Categorizing methods and approaches for generating and identifying paradata

Documenting the processes and practices of making and processing research data has been identified as key prerequisite of data reusability and intelligibility. A large number of methods and approaches for generating and identifying such information have been proposed, however, dispersed across the literature. Consequently, the current understanding of what types of approaches have been envisioned, how they differ and relate to each other, and what kind of paradata they produce is limited. This paper reports an initial study to increase understanding of the methods landscape through review and categorization of paradata generation and identification methods. We identified three major temporal categories of (1) prospective, (2) in situ, and (3) retrospective methods and approaches, and five categories of paradata artifacts generated: (1) structured metadata, (2) narratives, (3) snapshots, (4) diagrammatic representations, and (5) standard procedures.