Elementary Estimates: An Introduction to Nonparametrics
Keyword(s):
The paper presents a unified approach to some of the more popular nonparametric methods in current use. The approach provides the reader with new insights by exhibiting relationships to relevant population parameters, such as location and scale parameters for the one- and two-sample problems and regression parameters for bivariate data. For each parameter, a set of so-called elementary estimates is defined. The elementary estimates are then used to determine point estimates, confidence intervals, and test statistics for testing relevant nonparametric hypotheses. Among the tests discussed are the sign test, the Wilcoxon one- and two-sample tests, and Kendall’s test of independence.
2002 ◽
Vol 14
(4)
◽
pp. 459-476
◽
1992 ◽
Vol 60
(3)
◽
pp. 243-269
◽
1998 ◽
Vol 40
(1)
◽
pp. 53-64
◽
2005 ◽
Vol 35
(1)
◽
pp. 1-20
◽