By Laini Byfield

Ethics for the datasets that are too small to hide harm.

Small data is not low risk. In small systems, one incorrect merge, one delayed file, or one ambiguous metric can change a person’s payout, access, reputation, or sense of safety. Small Data Ethics treats these systems as moral infrastructure — not just technical pipelines.

ProportionalityContextExplainabilityRepair

Start here

Plain-language definition

Small Data Ethics is a values-driven approach to collecting, analyzing, and using human-scale data in ways that protect dignity, context, and trust — especially when people are identifiable and consequences are personal.

In small data, anonymity is often a myth. Technical choices become ethical choices.

How this site is organized

  • What is: definition, principles, and boundaries
  • Theory: foundational ideas you can cite
  • Policy: practical standards and guidance
  • Governance: committees, roles, and accountability
  • Practice: how to operate day-to-day
  • ETHICMAP: the implementation cycle

Attribution and term history

This site is authored by Laini Byfield. The earliest verifiable public use of "Small data ethics" found in this research pass is a 2021 call for abstracts for a University of Melbourne Digital Studio symposium. Earlier uses may exist in paywalled proceedings or unindexed documents.