[R] statistical advice

[Greg Snow]

I have had luck using the circular and the CircStats packages to look at sleep related data.  I have fit mixtures of vonmeises distributions to time of occurance responses (have not looked at duration with time of occurance).  There are also tools for doing regression models with circular responses (time of day) and either linear or circular predictors.  If you want to use a circular predictor to predict a non-circular response then there are tools for doing periodic splines which work in that case.

Hope this helps,  If you give more detail of what you want to accomplish (either in the mailing list, or offline), then we/I can give more detail.

--
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
(801) 408-8111



> -----Original Message-----
> From: r-help-bounces at r-project.org
> [mailto:r-help-bounces at r-project.org] On Behalf Of Allison Brager
> Sent: Wednesday, September 03, 2008 10:44 AM
> To: r-help at r-project.org
> Subject: [R] statistical advice
>
> Hello,
>
> I am a researcher in sleep and circadian rhythms who is
> having much trouble deciding on proper statistical analyses.
> Before I state my question, I provide a brief synopsis of the
> looming problem; I am interested in activity bout
> distributions across a 24 hr day. In addition to looking at
> the number of activity bouts across the day, I am also
> interested in the duration of each activity bout. My current
> method of analyzing numbers and durations of activity bouts
> involves; 1) documenting the number of activity bouts for
> each condition  (alcohol or water drinker) and then, using
> repeated measures ANOVA to account for three days of activity
> recording 2) displaying the number of each bout duration
> (expressed in 15 minute bins) in a frequency histogram, that
> again, documents differences in bout number between
> conditions. My concern lies in the size of the frequency bins
> for the bout durations. My advisor suggested 15  min bins, my
> co-advisor suggests 60 min, I think 10 min (from
> *Quantitative analysis of the age-related fragmentation of
> hamster 24-h activity rhythms* by Plamen D. Penev, Phyllis C.
> Zee, and Fred W. Turek) which leads me to my question: IS
> THERE A MORE CONTINUOUS WAY OF ANALYZING ACTIVITY DURATION
> DIFFERENCES THAT DOES NOT INVOLVE "COUNTING"?
>
> If you would like more insight on my experiment in order to
> have a more accurate understanding of the problem, please do
> not hesitate to contact me.
>
> Sincerely,
> Allison Brager
>
> --
> Graduate Student
> Department of Biological Sciences
> Kent State University
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at 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.
>