Test if Distributions are the same

[lancearmstrong1313]

n,.,.,.,.,.,

[jrai]

The maximums follow Gumbel distribution. I think that the normality assumption would be inaccurate.

[lancearmstrong1313]

What about running a Kolmogorov-Smirnov Test? I've never done that before though.

[jrai]

KS test is a non-parametric test & you don't have to specify the distributions. Therefore, it is a good idea. But as I was reading in some other post on this site that K-S statistic tests extreme deviations & in some cases might not be very reliable. Q-Q plot is a much more reliable method (according to the author of that post).