AJPS Author Summary: Analyzing Computational Models

Author Summary by David A Siegel

AJPS Author Summary - Siegel

Computational models have been underutilized as tools for formal theory development, closing off theoretical analysis of complex substantive scenarios that they would well serve. I argue that this occurs for two reasons, and provide resolutions for each. One, computational models generally do not employ the language or modes of analysis common to game-theoretic models, the status quo in the literature. I detail the types of insights typically derived from game-theoretic models and discuss analogues in computational modeling. Two, there are not widely established procedures for analysis of deductive computational models. I present a regularized method for deriving comparative statics from computational models that provides insights comparable to those arising from game-theoretic analyses. It also serves as a framework for building theoretically tractable computational models that can build on prior work. The method relies on techniques that are commonly employed by empirical scholars, including simulation and numerical optimization, making its use straightforward for theory development. I illustrate this last point with an example of a computational model intended to provide theoretical expectations for a program evaluation. Together, these contributions should enhance communication between models of social science and open up the toolkit of deductive computational modeling for theory-building to a broader audience.

About the Author: David A. Siegel is Associate Professor in the Department of Political Science at Duke University.  “Analyzing Computational Models” will appear in a forthcoming issue of the American Journal of Political Science and is now available in Early View.

Speak Your Mind



The American Journal of Political Science (AJPS) is the flagship journal of the Midwest Political Science Association and is published by Wiley.

%d bloggers like this: