VGAM (version 0.8-3)

lirat: Low-iron Rat Teratology Data

Description

Low-iron rat teratology data.

Usage

data(lirat)

source

Moore, D. F. and Tsiatis, A. (1991) Robust Estimation of the Variance in Moment Methods for Extra-binomial and Extra-Poisson Variation. Biometrics, 47, 383--401.

Details

The following description comes from Moore and Tsiatis (1991). The data comes from the experimental setup from Shepard et al. (1980), which is typical of studies of the effects of chemical agents or dietary regimens on fetal development in laboratory rats.

Female rats were put in iron-deficient diets and divided into 4 groups. One group of controls was given weekly injections of iron supplement to bring their iron intake to normal levels, while another group was given only placebo injections. Two other groups were given fewer iron-supplement injections than the controls. The rats were made pregnant, sacrificed 3 weeks later, and the total number of fetuses and the number of dead fetuses in each litter were counted.

For each litter the number of dead fetuses may be considered to be Binomial($N,p$) where $N$ is the litter size and $p$ is the probability of a fetus dying. The parameter $p$ is expected to vary from litter to litter, therefore the total variance of the proportions will be greater than that predicted by a binomial model, even when the covariates for hemoglobin level and experimental group are accounted for.

References

Shepard, T. H., Mackler, B. and Finch, C. A. (1980) Reproductive studies in the iron-deficient rat. Teratology, 22, 329--334.

Examples

# cf. Figure 3 of Moore and Tsiatis (1991)
plot(R/N ~ hb, data = lirat, pch = as.character(grp), col = grp,
     las = 1, xlab = "Hemoglobin level", ylab = "Proportion Dead")