A Method for Anticipation of Undesirable Interactions in Software for a Digital Society informed by a Thematic Analysis of Discovery Practice

Abstract

This research explores current user experience design practice in the IT sector through empirical studies with practitioners. The focus is how interactions that are undesirable are identified, because they are contrary to the interests of the users. The practice area of interest is the discovery stage when designers are working to understand the user’s aims and identifying opportunities to achieve the desired outcomes.

Two research questions are explored: what methods are used in current software design practice to identify undesirable interactions during discovery activities, and how can designers be helped to structure their work in a way that assists them in identifying undesirable interactions.

Three empirical studies were conducted with user experience practitioners. The first used Ketso workshops to gather data on discovery goals, practices, and challenges. These informed the second study, which used interviews to gather data on attitudes and practices. Reflexive thematic analysis was used to analyse findings. Using findings from the first two studies and lessons from the existing literature, I developed a new method of anticipating undesirable interactions by identifying ethical properties that the design should preserve and considering how they might be lost. This Jeopardy Analysis method was evaluated in the third study through remote workshops with user experience design practitioners who were asked to apply it to an unfamiliar scenario and provide feedback on its use.

Findings about current practice from the first two studies indicate that user experience practitioners favour methods that build a shared understanding, but select them to suit the context. They tailor their approach, and actively explore and experiment with new methods. There was some recognition of the need to anticipate problems, but no methods were applied at the discovery stage, instead relying on usability testing.

The evaluation of the Jeopardy Analysis method found that it helped to challenge assumptions. Practitioners found framing the problem in ethical terms unfamiliar and difficult, but felt they could use it by themselves with more practice. The generic properties used for the evaluation were found to be too abstract, so the method step tailoring them for the domain would be an important part of its application.

The research contributes insights into the goals practitioners have for their discovery activities, and their current approaches to identifying undesirable interactions. It identifies practitioner interest in recent ‘consequence scanning’ approaches to anticipating problems that differ from current practice, and are associated with a more risk averse mindset. It contributes a novel Jeopardy Analysis method, and reports encouraging results from its initial evaluation.

Further work is needed to refine Jeopardy Analysis for use in industry, and to evaluate practitioner selection of ethical properties tailored to their domain and product. Its natural domain of use is seen as software applications supporting life in our increasingly digital society, where the general public are co-opted into our designs, and the ethical case for intervention is most compelling. Extension of Jeopardy Analysis to involve prospective users in co-analysis and design would further address the potential imbalances of power in current practices. It is suggested that teaching Jeopardy Analysis in higher education settings would contribute to learning outcomes in inclusive design, societal impact, the making of ethical choices, risk management, and the recognition of responsibilities.