A theory of requirements documentation situated in practice

• Authors: Norah Power;Tony Moynihan
• Venue: Proceedings of the 21st annual international conference on Documentation
• Year: 2003

This paper presents a theoretical framework which attempts to explain the variety of styles of requirements documentation found in practice in relation to the variety of situations in which software products and systems are developed. It identifies situational factors that might be useful to categorize development situations from the point of view of the requirements documentation process. This framework is in contrast with much of the literature on requirements engineering, which takes a very prescriptive approach to documentation, and which takes very little account of the situation of use. The research was based on a qualitative study of requirements practitioners and their documentation practices. The empirical data collected from interviews in the study was systematically analyzed using the grounded theory method and a computer-based tool, ATLAS.ti.The framework is in three parts. The first part is an analysis of requirements documents as texts, categorizing the different constituent elements that might be used to specify requirements. The second part is a scheme for classifying system development situations with respect to the requirements documentation process. The third part of the framework takes each of these situation types and matches it with an appropriate style of requirements document that is found to be typical in that situation. The aim of the paper is to explain (or structure an explanation of) the diverse ways that system and software requirements are documented in practice. To date, this diversity has not been examined by any empirical study or theoretical framework. Most requirements specification techniques are regarded as application-independent, assumed to be context-independent, and presented as if they were universally applicable. Requirements specification standards follow a similar line. The paper concludes that standard prescriptive approaches have failed to identify the necessary and sufficient contents and style of a requirements document, because what is good enough in one situation may not be desirable or acceptable in another.