Investigating the potential impact of values on requirements and software engineering

This paper describes an investigation into value-based software engineering and proposes a comprehensive value taxonomy with interpretation of design feature implications. The value taxonomy is used to assess the design of Covid-19 symptom tracker applications, contrasting the UK's NHS phase 1 and 2 designs which adopted centralized, then decentralized, architectures. The value/ feature analysis is also applied to the King's/Zoe Covid app which does not detect proximity, instead relying on user self-reporting. Value analysis illuminated design choices but was insufficient to account for download acceptance of the apps. We argue that motivational cost-benefit analysis needs to complement a values-based approach.