Consider the MLEs (maximum likelihood estimators) of the parameters of the Gumbel distribution using SRS (simple random sample) and RSS (ranked set sample) and the MOMEs (method of moment estimators) and REGs (regression estimators) based on SRS. A comparison between these estimators using bias and MSE (mean square error) was performed using simulation. It appears that the MLE based on RSS can be a robust competitor to the MLE based on SRS.
Yousef, Omar M. and Al-Subh, Sameer A.
"Estimation of Gumbel Parameters under Ranked Set Sampling,"
Journal of Modern Applied Statistical Methods: Vol. 13
, Article 24.
Available at: http://digitalcommons.wayne.edu/jmasm/vol13/iss2/24