In Praise of Small Data: When You Might Consider N-of-1 Studies

The era of big data is well upon us, and understandably, there is much excitement over what we can learn from so much data and new big data analyses. However, big data often comes with big challenges, such as issues around data privacy and retention and resource constraints like battery life and storage. Moreover, big data may often not actually be able to tell us what we want to know. Many of the analyses based on big data may tell us only correlations, not causes. Or, they may only tell us about how a large group of people behaves, not necessarily an individual. In this article, I'll describe some reasons why you might want to take a small data approach, sometimes called Single Case designs or N-of-1 studies, how you might do it, and provide an example of a mobile app we built that applied these designs for self-experimentation with gastrointestinal health.