Standard error

Prism uses a standard method to compute the standard error and confidence interval for each parameter fit with nonlinear regression.

Each parameter's standard error is computed using this equation:

  SE(Pi) = sqrt[ (SS/DF) * Cov(i,i) ]

where:

 Pi : i-th adjustable(non-constant) parameter
 SS : sum of squared residuals
 DF : degrees of freedom (the number of data points minus number of parameters fit by regression)
 Cov(i,i) : i-th diagonal element of covariance matrix
 sqrt() : square root