Towards functional change decision support based on COSMIC FSM method

Abstract

Abstract Context: Managing requirements change is a central issue in the software development industry. In fact, inappropriate decisions about a change request may jeopardize the project development progress by going over budget/time or delivering a software with functional requirements that do not fully meet the user’s needs. Hence, a change decision support is required for the success of the software development. Objective: This paper has a three-fold objective: (i) explore the applicability of the ISO standard COSMIC FSM method to evaluate a change request; (ii) investigate the use of estimation models to predict the effort required to handle a functional change and its impact on the initially estimated software development effort; and (iii) propose a decision support method that offers the appropriate information for the change advisory board members to decide whether to accept, deny or defer a functional change request. Method: To guide the decision on a change request, the method proposed in this paper accounts for the most important factors when evaluating a change request, namely the functional change status, the preference of the change requester, and the effort required to handle the change. The functional change status is identified based on the sensitivity of the changed functionality and the functional size of the functional change. The functional change effort can be estimated using several ways including the COCOMO II model, the Simple Linear Regression Model and expert judgment. Furthermore, this paper proposes a prototype to determine automatically the functional change status and offers pertinent information that the change advisory board can use to determine how to handle a change request. The use of the decision support method and tool is illustrated through three case studies. Results: A decision support method to help decision-makers respond to a functional change request is provided. This method takes into account the functional change status, the preference of the change requester and the functional change effort. The empirical evaluation of the proposed method is illustrated through three case studies. The role of experiments here is primarily to provide a proof-of-concept rather than an exhaustive evaluation. Conclusion: Using COSMIC FSM method, it is possible to identify functional changes leading to a potential impact on the software development progress. Based on the evaluation of the functional change, the change advisory board members can make judicious decisions about whether to accept, defer or deny a functional change request.