John P. Burkett
2008-Mar-25 19:16 UTC
[R] quasi-Poisson regression in stats glm vs aod quasipois
For quasi-Poisson regression, I have tried to compare the output of the quasipois function in the aod package to the output of the glm function with family=quasi(link="log", variance="mu^2") in the stats package. The standard errors are about 32.5 times larger in the former than in the latter. If anyone can explain the difference, I would be most grateful. I am running R version 2.6.1 and aod version 1.1-24 under Gentoo Linux. In case they are of diagnostic value, I am attaching my R code (compare.R), my data set (visitsbymonth1990_2006.csv), and my output file (compare.Rout). Best regards, John -- John P. Burkett Department of Environmental and Natural Resource Economics and Department of Economics University of Rhode Island Kingston, RI 02881-0808 USA phone (401) 874-9195 -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: compare.R Url: https://stat.ethz.ch/pipermail/r-help/attachments/20080325/45c90777/attachment.pl -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: compare.Rout Url: https://stat.ethz.ch/pipermail/r-help/attachments/20080325/45c90777/attachment-0001.pl