (1) the sign test, (2) the Wilcoxon (Mann-Whitney) test, (3) the Wald-Wolfowitz Runs test, (4) The Moses test, (5) the Kruskal-Wallis test, (6) Kendall's rank correlation coefficient test, and {7) Kendall's coefficient of concordance test. It is found that most of the statisticians recommend the mid-rank method of treating ties.

Degree

thesis:*
Name thesis:degree_name
Master of Arts
Level thesis:degree_level
masters
Discipline thesis:degree_discipline
Mathematics
Grantor dc:publisher
Boston University
Year dc:date.issued
1959

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Duenas, Jeannene

Rights

dc:rights
Statement dc:rights
  • Based on investigation of the BU Libraries' staff, this work is free of known copyright restrictions.
Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Dc Identifier Other
b14702459

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
related terms
citation

Duenas, Jeannene. The treatment of ties in non-parametric tests.. masters thesis, Boston University, 1959. https://hdl.handle.net/2144/14444