Many thanks indeed Bill. I understand now what you're getting at with the a/b-1!
Many thanks again. Graeme Duncan Bill.Venables wrote: > > The trick is to fit the model in a form which has the two separate > intercepts and the two separate slopes as the parameters. > > You do have to realise that a*b, a*b-1, a/b, a/b-1, ... all specify the > same model, they just use different parameters for the job. (Yes, > really!) > >> dat > ldose sex numdead > 1 0 M 0 > 2 1 M 3 > 3 2 M 9 > 4 3 M 16 > 5 4 M 18 > 6 5 M 20 > 7 0 F 0 > 8 1 F 2 > 9 2 F 6 > 10 3 F 10 > 11 4 F 11 > 12 5 F 14 >> dat <- transform(dat, Tot = 20) >> fm <- glm(numdead/20 ~ sex/ldose-1, binomial, dat, weights = Tot) >> coef(fm) > sexF sexM sexF:ldose sexM:ldose > -2.7634338 -3.4853625 0.7793144 1.5877754 >> dose.p(fm, c(1,3)) ## females > Dose SE > p = 0.5: 3.545981 0.3025148 >> dose.p(fm, c(2,4)) ## males > Dose SE > p = 0.5: 2.195123 0.1790317 >> > > In fact if you look at the book from which dose.p comes, you will find > an example not unlike this one done this way. At least I think that's > what I, er, read. > > W. > > Bill Venables > CSIRO Laboratories > PO Box 120, Cleveland, 4163 > AUSTRALIA > Office Phone (email preferred): +61 7 3826 7251 > Fax (if absolutely necessary): +61 7 3826 7304 > Mobile: +61 4 8819 4402 > Home Phone: +61 7 3286 7700 > mailto:[EMAIL PROTECTED] > http://www.cmis.csiro.au/bill.venables/ > > -- View this message in context: http://www.nabble.com/Finding-LD50-from-an-interaction-Generalised-Linear-model-tp15436597p15452544.html Sent from the R help mailing list archive at Nabble.com. ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.