Hi :) I hope someone can help me figure out what statistical test(s) are relevant in the following situation. It is new for me working with binary data.
10 patients have blood submitted to 4 laboratories testing 14-21 different parameters each. The test results are given as positive or negative. I would like to find out if there is a statistical difference in the results amongst laboratories when testing the same parameters on the same patient. I was thinking of doing pairwise comparisons between laboratories since not all the parameters are available for testing in all 4 labs.
Tests I have considered: - Similarity coefficients for binary data e.g. Phi - Cohens kappa - Chi/fischers/Mcnemars test
Given the small sample size (10 patients/140-210 parameters per. lab), the fact that the data is binary, paired/repeated (blood from the same patient) - what statistical tests are relevant?
Thanks :)
