Managing and analysing software requirements changes

Submission note: A thesis submitted in total fulfilment of the requirements for the degree of Doctor of Philosophy (by published work) to the School of Engineering and Mathematical Science, College of Science, Health and Engineering, La Trobe University, Bundoora, Victoria.

Throughout the development of a software system, there is a tendency for new requirements to emerge and existing ones to change. The management of changing requirements during the requirements engineering process and system development is known as requirements change management (RCM). The literature suggests that unmanaged changes can cause budget and time overruns and lead to poor quality products, which will not satisfy customer requirements. In order to address some of the existing issues of RCM, this thesis aims to create a better understanding of RCM and presents a requirements change management process (RCMP) that encompasses change identification, analysis and rework which will be useful in providing a more rounded solution. RCM has many aspects that have not been explored and/or understood in depth. A systematic review presented in this thesis brings together research relevant to RCM, providing a holistic picture that encompasses the causes of requirements changes, current issues, solutions provided and the existing knowledge gaps. In the RCMP, change identification is accomplished by a change specification method and a change classification method. The outcome of these methods results in less communication ambiguities and a better understanding of the need for the change. A method of requirement change analysis is developed to identify how requirement changes propagate through the existing system design and also to identify the system activities which are affected due to the changes. The third method presented as part of the RCMP is to analyse the extra work /rework required to implement a requirements change. The work presented gives a clear understanding of rework in the context of RCM with an assessment of the rework calculated using the interactions caused by the changes with the system activities and their connections. To demonstrate the usefulness of the methods, several methods are applied to a running example which explains the application of the methods step by step. This allows for a better understanding of the mechanics of the methods. The demonstration is further extended by applying all the methods to a larger case studyThroughout the development of a software system, there is a tendency for new requirements to emerge and existing ones to change. The management of changing requirements during the requirements engineering process and system development is known as requirements change management (RCM). The literature suggests that unmanaged changes can cause budget and time overruns and lead to poor quality products, which will not satisfy customer requirements. In order to address some of the existing issues of RCM, this thesis aims to create a better understanding of RCM and presents a requirements change management process (RCMP) that encompasses change identification, analysis and rework which will be useful in providing a more rounded solution. RCM has many aspects that have not been explored and/or understood in depth. A systematic review presented in this thesis brings together research relevant to RCM, providing a holistic picture that encompasses the causes of requirements changes, current issues, solutions provided and the existing knowledge gaps. In the RCMP, change identification is accomplished by a change specification method and a change classification method. The outcome of these methods results in less communication ambiguities and a better understanding of the need for the change. A method of requirement change analysis is developed to identify how requirement changes propagate through the existing system design and also to identify the system activities which are affected due to the changes. The third method presented as part of the RCMP is to analyse the extra work /rework required to implement a requirements change. The work presented gives a clear understanding of rework in the context of RCM with an assessment of the rework calculated using the interactions caused by the changes with the system activities and their connections. To demonstrate the usefulness of the methods, several methods are applied to a running example which explains the application of the methods step by step. This allows for a better understanding of the mechanics of the methods. The demonstration is further extended by applying all the methods to a larger case study.

Thesis with publications.