Examining expectations about user involvement in software development and factors that influence high quality user involvement
In contemporary software development, frequent user engagement throughout the development process is commonly viewed as good practice, leading to increased development productivity and user satisfaction with the product. Contemporary Agile software development methodologies, as adopted by many practitioners, promote such frequent and timely involvement of users. The quality of this user involvement may be variable in its contribution to the software project, however. While there are many factors that may influence the quality of the user involvement, in this thesis it is posited that the degree of alignment or misalignment of expectation of user involvement among the development and user groups is an important factor in the quality of user involvement. There is little documented research in this alignment of expectations of user involvement, and an investigation of this in practice is the basis of this thesis. Furthermore there is no consistent meaning to the notion of “quality of user involvement” in related literature and so another aim of the thesis is to get a deeper understanding of the meaning of “quality” in user involvement.
The scope of the investigation presented in this thesis is to develop an instrument for comparing expectations and use this to identify patterns of alignment in expectations by role. The linking of alignment of expectations to project success is beyond the scope of this investigation. A framework for conceptualising “quality” in the context of user involvement, as well as characterising significant barriers and enablers to high quality user involvement, are also goals of this research.
The approach to the investigation is based on a systematic, role-by-role comparison of user involvement expectations using Repertory Grid techniques. The standardised grids of constructs and elements related to user involvement were constructed from a synthesis of relevant current literature. Three grids were used for comparison of expectations by role: the nature of the involvement, the modes of communication associated with the involvement, and the charactersitics of the user that influenced user involvement. In addition to the Repertory Grid data, interview data were gathered from participants through a series of semi-structured interviews. Thematic analysis of the interviews was used to develop conceptual frameworks related to the notion of “high quality user involvement” as well as to identify the significant enablers and barriers to high quality user involvement.
Two Agile software development organizations based in New Zealand, were the case organizations for this study. A total of nineteen interviews were conducted with technical roles including Developer, Project Manager, Business Analyst, Tester and user roles such as the Product Owner and Subject Matter Expert.
The results show no strong misalignment of expectations of user involvement across both the user roles and development roles for the case organizations studied, and some interesting patterns were uncovered. Some significant barriers and enablers of high quality user involvement are also identified, providing some useful insights to guide the design of future techniques and tools to support high quality user involvment.