non-parametric equivalent

Hi,

First I am not a stats guy and my knowledge on stats is limited, I am just a
c++ programmer.

In a project I was asked to run a test to see if the mean response to a
medical event was different from 0, I used a simple t-test, which I believe
to be a parametric test. And it seems to giving the answers that the boss
wants. but now he asked me as a robustness to implement a similar
non-parametric test... any suggestion? I was reading a stats book and
apparently the equivalent are things like Wilcoxon, kolmogorov, but from
other book are things like signed rank test or signed test for a median... I
am just worried that I might be complicating the issue...

thank



.