Cliff's delta Graphics and Interpretation

Share:

Description

Returns a graphical representation and interpretation of Cliff's delta

Usage

1
delta_gr(x,y, ... ,dv=2)

Arguments

x

A 1-column matrix with optional column name containing all n_x values or scores of group X or 1 (e.g. control or pretest group.).

y

A 1-column matrix with optional column name containing all n_y values of group Y or 2 (e.g. experimental or post-test group). For paired comparisons (e.g. pre-post), n_x = n_y is required.
See orddom for details.

...

Other arguments to be passed on to the orddom function, such as (for example):
- paired: to compare dependent data (e.g. pre-post) set to paired=TRUE ,
- alpha for the respective significance level to be used, e.g.alpha=.01 for 1 - onetailed to generate one-sided testing p and confidence interval (CI) values set to onetailed=TRUE ,
- studdist to obtain CI based on normal distribution z values (instead of Student distribution t) set to studdist=FALSE for 1 - symmetric to obtain symmetric rather than asymmetric CIs (see orddom for details) set to symmetric=TRUE .
- onetailed for one-sided rather than the default two-tailed testing.
- x.name to assign an individual label to group x (i.e. 1st or control or pretest group).
- y.name to label the y input matrix or group y (i.e. 2nd or experimental or posttest group).
- description This argument allows for assigning a string (as title or description) for the ordinal comparison outputs.

dv

(For paired comparisons (dv=3) only.) Determines which ordinal δ statistics are to be returned. Set to:
- dv=1 [within] to return an analysis for the n_x=n_y within-pair changes,
- dv=2 [between] to return an analysis for the overall distribution changes, based on all n^2-n = n(n-1) score comparisons between y and x where i \ne j,
- dv=3 [combined] to return an analysis for the combined inference d_w + d_b. It is advisable to use dv=3 in combination with symmetric=TRUE.

Value

Returns a graphical representation and text interpretation of Cliff's delta.

Author(s)

Jens Rogmann

See Also

orddom

Examples

 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
## Not run: 
#Paired comparison combined inference (Data taken from Long et al. (2003), Table 4)
x2<-t(matrix(c(2,6,6,7,7,8,8,9,9,9,10,10,10,11,11,12,13,14,15,16),1))
colnames(x2)<-c("Incidental")
y2<-t(matrix(c(4,11,8,9,10,11,11,5,14,12,13,10,14,16,14,13,15,15,16,10),1))
colnames(y2)<-c("Intentional")
delta_gr(x2,y2,paired=TRUE,studdist=FALSE,dv=3)
#
#Journal of Statistics Education Dataset: Oral Contraceptive Drug Interaction Study
#Journal of Statistics Education, Volume 12, Number 1 (March 2004).
columns<-c("SubjectNo","Seq","Period","Treatment","EEAUC","EECmax","NETAUC","NETCmax")
data<-read.table("http://www.amstat.org/publications/jse/datasets/ocdrug.dat",col.names=columns)
#
#returns delta (between) and 95
x<-subset(data,data$Treatment==0)[6] #Placebo EECmax
colnames(x)<-"Placebo Phase"
y<-subset(data,data$Treatment==1)[6] #Treatment EECmax
colnames(y)<-"Treatment Phase"
delta_gr(x,y,paired=TRUE,onetailed=TRUE,dv=2)
#
#checks treatment groups delta equivalence in placebo phase
#returns delta and 95
plac<-subset(data,data$Treatment==0)
x<-subset(plac,plac$Period==1)[6] #control (placebo before drug)
colnames(x)<-"Control (before Drug)"
y<-subset(plac,plac$Period==2)[6] #experimental (placebo after drug)
colnames(y)<-"Exp (Placebo after Drug)"
delta_gr(x,y)
#

## End(Not run)