On 6/05/2008, at 11:19 AM, <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> wrote:
Dear list,
I have a time-series Y of length n which has significant auto-
correlation at lag 1, as indicated by acf plots.
According to certain criteria, I have defined two groups of
observations n1 and n2. The objective is to estimate the difference
of the means between these groups and test for significance. The
problem is how to correct for auto-correlation within the groups.
I guess one way would be to delete consecutive observations within
the groups?
Could maybe bootstrap, without resampling, be another option?
In that case, could I use twot. permutation ( ) function of DAAG
package? How would it be possible to obtain the SE of the difference?
It seems to me that your ideas are rather inchoate. You need to get
a more clearly
formulated model for the structure of the time series. Is it
stationary? If not,
then autocorrelation is not directly meaningful. If so, then the
mean is constant,
and the means of any ``groups'' will be the same.
If your model is properly formulated, then the question of what
software to use to
analyze the data usually answers itself.
cheers,
Rolf Turner
######################################################################
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}
______________________________________________
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.