Machine-Level Policy Implementation by Data Managers and Data Scientists, and the Impact on Digital Stewardship: A Mixed-Methods Content Analysis

This study examined how data managers and data scientists address digital stewardship with regards to the data repositories they manage and use, by determining what they may do at the machine-level with regards to human-readable and human-implemented preservation policies. The main significance of the research was that no existing studies have examined the actual implementation of these standard preservation policies at the machine level with the intent of demonstrating whether or not a particular preservation system can “prove” that it is actually implementing those standards. It also contributed to general knowledge about what kinds of policies tend to be implicit (e.g., authentication) and implemented in machine code, and, which policy types tend to be explicit (e.g., planning) and are implemented manually, outside the preservation system. Thus far, researchers and practitioners have not taken a “bottom-up” approach and examined what data managers and data scientists are actually doing with the data repositories and content they manage and examined the results against recommended practices. This study addressed this gap.