Hello all, I'm trying to do a nested linear model with a dataset that incorporates an observation for each of several classes within each of several plots. I have 219 plots, and 17 classes within each plot. data.frame has columns "plot","class","age","dep.var" With lm(dep.var~class*age), The summary(lm) function returns t-test and F-test values evaluated as though I were working with 219*17-17=3706 degrees of freedom, when in fact I have but 219-17=202 df. I'm probably being dense on this one, but is there a way I can set df to the proper number so that summary.lm does the correct significance test? Or should I be doing an entirely different anlaysis? Thanks, -jlb -- ************************************ Joseph P. LeBouton Forest Ecology PhD Candidate Department of Forestry Michigan State University East Lansing, Michigan 48824 Office phone: 517-355-7744 email: lebouton at msu.edu
I believe your difficulties will be greatly enlightened by using either lme in library(nlme) or lmer associated with the lme4 package. Essential documentation is Pinheiro and Bates (2000) Mixed-Effects Models in S and S-Plus (Springer). hope this helps. spencer graves Joseph LeBouton wrote:> Hello all, > > I'm trying to do a nested linear model with a dataset that incorporates > an observation for each of several classes within each of several plots. > I have 219 plots, and 17 classes within each plot. > > data.frame has columns "plot","class","age","dep.var" > > With lm(dep.var~class*age), > > The summary(lm) function returns t-test and F-test values evaluated as > though I were working with 219*17-17=3706 degrees of freedom, when in > fact I have but 219-17=202 df. I'm probably being dense on this one, > but is there a way I can set df to the proper number so that summary.lm > does the correct significance test? Or should I be doing an entirely > different anlaysis? > > Thanks, > > -jlb