Dear R community, I am wondering what experimental design I am dealing with? I study the effect of daily air temperature on daily body temperature of the overwintering turtles (i.e. sleeping in soil). The model is a cosine wave with the air temperature as a covariate. Objects are 19 overwintering turtles. Data were combined over the three years. Each turtle was studied only once during these years, while 2 turtles were studied during two years (see table below). The study area was the same for these three studied years. Turtles were overwintering in different sites within this area. Is that correct to treat the study years as a random effect? I understand that I should incorporate autocorrelation model to account for daily variations in turtle’s body temperature. Is there a spatial correlation between the three study years, since the study was done within the same area? Study year turtle# 2005 year 1 2 3 4 5 6 7 2006 year 8 9 0 11 12 2006 year 13 14 15 4 16 17 18 12 19 Thanks for your help, Julia Internet Explorer 8 – Now Available. Faster, safer, easier. Download FREE now! _________________________________________________________________ Rediscover Hotmail®: Now available on your iPhone or BlackBerry Mobile2_042009 [[alternative HTML version deleted]]
Dieter Menne
2009-Apr-24 16:00 UTC
[R] what is this experimental design (mixed-effects model)?
J S-19 wrote:> > > I am wondering what experimental design I am dealing with? I study the > effect of daily air temperature on daily body temperature of the > overwintering turtles (i.e. sleeping in soil). The model is a cosine wave > with the air temperature as a covariate. > >Not sure if I fully got it, but check the "estus cycle" example in Pinheiro/Bates that looks similar, or the beaver example in 13.5 of MASS (probably less likely). Dieter -- View this message in context: http://www.nabble.com/what-is-this-experimental-design-%28mixed-effects-model%29--tp23218975p23219579.html Sent from the R help mailing list archive at Nabble.com.