The Ability of Engineers to Extract Requirements from Models

The Department of Defense is adopting model-based systems engineering in which models will replace the extensive amounts of documentation generated in developing a new system. This research examines how this shift from textual description of requirements to a model-based description will effect the requirements engineering process. Specifically, we ask whether engineers will be able to extract the same understanding of the system requirements from the models as they can from the traditional textual requirements specifications. This paper describes the theory and related work on the understandability of models and the performance of cognitive tasks such as requirements engineering. Our research into model representation is part of a larger effort on a theory of model relativity postulating that models affect how we think about the system of interest. In this paper, we present our exploratory research studies, discuss our research protocol, describe the research plan, and present the current status of our study.