Introduction

Say that we measure radon concentration in 100 houses in a given county. It is reasonable to assume that perhaps houses that are closer to each other will have similar concentration of this gas when compared to houses that are farther apart.

Steps

To keep things simple, we will assume that no covariates are available to model the mean. As a result, the mean will be represented by a single parameter \(\mu\). Aside from \(\mu\), the goal is also to estimate the overall variance \(\sigma^2\) and the exponential decay parameter \(\rho\).

1. Simulate data with spatial correlation. To this end, we will have to create the coordinates of each house to then calculate the distance between each pair of houses.

2. Create the geospatial model in JAGS. Does it work? Are we able to accurately estimate the parameter values used in the simulated data?

Back to main menu