Analysis and Modelling Techniques

Research Question: How can we improve software engineering (SE) practices, in industry and education, with tools and techniques developed from the systematic study of software related stories?
With high profile technology stories and scandals in the media, the software engineering (SE) community is under increasing scrutiny with demands for better practices and even regulation to demonstrate we are mindful of the wider ethical, social and human impact of the systems we design and build. Can we design approaches to explore alternative perspectives on SE stories that will help software professionals reflect on and anticipate the impact and consequences of technical decisions in their own work and working practices?

This project is a systematic study of the perceptions of the roles, responsibilities and activities of technology, software engineering teams, end users and wider society in software related stories. It will pilot processes to capture, document and analyse the understanding and interpretation of software stories by software practitioners, end-users and wider society. In particular to look at how a software engineer's understanding of a story compares to that of a non-software engineer. What can we learn from the similarities and differences, and how does using stories help us with this understanding? Research suggests that stakeholders in a SE project may have differing mental models regarding the design and use of the software, and that these differing mental models could lead to software vulnerabilities that if exploited, through misunderstanding or misuse, can result in unintended consequences and potentially negative outcomes. If software engineering teams explore the unintended consequences of their code, products or services through the use of stories, how might this impact their future design and coding practices? The research is particularly concerned with aspects of privacy, trust and security of software systems and their impact on end users and wider society.

The following research activities will be conducted
1. Story exploration, with a broad range of participants, of carefully selected public software stories (e.g. Volkswagen emissions scandal, WannaCry, Tesla car crash, or the Facebook-Cambridge Analytica story).
2. Story capture and creation - capturing and creating software stories from the different perspectives of roles in SE teams and other non-developer stakeholders.
3. Based on iterations and analysis from 1 and 2, to co-design and pilot new tools and techniques to support the presentation and exploration of stories as a learning approach.

The following methods will be used to run and analyse these activities:
1. Qualitative and Quantitative methods: Focused discussion approaches will be designed and piloted as a method to structure interviews and workshops that explore facts, feelings and insights about how stories impact self, team, organisation and wider society (qualitative data). Questionnaires (quantitative and qualitative data) will be designed to measure the impact of both the session approach and the stories used.
2. Q-methodology to investigate participant and team perspectives on a theme (e.g. reflections on a project or story). Participants rank a series of statements, represented as physical cards, whilst discussing their choices. Two outputs: a narrative (qualitative) and a digital image of the resulting ranked position of q-sort cards (quantitative). Q-Sort data is analysed using an appropriate statistical package to look for common factors.
3. Thematic analysis will be used to identify and analyse patterns in the qualitative interview, workshop and Q-Sort data.