[R-sig-ME] Weights in lmer

Roberts, Kyle kyler at mail.smu.edu
Fri Sep 19 18:27:47 CEST 2008

Dear All,

Still trying to use the weights function with the ECLS (Early Child Longitudinal Study) with lmer. I think that I have identified the problem, but I am not sure. If a person has a "0" for a weight (they don't have all years of data), how does lmer handle this case? Also, how is the weight treated when you have a stacked format (measurement occasions nested inside students)? For 4 measurement occasions, is the weight multiplied by each of the 4 time points, or does lmer somehow apply the weight at the student level (or does it matter)?


Dr. J. Kyle Roberts
Department of Literacy, Language, and Learning
School of Education and Human Development
Southern Methodist University
P.O. Box 750381
Dallas, TX  75275

More information about the R-sig-mixed-models mailing list