to anyone who can assist
my understanding of the anova (aov) model is that the
response is continuous while the predictor/s
are factors (categorical variables).
thus, if i am using an anova model, and i have
a random factor which i wish to adjust for which
is continuous, it appears to me that it is
inappropriate to place it within the error
term of aov since it is not a categorical variable.
i think my only option here in splus is to fit
some sort of linear regression, such as lme which
allows for both categorical and continuous
random variables, and then compare successive
models using the anova function from which i can
determine p values for fixed factors.
would anyone like to comment on this.
any assistance as always is much appreciated.
thanks
andrew
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
|