Contents
- Index
Variance Estimation Methods
Two different procedures are provided in MARK to estimate the variance-covariance matrix of the estimates. The first is the inverse of the Hessian matrix obtained as part of the numerical optimization of the likelihood function. This approach is not reliable, and should only be used when you are not interested in the standard errors, and already know the number of parameters that were estimated. The only reason for including this method in the program is that it is the fastest -- no additional computation is required for the method.
The second method (the default) computes the information matrix directly using central difference approximations to the second partial derivatives. This method (labeled the 2ndPart method) provides the most accurate estimates of the standard errors, and is the default and preferred method.
Because the rank of the variance-covariance matrix is used to determine the number of parameters that were actually estimated, using different methods will sometimes result in a different number of parameters estimated, and hence a very different value of the AIC and AICc.