Over the last decade, research on the privacy of user information has shown that often a) ordinary users pay little attention to privacy policies and b) when considering policies, people have a hard time understanding their meaning and practical implications. Usable computational solutions to this problem have been developed. We observe a similar trend with respect to the ethics, norms and regulations for using public digital data at any scale; big and small. By this we mean that researchers may have little awareness of the different types of regulations beyond IRBs that might apply to their work, and difficulties to fully comprehend and implement applicable rules. This article focuses on practical issues with properly using social trace data for research and proposes solutions. For the purpose of this article, we define publicly available social trace data as information about people interacting with a) other social agents (e.g. social networks data from Facebook and Twitter), b) pieces of information (e.g. product review sites and discussion forums), and c) infrastructures (e.g. people checking in to places, geolocation services), and natural language text data (e.g. the content of posts and tweets) that all can be collected without intervention or interacting with users (also called passive measurement).

Download