I'm more of a programmer, so perhaps my confusion is OK.

I have 66 subjects, testing 8 products, (a different one each visit.)

I'm doing Wilcoxon tests, comparing each product against another.

Client has asked for Confidence Intervals at both 90% and 95%, showing statistical advantage.

Of course, both runs give the same p-value, and more or less, the exact same confidence interval. Doing Top 1 Box, Top 2 Box, Top 3 box, Middle 1 Box, Bottom 2 Box. Even though the P-values are the same, there are a couple of CI differences.

Most of my CI values are 0 or -.5

Anyway, what is my clinic looking for with the CI difference?