I am working with data that have values below the limit of detection (otherwise known as left-censored data). I used dinterval to fit my model to estimate the parameters. Below is an example

model {

for (i in 1:N) {

above.lod[i] ~ dinterval( x[i] , llodVec[i] )

x[i] ~ dnorm( mu , tau)

}

...

}

My question is how does jags estimate the censored values or solve this model involving censored value. In other words, what goes on behind the scene?

Thank you