Isabelle, I don't think any statistical method will do you much good if you only have four subjects.
If you have a larger sample size such that some general inference about a treatment effect is reasonable, here's two references worth checking out: Gelman and Hill. 2007. Data analysis using regression and multilevel/hierarchical models. Cambridge. Pinheiro and Bates. 2004. Mixed-effects models in S and S-PLUS. Springer. -- Dave Hewitt > Date: Thu, 11 Jun 2009 17:54:35 -0400 > From: =?iso-8859-1?Q?Isabelle-Anne_Bisson?= <[email protected]> > Subject: Mixed Linear Models > > Fellow Eco-loggers, > > I would greatly appreciate your input on a statistical question that I > have. My study design is as follows: > > Four subjects receive NO treatment on day 1 and a treatment on day 2. The > treatment is of 4 hours. I take a measurement (e.g., blood pressure) every > minute for each subject on both days. I would like to test for the effect > of treatment while taking into account change over time. Are 'linear mixed > models' analyses appropriate for this type of data set? > > I will gladly provide more information if need be.
