Ordanova

Statistical test ORDANOVA for nominal data.

The test takes a table in which each line represents a nominal category and each column represents a analysis of the data. Each element table(i,j) is the number of occurrences of the data of that category i for the method j.

H_0 : p_{k1} = p_{k2} = \dots = p_{kM}
The probabilities are the same in each column
H_1 : one or more frequencies of the same category k (represented by
lines) are not equal to the others

INPUTS:
table : in format (number of categories) lines and
(number of treatments) columns
alpha : desired significance value (default = 0.05)

OUTPUTS:
p : p-value of the test
h : hyphoteses (h=1: values are different)
i : test value

EXAMPLE 1:
table = [4,2,3;83,90,85;10,6,7;3,2,5];
[p,h,i] = ordanova(table)
p =
0.5237
h =
0
i =
0.6468

EXAMPLE 2:
table = [4,2,3;83,90,7;10,6,83;3,2,5];
[p,h,i] = ordanova(table)
p =
0
h =
1
i =
70.7250

The test is described by :
Emil Bashkansky, Tamar Gadrich and Ilya Kuselman
in
Interlaboratory comparison of test results of an ordinal or nominal
binary property: analysis of variation