# [R] post hoc test to a significant Kruskal-Wallis test package asbio function pairw.kw()

arun smartpink111 at yahoo.com
Tue Oct 22 05:12:26 CEST 2013

```Hi,
Try:
pairw.kw(y,factor(x),conf=.95)

A.K.

On Monday, October 21, 2013 7:10 PM, luigi marongiu <marongiu.luigi at gmail.com> wrote:
Dear all,
I am trying to apply a post hoc test to a significant Kruskal-Wallis test.
I eventually found the function ‘pairw.kw()’ of the package “asbio”. This
function has the following syntax:
pairw.kw(y, x, conf)
Arguments
y     The response variable. A vector of quantitative responses.
x     An explanatory variable. A vector of factor levels.
conf     The level of desired conﬁdence, 1 - P(type I error).

The example reported indicated 2 vectors of length 12 arranged in a data
frame:
rye.data <- data.frame(rye = c(50, 49.8, 52.3, 44.5, 62.3, 74.8, 72.5, 80.2,
47.6, 39.5, 47.7,50.7), nutrient = factor(c(rep(1, 4), rep(2, 4), rep(3,
4))))
with(rye.data, pairw.kw(y = rye, x = nutrient, conf = .95))

I tried to arrange the data in this way, as well as using my own dataframe
obtained from the read.table() function and directly with two vectors, but
I always obtained an error. Essentially the two vectors are not of the same
length; this is incorrect since there are both 64 elements.

Is there somebody who could tell me how to use this function?

Many thanks,
Luigi

EXAMPLE GIVEN
Here I am using the vector example.

### define vectors
x<-c(    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
0,    1,    1,    1,    1,    1,    1,    1,    1,    1,    1,    1,
1,    2,    2,    2,    2,    2,    2,    2,    2,    2,    2,    2,
2,    2,    2,    2,    2,    3,    3,    3,    3,    3,    3,    3,
3,    3,    3,    3,    3,    3,    3,    3,    4,    4,    4,    4,
4,    4,    4,    4,    4)

y<-c(    0.002515324,    0.005128344,    0.008474943,    0.009992525,
0.01460982,    0.020567249,    0.069597233,    0.192428254,
0.636910481,    0.673830995,    4.354055891,    4.847755691,
0.403135211,    0.539626363,    0.892722111,    0.993636963,
1.289747804,    3.766986726,    5.623002684,    8.52186547,
17.24684431,    19.84624792,    25.9529251,    35.64632645,
0.437492775,    0.594863773,    0.64377689,    0.88033664,
3.199906591,    5.745723357,    6.339129057,    6.583851172,
11.27921212,    13.01230334,    14.61151433,    26.68041064,
52.42360789,    71.88062442,    91.34132781,    115.1038548,
0.059333604,    0.068746428,    0.1473799,    0.220983138,
0.959677864,    1.21197612,    2.860983172,    4.402992963,
5.122748306,    8.711530662,    25.6245157,    44.0798333,
45.24736747,    66.4490478,    86.60427738,    0.10348061,
0.390758249,    0.634667258,    1.122594755,    2.726868877,
4.15194739,    5.361175006,    8.825635885,    45.75640999)

### confirm length:
length(x)    # 64
length(y)    # 64

### KRUSKAL-WALLIS to test inter-group differences
kruskal.test(y, x)     # p-value = 0.0008701

### Multiple pairwise comparison procedure to accompany a Kruskal-Wallis
test
pairw.kw(y, x, conf = .95)

# Error in outer(levels(x), levels(x), function(x1, x2) { :
# dims [product 0] do not match the length of object [1]

[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help