One of the key ideas behind the "SCID project" http://cfpm.org/scid (Social Complexity of Immigration and Diversity) is that of starting with a simulation that reflects the evidence and what is known about the social model being modelled, however complex this makes it. *Then*, doing a simpler model of this complex model to get understanding of that model (and indeed models of the model of the complex model etc.). In this way we can stage abstraction rather than attempting it in one leap. This retains reference between the models, allows those doing the simplification free reign in what they simplify (under the constraint it has to match the significant results of the more complex model), and combines some of the advantages of rigour _and_ relevance.
This is described in three papers, one for each modelling stage (all Open Access):
Fieldhouse, E; Lessard-Phillips, L; and Edmonds, B. (2016) Cascade or echo chamber? A complex agent-based simulation of voter turnout. Party Politics. 22(2):241-256. DOI:10.1177/1354068815605671This describes a complex model of voter behaviour that includes all the processes thought to be relevant to whether people vote. This is the first stage of abstraction from evidence and data to descriptive simulation. The model itself can be found and downloaded from http://www.openabm.org/model/4368
Lafuerza LF, Dyson L, Edmonds B, McKane AJ (2016) Staged Models for Interdisciplinary Research. PLoS ONE, 11(6): e0157261. doi:10.1371/journal.pone.0157261(But please read the correction at the start since PLoS messed up the formatting and they don't fix the main paper after publication!_. A better formatted version is at: http://arxiv.org/abs/1604.00903)
This described the second stage of abstraction, from complex simulation to a simpler one that can be investigated more thoroughly. This reveals some hypotheses about the model that would not have otherwise have been discovered.
Lafuerza, LF, Dyson, L, Edmonds, B & McKane, AJ (2016) Simplification and analysis of a model of social interaction in voting, European Physical Journal B, 89:159. DOI:10.1140/epjb/e2016-70062-2Describes a further simplification stage from simpler simulation model to analytic model.