[R] Interpreting ANOVA / quantile regression results
Yunqi Zhang
yqzhang at ucsd.edu
Sat Nov 22 05:56:17 CET 2014
Hi all,
I think this is probably a very simple question related to ANOVA or
quantile regression.
Let's say if I'm doing ANOVA on 2 factors (f_a and f_b) and trying to
figure out whether they have an impact on the output. The result shows me
that f_a has a significant impact while f_b does not as following, where
the std. error (confidence interval) is
estimate std. error p-value
f_a 10 0.2 0.001
f_b 0.5 2 0.56
If I take this result, and I want to make my best guess for the output of a
pair of f_a and f_b values. Should I calculate the output using both f_a
and f_b or only the f_a because f_b is statistically insignificant? And
what the std. error (confidence interval) would be?
Specifically, should I do
guessed output = 10 * f_a + 0.5 * f_b and the std. error is 0.2 + 2 = 2.2
or
guessed output = 10 * f_a and the std.err is 0.2
Thank you very much!
Yunqi
[[alternative HTML version deleted]]
More information about the R-help
mailing list