Hello, afraid stats was never my strongest maths subject, so hoping someone can help me with this.

I'm fitting a non-linear regression line to some data. The line has the form

y=a0+1.086*(x/a1)^a2

where a0, a1 and a2 are free parameters. I'm currently using the program xmgrace to plot graphs and carry out function fitting. A typical output would be

Fitting with formula: y=a0+1.086*(x/a1)^a2

Initial guesses:

a0 = 1

a1 = 1

a2 = 1

Tolerance = 1e-14

Relative error in the sum of squares is at most tol.

Computed values:

a0 = 19.5413

a1 = 8.41693

a2 = 0.974081

Chi-square: 0.932071

Correlation coefficient: 0.974664

RMS per cent error: 0.00781243

Theil U coefficent: 0.00796977

What I would love to know is would any of the above allow me to calculate the individual errors for each of the parameters?

Thanks for reading,

Rai