Beetles: choice of link function
Dobson (1983) analyses binary dose-response data published by Bliss (1935), in which the numbers of beetles killed after 5 hour exposure to carbon disulphide at N = 8 different concentrations are recorded:
We assume that the observed number of deaths r
i at each concentration x
i is binomial with sample size n
i and true rate p
i. Plausible models for pi include the logistic, probit and extreme value (complimentary log-log) models, as follows
p
i = exp(
α +
βx
i) / (1 + exp(
α +
βx
i)
p
i = Phi(
α +
βx
i)
p
i = 1 - exp(-exp(
α +
βx
i))
The corresponding graph is shown below:
model
{
for( i in 1 : N ) {
r[i] ~ dbin(p[i],n[i])
logit(p[i]) <- alpha.star + beta * (x[i] - mean(x[]))
rhat[i] <- n[i] * p[i]
}
alpha <- alpha.star - beta * mean(x[])
beta ~ dnorm(0.0,0.001)
alpha.star ~ dnorm(0.0,0.001)
}
Data
list( x = c(1.6907, 1.7242, 1.7552, 1.7842, 1.8113, 1.8369, 1.8610, 1.8839),
n = c(59, 60, 62, 56, 63, 59, 62, 60),
r = c(6, 13, 18, 28, 52, 53, 61, 60), N = 8)
Inits for chain 1
list(alpha.star=0, beta=0)
Inits for chain 2
list(alpha.star=1, beta=1)
Results Logit model
Probit model
Extreme value (cloglog) model