Journal on Policy and Complex Systems
2 . How components of theories and systems give rise to patterns .
3 . Identifying critical ( necessary and sufficient — rather than all ) variables .
4 . Visually show how variables are related .
Simulations function as a tool to envision and deduce possible states from the map . Modelers can set the parameters of how variables are to be represented along with probability distributions to mirror real world population samples . Capitalizing on simulation capabilities , modelers can account for simultaneity of variable interactions , rule-based criteria from equation , theories , principles , assumptions , and / or a combination , showing actual , past , future , and counterfactual states ( Johnson , 2016 ).
Translating maps into simulations , whether simple or complex , is challenging . Conceptualizing , developing , implementing , and analyzing a simulation takes creatively , solid logic , and an effective map to match the model to social reality . Developing an effective simulation , to build theory , is process-oriented and an art that takes practice abstracting and identifying essential elements , along with system structures , attributes , agent actions , and system-wide actions within a social collective . Consider an interdisciplinary approach in order to gain feedback and input on possible model strategies .
Implementation of a simulation necessitates translating model specification into computer programing language . There are a variety of software programs that can model the simulation of networks , systems dynamics , scenario modeling , and agent-based modeling . Verification and validation are the next steps in model development to ensure there are no programming errors and ; does the model perform as intended and for the purpose of theory building ( Johnson , 2016 ).
Finally , the tools to analyze simulations are similar to those used in tradition research . Simulations can be programmed to include output plots , graphs , for individual simulation runs or numerous runs . Outcomes of simulation runs can then be compared and analyzed ( Johnson , 2016 ). Analyses can show patterns , trends , tipping points ,
192