![[banner]](images/banner.png "Batsto Lake, Wharton State Forest, New Jersey")
Examples in Summary and Analysis of Extension Program Evaluation
SAEPER: Two-sample Paired Rank-sum Test
SAEPER: Sign Test for Two-sample Paired Data
Packages used in this chapter
The following commands will install these packages if they are not already installed:
if(!require(BSDA)){install.packages("BSDA")}
When to use it
The poplar example is shown below in the “How to do the test” section.
Null hypothesis
How it works
Examples
Graphing the results
See the Handbook for information on these topics.
Similar tests
Paired t-test and permutation test are described in the Paired t–test chapter. The sign test is described below.
How to do the test
Wilcoxon signed-rank test example
###
--------------------------------------------------------------
### Wilcoxon signed-rank test, poplar example, p. 189
### --------------------------------------------------------------
Input = ("
Clone August November
Balsam_Spire 8.1 11.2
Beaupre 10.0 16.3
Hazendans 16.5 15.3
Hoogvorst 13.6 15.6
Raspalje 9.5 10.5
Unal 8.3 15.5
Columbia_River 18.3 12.7
Fritzi_Pauley 13.3 11.1
Trichobel 7.9 19.9
Gaver 8.1 20.4
Gibecq 8.9 14.2
Primo 12.6 12.7
Wolterson 13.4 36.8
")
Data = read.table(textConnection(Input),header=TRUE)
wilcox.test(Data$August,
Data$November,
paired=TRUE)
Wilcoxon signed rank test
V = 16, p-value = 0.03979
### Matches “Signed Rank” p-value in SAS output
Simple 1-to-1 plot of values
plot(Data$August, Data$November,
pch = 16,
xlab="August",
ylab="November")
abline(0,1, col="blue", lwd=2)
Plot of paired samples from a Wilcoxon signed-rank test. Circles above and to the left of the blue one-to-one line indicate observations with a higher value for November than for August.
# # #
Sign test example
The following is an example of the two-sample dependent-samples sign test. The data are arranged as a data frame in which each row contains the values for both measurements being compared for each experimental unit. This is sometimes called “wide format” data. The SIGN.test function in the BSDA package is used. The option md=0 indicates that the expected difference in the medians is 0 (null hypothesis). This function can also perform a one-sample sign test.
###
--------------------------------------------------------------
### Two-sample sign test, poplar example, p. 189
### --------------------------------------------------------------
Input = ("
Clone August November
Balsam_Spire 8.1 11.2
Beaupre 10.0 16.3
Hazendans 16.5 15.3
Hoogvorst 13.6 15.6
Raspalje 9.5 10.5
Unal 8.3 15.5
Columbia_River 18.3 12.7
Fritzi_Pauley 13.3 11.1
Trichobel 7.9 19.9
Gaver 8.1 20.4
Gibecq 8.9 14.2
Primo 12.6 12.7
Wolterson 13.4 36.8
")
Data = read.table(textConnection(Input),header=TRUE)
library(BSDA)
SIGN.test(x = Data$ August,
y = Data$ November,
md = 0,
alternative = "two.sided",
conf.level = 0.95)
Dependent-samples Sign-Test
S = 3, p-value = 0.09229
### Matches “Sign” p-value in SAS output
# # #