Thank-you very much Samuel Buttery and Renaud Lancelot for answering my
questions.
Q1. Does the Cp statistic work with variables that have different degrees
of freedom:
A1. Both Samuel and Renaud replied that Yes, the Cp should account for
degrees of freedom however, it is preferable to use Ripley's AIC as
criteria for adding or droping terms.
Procedure is as follows: In the mass library, library(Mass)
you can use stepAIC(), addterm() and dropterm()
"AIC (= -2 * log(maximized likelihood) + 2 * number of
parameters: smaller AIC value is better) or related criteria (to account
for small sample size and / or overdispersion) instead of Cp. (Renaud)"
Renaud also cited the following references:
Burnham, K.P., Anderson, D.R., 1998. Model selection and inference: a
practical information-theoretic approach. New-York, Springer-Verlag, 353
p.
Lebreton, J.-D., Burnham, K.P., Clobert, J., Anderson, D.R., 1992.
Modeling survival and testing biological hypotheses using marked
animals: a unified approach with case studies. Ecological Monographs, 62
(1): 67-118.
Q2. What is the proper syntax for nested variables?
A2. Renaud replied that the proper syntax is:
For nested variables, the syntax is as follows:
Fb %in% Fa ### meaning Fb is nested within Fa
Fa / Fb ### meaning Fa + Fb %in% Fa
Thanks again.
Suzanne
|