Values in Computational Models Revalued
Computational models are mathematical representations that are designed to study the behaviour of complex systems. Systems under study are usually nonlinear and complex to the extent that conventional analytics cannot be used. Scholars have tried to establish the role played by trust and values in the use of such models in the analysis of public administration.
Public decision-making is itself a complex endeavour that involves the input of multiple stakeholders. Usually, there are a lot of conflicting interests that influence the final outcome of such decision-making processes (Klabunde & Willekens, 2016). In a computational model, a number of factors equally influence the outcome of the process. One of them is the number of actors involved the presence of more actors normally implies increased mistrust. Another factor is the amount of trust that already exists among the decision makers. In cases where the group is homogenous, there is likely to be more trust and thus, less concern about the number of actors involved.
Given the importance of these two factors, the designer of any such model bears the largest burden in assuring the value of the model. He or she can choose to implement agency by humans or by technology depending on the number of actors and trust among them. Also, model designer determines the margins of error from each scenario while modelling (Gershman, Markman & Otto, 2014). Since in conventional decision-making processes different actors have different roles, the model designer may decide to accord different levels of authority to different actors. Nevertheless, they must ensure that such a decision does not affect the trust of the system. Overall, what values are sought from a computational model in a public decision-making context?
Gershman, S. J., Markman, A. B., & Otto, A. R. (2014). Retrospective revaluation in sequential decision making: A tale of two systems. Journal of Experimental Psychology: General, 143(1), 182-194.
Klabunde, A., & Willekens, F. (2016). Decision-making in agent-based models of migration: state of the art and challenges. European Journal of Population, 32(1), 73-97.