A Framework for Evaluating Inductive Models of Software Development
Scott Serich · Journal of the Association for Information Systems · 2001
This study introduces a framework for evaluating decision models in organizations that conduct custom software development. The framework takes the form of a metamodel into which decision models can be embedded and assessed. In response to the turbulent, heterogeneous task environments facing software firms, the framework targets each model’s self-adaptive or inductive features for analysis. The evaluation mechanism is comprised of homomorphisms from abstract algebra and the transition function, observability and controllability features of control systems theory. The meta-model is tested on three candidates, two static models and a dynamic model based on Simon’s behavioral model of rational choice. It correctly distinguishes the former models as having weak induction features and the latter as being strong on this aspect.