Re: [R] help with multivariate analysis

From: Spencer Graves <spencer.graves_at_pdf.com>
Date: Sat 17 Dec 2005 - 10:48:01 EST

Dear Dr. Vidali:

  1. I would first plot the data a number of different ways. Maybe you've already done plenty of this, but I have had a tendancy to rush to the model I thought was the "best" without adequate data integrity checks. I typically want to start with normal probability (using "qqnorm"). Outliers, the need for transformations, mixtures, etc., will jump off the page when I do this. Sometimes, I find that I didn't get the data read properly. Also, I may consider log(heart.frequency) and 1/heart.frequency as well as heart.frequency. In most cases, I will base the subsequent analysis on whichever looks most nearly normally distributed.
  2. I will also want to make scatterplots of the two variables as well as plotting each vs. time (with one line for each patient and separate plots for the two groups, but on the same scale). This can tell me how much correlation to expect as well as what kinds of effects I want to model.
  3. Have you consulted Pinheiro and Bates (2000) Mixed-Effects Models in S and S-Plus (Springer)? The first half of that book describes how to analyze data sets like you describe when effects are linear and additive. The latter half describes what to do with nonlinear effects.
  4. If you still would like further assistance from this group, please submit another post, but PLEASE do read the posting guide! "www.R-project.org/posting-guide.html". Anecdotal evidence suggests that posts more consistent with that guide tend to get more useful replies quicker.
	  hope this helps.
	  spencer graves

Matteo Vidali wrote:

> dear R users,
> I need some help for multivariate analysis.
> I have 2 anaesthetic treatment groups (20 patients/group) where I
> register heart frequency and pressure for 60 min (repeated measures
> every 5 minutes). I would like to perform a test to check if treatments
> are different in controlling freq and pressures during the anaesthesia,
> but i would like to have also an overall measure and not only multiple p
> for different time intervals. I also think I should choose a test in
> which time is meaningful since the measures are not simple repeated
> measurements but measurements taken at specific time points.
> 1 million dollar question.... how to do in R?
> thanks in advance
>
> Dr Matteo Vidali
> Dep. of Medical Sciences
> University of East Piedmont "A. Avogadro"
> ITALY
>
>
> --
>
> ______________________________________________
> R-help@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

-- 
Spencer Graves, PhD
Senior Development Engineer
PDF Solutions, Inc.
333 West San Carlos Street Suite 700
San Jose, CA 95110, USA

spencer.graves@pdf.com
www.pdf.com <http://www.pdf.com>
Tel:  408-938-4420
Fax: 408-280-7915

______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Received on Sat Dec 17 10:55:03 2005

This archive was generated by hypermail 2.1.8 : Sat 17 Dec 2005 - 14:34:25 EST