Error propagation for non-linear regression parameters

#1
Hi,

I've been doing non-linear regression on some datasets using R. The regression outputs parameter estimation and parameter standard error (SE) for the three parameters a, n and k. I calculate the value for a new parameter f = n/k. I want to find the SE for f using error propagation. I found formulas at wikipedia for error propagation, but all are using standard deviation and not standard error. Can I use the same formulas for standard error?

SE formula:
f SE = SQRT((n/n SE)^2+(k/k SE)^2-2*covariance[n/k]/(n*k))*f

based on: http://upload.wikimedia.org/wikipedia/en/math/c/1/3/c1369adf1eba9c3d585151de58bac4d6.png
and :
http://upload.wikimedia.org/wikipedia/en/math/0/7/6/076d3820a46afe55ee680f3c85e34c76.png

Regression formula:
c~a+log10(1-(1-10^(k*day))^(10^n))
(c and day found in dataset)

Thanks!
/Jörgen