Dear all, In the regression literature we can specify yt= f(xt,b)+ut where f(.,b) can represent a monotonic nonlinear function with b a scalling factor. In nonparamentic analysis the same specification can be applied however now the f(.,.) is an unspecified "nonparametric" function. Due to its nonparametric nature the second methodology does not use any parametric assumption, Can we use the same principles? e.g. keep parametric assumptions at the minimal level (not using a location parameter)in the first case? Will that (not using a location parameter)help in better revealing the yt,xt relationship than using a non-systematic component like the location parameter? I'll appreciate your reply Rudolf ____________________________________________________________ Do You Yahoo!? Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk or your free @yahoo.ie address at http://mail.yahoo.ie