Weights: Stanardized vs. raw

[SPrince]

Hello,
I am analysing a two level model with complex survey data. I have surey weights for level 1 variables that I would like to apply and keep level 2 equal. In MLWIN's weights menu I am given the option of raw vs. standardizing the weights. I am not sure which one to use in which case?

Also, the MLWin HELP menu indicates that weighting is still quite experimental and so I am wondering if they should even be used at all?

Thank you,
Stephanie