Characterizing Human Aspects in Reviews of COVID-19 Apps

Abstract

To successfully satisfy user needs, software developers need to suitably capture and implement user requirements. A critical and often overlooked characteristic of user requirements are 'human aspects', which are personal circumstances affecting the use of software (e.g., age, gender, language, etc.). To better understand how human aspects can impact the use of software, this work presents an empirical study focusing on app reviews of COVID-19 contact tracing apps. We manually analyzed a dataset of 2,611 app reviews sampled from the reviews associated with 57 COVID-19 apps. To analyze the reviews, we performed qualitative and quantitative analyses. The analyses characterize the human aspects contained in the reviews and investigate whether the apps suitably address the human aspects. We identified 716 reviews related to human aspects and grouped these into nine categories. Of these 716 reviews, 8% report bugs, 14% describe future/improvement requests, and 22% detail the user experience. Our analysis of the results reveal that human aspects are important to users and we need better support to account for them as software is developed.