Do User Backgrounds Matter? Impact of Gender and Countries on App Reviews and Tweets

Previous research found that socio-demographic factors, like national culture or gender have an impact on the frequency and type of given user feedback for software requirements. However, few exploratory studies have analyzed this impact and focused their analyses on individual factors separately. In this study, we add knowledge about the influence of socio-demographics on given user feedback. Our work investigates the effect of two independent factors, gender and country of residence, on user feedback characteristics from Twitter and Apple App Store reviews. We analyzed 2,911 tweets and reviews from different countries, languages, and apps using statistical methods. Our analyses show that (1) gender significantly influences the content of user feedback in tweets such as women submitting more bug reports than men, (2) users' gender also has an interaction effect with the country of residence on Apple App Store reviews' and Twitter content, (3) users' country of residence impacts the text length and ratings in app reviews, and (4) users' country of residence also influences the sentiment and content in tweets and reviews. Our work raises awareness of potential biases in automatic approaches used during software development and evolution that process diverse user feedback and highlight the necessity to include users with (intersecting) marginalized identities during software evolution.