Usability based risk assessment model for sotware development process

Abstract

Software usability is an important factor in ensuring the development of quality and usable software. Ignorance, unawareness and failure to address usability during the software development process lead to poor quality software that is associated with potential usability risks. Risk management can be used to assess and control these usability risks. However, currently knowledge on usability risks is still insufficient and model to assess these risks is also lacking, leading to ignorance in managing usability risks in the software development lifecycle process (SDLC). This thesis proposes to develop a new Usability Risk Assessment Model to assessment of usability risks during the SDLC. Initially, elements of the Usability Risk Assessment Model were identified using Systematic Literature Review (SLR) whereby five major elements, namely, Risk Identification, Risk Analysis, Risk Prioritization, Risk Classification and Risk Mitigation were included in the model. Subsequently, feedback from 270 respondents of a survey questionnaire was utilized to identify 38 possible usability risk factors, which were then used to define 42 potential usability risks. These usability risks were used as keywords in identifying 85 initial usability vulnerabilities from the literature, which were grouped into four main categories that influence software development outcomes: Institutional Context, Software Project Content, People and Action, and Development Processes. The above usability risks and their vulnerabilities were then validated by four selected experts from the Public Sector. After validation, a total of 88 distinct usability vulnerabilities for various usability risks were identified. The usability risks were analysed using the Delphi method, involving seven experts to identify the probability of occurrences, impact on SDLC phases and mitigation plans for usability risks. Aided by the probability of occurrences and impact on SDLC phases, the usability risk exposure level was quantified, and used to classify and prioritize usability risks on SDLC phases. A Web-based Usability Risk Assessment Tool as a proof-of-concept was developed using ASP.Net to automate detailed elements in order to support the implementation of the model. Using this tool, multiple case study evaluations on four software projects in the Public Sector of Malaysia had demonstrated an inverse relationship between number of usability risks and usability of software. Thus, with the proposed Usability Risk Assessment Model, usability risks can effectively identified, analysed, prioritized, classified and mitigated during software development process to reduce these risks in order to enhance the usability of software. The contributions of this research are; first, a validated list of potential usability risks, usability vulnerabilities and possible mitigation plans for the usability risks; second, classification and prioritization of usability risks on SDLC phases; and third, empirically evaluated the Usability Risk Assessment Model.