[R-sig-ME] Fw: Meaning of parameters in nlme output

[Juliet Ndukum]

----- Forwarded Message ----
From: Juliet Ndukum <jpntsang at yahoo.com>
To: R-SIG-Mixed-Models at R-project.org
Sent: Fri, January 22, 2010 4:05:58 PM
Subject: Meaning of parameters in nlme output




The data set Soybean.csv is the modified Soybean data set in the R nlme library. I wish to test the effect of the fixed effect Svariety which is a factor with three levels allen, fowler and glenn. In other words, I wish to investigate the effect of Svariety on the individual parameters i.e. beta, theta and lambda.

The file project.R contains the R code, whereby the self-starting three parameter logistic model SSlogis is used.

First concern, what is the meaning of the intercept term?
What is the meaning of the parameter estimates?

How do I compare variety allen with variety glenn for example? Since there are three levels, how many comparisons can I possibly have?

Thank you in advance for your help.