I am comparing the results obtained from three methods. I want to use the Wilcoxon Rank-Sum test ,with a confidence value equal to 0.95.
In a reference I found it gives the result in the form +, - or =.
My data is in the form of real values, e.g:
data1=[0.26 15.2 26]
data2=[15.36 25.1 13.00]
data3=[254.3 265.1 0.0041]
Can anyone please tell me how to use it? Have in that the less the result is the better it is (a minimisation problem).
Analysis of non parametric data is not easy at all, if you have a one way of classification analysis, i asumme that it would be not a problem. This is easy to solve. First determine if the PROC NPAR1WAY, ANOVA gives differences among the k populations of means, after that the most rational way is to perform the analysis by using non parametric tests for comparing intersting hypothesis, if any, say por example, by using the appropiate contrast, according with your software, or worse by hand.
If not see Dr Agresti work.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics