Scenario:
You study asthma in a relatively homogeneous suburban area (the population
density, housing styles, etc. are effectively constant over the square
study area - 10 miles by 10 miles).
Along one side of the area a heavily travelled interstate
runs. Hypothesizing that emissions from heavy volume of vehicle traffic will
cause clustering of asthma cases nearer that road, you obtain maps of the
addresses of asthma suffering emergency room patients from the local
hospitals.
Questions:
- If the region is as uniform as suggested, is a Poisson model for the
distribution of cases a good one? How could you test a Poisson
model?
- Suppose traffic is contributing to increased incidence of asthma.
- What will be the consequences for the model? How might you
adjust a Poisson model to make it more appropriate?
- Would true or apparent contagion models be appropriate?
Return to Scenerios