[BIOINFORMATICS?] Compare two methods

Question

[ravihansa82]

Dear Friends

I am going to analyze a new way of processing of pathological samples for microscopic examination compared to the conventional method. I am going to look for, microtony characteristics(V1), mounting characteristics (V2), etc. for each I will give a score varying from 1 to 3(1 = poor ,2= fair, 3=good). for example, I will look for microtony characteristics for both conventional and new method and give a score. I have observed 300 tissue samples including 5 different tissue types. my results out put has ordinal data. now I need to look whether the new sample preparation method is good compared to the conventional method. Since I have many ordinal variable how I proceed statistical test for this? I need your suggestions.

Here is the sample out put. I compare same set of variable(V1, V2, etc) for both methods and give respective score for each variable

        conventional method                                new method
V1    V2    V3    V4    V5    V6                V1    V2    V3    V4    V5    V6
2     2     3     2     2     2                 3     2     3     3     3     3

[Adrian Pelin]

I don't see how ChiSquare would be wrong here.

[Ram]

Maybe see how values for V1, V2 etc are distributed among the samples? Also, not sure if bioinformatics, other mods please advise.

[Nicolas Rosewick]

I guess a Mann-Whitney U test should also work

[Jean-Karim Heriche]

You could also consider Cohen's Kappa. It's a measure of agreement between the two methods taking into account that they could agree by chance.