All simulations bear the problem of how to assess uncertainty. The way chosen here is via internal validation, another could be via behavioural analysis. Does SIENA behave in the same way as real world
cities would? This can be assessed by carrying out analysis with SIENA and comparing the results with real-world studies.
Behavioural analyses do not only verify that SIENA provides results of realistic magnitude, they also require that the statistical and spatial distribution of the underlying components and relationships are simulated realistically.