I am using the R package 'drc' to calculate a dose-response model for continuous data (with the function 'drm'). I then reverse-predict the ED50 value using the 'ED' function. The function returns the value of the ED50 and a standard error, but I cannot figure out how the standard error is calculated. I know how a standard error in general is calculated, but what is the sample on which it is calculated in this case? The papers to which the manual of the package refer did not help. Thanks in advance.