GENERATIVE EXPLANATIONS OF CRIME: USING SIMULATION TO TEST CRIMINOLOGICAL THEORY
A Coggle Diagram about Computational Models, The development of criminal justice policy typically uses a top-down analytic strategy; broad trends of crime and offending are observed, and inferences are made about the underlying mechanisms and processes giving rise to them., Traditional statistical or equation-based explanations operate in a top-down manner, with the identification of correlations between aggregate observations used to make inferences about underlying mechanisms.
In contrast, generative social science operates from the bottom-up, identifying generative explanations as those hypothesized individual-level behaviors that produce macrolevel output patterns consistent with the observed regularities of the real-world system (the "target").
Applying the ABM as its principal scientific instrument, generative social science assesses the generative sufficiency of theory.
Generatively sufficient mechanisms are those that when employed by an agent population are sufficient to generate macrolevel patterns congruent with the real-world system and This study assesses whether the principal theories of environmental criminology provide a viable generative explanation of several commonly observed patterns of crime.