Managing fixity and fluidity in data repositories

Data repositories walk a fine line between the fixity and fluidity of the data they curate. Change is constant, but too much change affects the integrity of data. This paper examines data transformations in three repositories, serving the zoological, archaeological, and quantitative social science research communities. Based on in-depth analysis of 27 interviews, we identify a typology of changes: adding value; correcting errors; creating consistency; changing representations of data to reflect new knowledge; responding to designated communities; and evolving practices around collecting. Then we discuss the nature of these changes in terms of the data and collections. Our findings indicate that organizational differences and the diverse needs of the repositories' designated communities play a large role in how they manage change.