A new package `survey' for analysing complex survey samples is on CRAN. It handles stratification, clustering, and unequal sampling probabilities in descriptive statistics, glms, and general maximum likelihood fitting. The package is still under development: - it doesn't do the finite population correction to variances - it needs some real life worked examples
Most importantly, though, I don't do this sort of analysis routinely, so it's possible that some part of the interface is completely insane from the viewpoint of practising survey statisticians. Now would be an excellent time to complain. -thomas Thomas Lumley Asst. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle ______________________________________________ [EMAIL PROTECTED] mailing list http://www.stat.math.ethz.ch/mailman/listinfo/r-help