quantileTestPValue {EnvStats} | R Documentation |
Compute the p-value associated with a specified combination of
m
, n
, r
, and k
for the
quantile test (useful for determining r
and
k
for a given significance level \alpha
).
quantileTestPValue(m, n, r, k, exact.p = TRUE)
m |
numeric vector of integers indicating the number of observations from the
“treatment” group.
Missing ( |
n |
numeric vector of integers indicating the number of observations from the
“reference” group.
Missing ( |
r |
numeric vector of integers indicating the ranks of the observations to use as the
lower cut off for the quantile test. All values of |
k |
numeric vector of integers indicating the number of observations from the
“treatment” group contained in the |
exact.p |
logical scalar indicating whether to compute the p-value based on the exact
distribution of the test statistic ( |
If the arguments m
, n
, r
, and k
are not all the same
length, they are replicated to be the same length as the length of the longest
argument.
For details on how the p-value is computed, see the help file for
quantileTest
.
The function quantileTestPValue
is useful for determining what values to
use for r
and k
, given the values of m
, n
, and a
specified significance level \alpha
. The function
quantileTestPValue
can be used to reproduce Tables A.6-A.9 in
USEPA (1994, pp.A.22-A.25).
numeric vector of p-values.
See the help file for quantileTest
.
Steven P. Millard (EnvStats@ProbStatInfo.com)
See the help file for quantileTest
.
quantileTest
, wilcox.test
,
htest.object
, Hypothesis Tests.
# Reproduce the first column of Table A.9 in USEPA (1994, p.A.25):
#-----------------------------------------------------------------
p.vals <- quantileTestPValue(m = 5, n = seq(15, 45, by = 5),
r = c(9, 3, 4, 4, 5, 5, 6), k = c(4, 2, 2, 2, 2, 2, 2))
round(p.vals, 3)
#[1] 0.098 0.091 0.119 0.089 0.109 0.087 0.103
#==========
# Clean up
#---------
rm(p.vals)