Hello,
I'd like to compare the predictive power of two independent simple
regression models, which relate to data sets with different numbers of
points (n1 and n2). I think I could use a F-test to compare both
residual variances, but I don't know whether it's a good idea or not.
To my opinion, the numbers of degrees of freedom for the F-test would
be respectively the number of degrees of freedom of the greater
residual variance (in the numerator of the F-statistic), e.g. n1-2,
and the number of degrees of freedom for the smaller residual variance
(in the denominator of the F-statistic), e.g. n2-2 ; but I'm not sure
about what quantities to put in the numerator and the denominator of
the F-statistic. Indeed, for two sample variances, the formula to
calculate the F-statistic would be (n1*s1²/(n1-1))/(n2*s2²/(n2-1)),
with s1 and s2 being sample variances, but what happens when variances
are residual variances coming from two independent simple regression
analyses ? Is there any function in R that would enable me to perform this kind
of test ?
Could anyone help me to solve these issues ?
Thank you in advance,
Scoubi.
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
