1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
|
\name{Steel.test}
\alias{Steel.test}
\title{
Steel's Multiple Comparison Wilcoxon Tests
}
\description{
This function uses pairwise Wilcoxon tests, comparing a common control sample with each of
several treatment samples, in a multiple comparison fashion. The experiment wise
significance probabity is calculated, estimated, or approximated, when
testing
the hypothesis that all independent samples arise
from a common unspecified distribution, or that treatments have no effect when assigned
randomly to the given subjects.
}
\usage{
Steel.test(\dots, data = NULL,
method = c("asymptotic", "simulated", "exact"),
alternative = c("greater","less","two-sided"),
dist = FALSE, Nsim = 10000)
}
\arguments{
\item{\dots}{
Either several sample vectors, say
\eqn{x_1, \ldots, x_k},
with \eqn{x_i} containing \eqn{n_i} sample values.
\eqn{n_i > 4} is recommended for reasonable asymptotic
\eqn{P}-value calculation. The pooled sample size is denoted
by \eqn{N=n_1+\ldots+n_k}. The first vector serves as control sample,
the others as treatment samples.
or a list of such sample vectors.
or a formula y ~ g, where y contains the pooled sample values
and g (same length as y) is a factor with levels identifying
the samples to which the elements of y belong. The lowest factor level
corresponds to the control sample, the other levels to treatment samples.
}
\item{data}{= an optional data frame providing the variables in formula y ~ g or y, g input
}
\item{method}{= \code{c("asymptotic","simulated","exact")}, where
\code{"asymptotic"} uses only an asymptotic normal approximation
to approximate the \eqn{P}-value, This calculation is always done.
\code{"simulated"} uses \code{Nsim} simulated standardized
Steel statistics based on random splits of the
pooled samples into samples of sizes
\eqn{n_1, \ldots, n_k}, to estimate the \eqn{P}-value.
\code{"exact"} uses full enumeration of all sample splits with resulting
standardized Steel statistics to obtain the exact \eqn{P}-value.
It is used only when \code{Nsim} is at least as large as the number
\deqn{ncomb = \frac{N!}{n_1!\ldots n_k!}}{N!/(n_1!\ldots n_k!)}
of full enumerations. Otherwise, \code{method} reverts to
\code{"simulated"} using the given \code{Nsim}. It also reverts
to \code{"simulated"} when \eqn{ncomb > 1e8} and \code{dist = TRUE}.
}
\item{alternative}{= \code{c("greater","less","two-sided")}, where for \code{"greater"} the
maximum of the pairwise standardized Wilcoxon test statistics is used and
a large maximum value is judged significant.
For \code{"less"} the minimum of the pairwise standardized Wilcoxon test
statistics is used and a low minimum value is judged significant.
For \code{"two-sided"} the maximum of the absolute pairwise standardized Wilcoxon test
statistics is used and a large maximum value is judged significant.
}
\item{dist}{\code{= FALSE} (default) or \code{TRUE}. If \code{TRUE}, the
simulated or fully enumerated null distribution vector \code{null.dist}
is returned for the Steel test statistic, as chosen via \code{alternative}.
Otherwise, \code{NULL} is returned. When \code{dist = TRUE} then
\code{Nsim <- min(Nsim, 1e8)}, to limit object size.
}
\item{Nsim}{\code{= 10000} (default), number of simulation sample splits to use.
It is only used when \code{method = "simulated"},
or when \code{method = "exact"} reverts to \code{method =}
\code{ "simulated"}, as previously explained.
}
}
\details{
The Steel criterion uses the Wilcoxon test statistic in the pairwise comparisons of the
common control sample with each of the treatment samples. These statistics are used in
standardized form, using the means and standard deviations as they apply conditionally
given the tie pattern in the pooled data, see Scholz (2016). This conditional treatment allows for
correct usage in the presence of ties and is appropriate either when the samples are independent
and come from the same distribution (continuous or not) or when treatments are assigned
randomly among the total of \code{N} subjects. However, in the case of ties the significance probability
has to be viewed conditionally given the tie pattern.
The Steel statistic is used to test the hypothesis that the samples all come
from the same but unspecified distribution function \eqn{F(x)}, or, under random treatment
assigment, that the treatments have no effect. The significance probability is the probability
of obtaining test results as extreme or more extreme than the observed test statistic,
when testing for the possibility of a treatment effect under any of the treatments.
For small sample sizes exact (conditional) null distribution
calculations are possible (with or without ties), based on a recursively extended
version of Algorithm C (Chase's sequence) in Knuth (2011), which allows the
enumeration of all possible splits of the pooled data into samples of
sizes of \eqn{n_1, \ldots, n_k}, as appropriate under treatment randomization. This
is done in C, as is the simulation of such splits.
NA values are removed and the user is alerted with the total NA count.
It is up to the user to judge whether the removal of NA's is appropriate.
}
\value{
A list of class \code{kSamples} with components
\item{test.name}{\code{"Steel"}}
\item{alternative}{ "greater", "less", or "two-sided"}
\item{k}{number of samples being compared, including the control sample as the first one}
\item{ns}{vector \eqn{(n_1,\ldots,n_k)} of the \eqn{k} sample sizes}
\item{N}{size of the pooled sample \eqn{= n_1+\ldots+n_k}}
\item{n.ties}{number of ties in the pooled sample}
\item{st}{2 (or 3) vector containing the observed standardized Steel statistic,
its asymptotic \eqn{P}-value,
(its simulated or exact \eqn{P}-value)}
\item{warning}{logical indicator, \code{warning = TRUE} when at least one
\eqn{n_i < 5}
}
\item{null.dist}{simulated or enumerated null distribution vector
of the test statistic. It is \code{NULL} when \code{dist = FALSE} or when
\code{method = "asymptotic"}.
}
\item{method}{the \code{method} used.}
\item{Nsim}{the number of simulations used.}
\item{W}{vector
\eqn{(W_{12},\ldots, W_{1k})}
of Mann-Whitney statistics comparing each observation under treatment \eqn{i (> 1)}
against each observation of the control sample.
}
\item{mu}{mean vector \eqn{(n_1n_2/2,\ldots,n_1n_k/2)} of \code{W}.}
\item{tau}{vector of standard deviations of \code{W}.}
\item{sig0}{standard deviation used in calculating the significance probability
of the maximum (minimum) of (absolute) standardized Mann-Whitney statistics, see Scholz (2016).}
\item{sig}{vector
\eqn{(\sigma_1,\ldots, \sigma_k)}
of standard deviations used in calculating the significance probability
of the maximum (minimum) of (absolute) standardized Mann-Whitney statistics, see Scholz (2016).}
}
\references{
Knuth, D.E. (2011), \emph{The Art of Computer Programming, Volume 4A
Combinatorial Algorithms Part 1}, Addison-Wesley
Lehmann, E.L. (2006),
\emph{Nonparametrics, Statistical Methods Based on Ranks, Revised First Edition},
Springer Verlag.
Scholz, F.W. (2023), "On Steel's Test with Ties", \url{https://arxiv.org/abs/2308.05873}
}
\section{warning}{\code{method = "exact"} should only be used with caution.
Computation time is proportional to the number of enumerations.
Experiment with \code{\link{system.time}} and trial values for
\code{Nsim} to get a sense of the required computing time.
In most cases
\code{dist = TRUE} should not be used, i.e.,
when the returned distribution objects
become too large for R's work space.}
\examples{
z1 <- c(103, 111, 136, 106, 122, 114)
z2 <- c(119, 100, 97, 89, 112, 86)
z3 <- c( 89, 132, 86, 114, 114, 125)
z4 <- c( 92, 114, 86, 119, 131, 94)
y <- c(z1, z2, z3, z4)
g <- as.factor(c(rep(1, 6), rep(2, 6), rep(3, 6), rep(4, 6)))
set.seed(2627)
Steel.test(list(z1, z2, z3, z4), method = "simulated",
alternative = "less", Nsim = 1000)
# or with same seed
# Steel.test(z1, z2, z3, z4,method = "simulated",
# alternative = "less", Nsim = 1000)
# or with same seed
# Steel.test(y ~ g, method = "simulated",
# alternative = "less", Nsim=1000)
}
\keyword{nonparametric}
\keyword{htest}
\keyword{design}
|
