Background: Preprints, a version of a manuscript that is self-archived and shared publicly (often) before publication in a journal, can be elusive and irregularly indexed which influences discovery. Evidence and best practice guidelines for the discovery of scholarly articles exist, but less guidance exists around discovery of preprints. The Cochrane Handbook recommends the inclusion of preprints and notes practical problems; individual preprint servers and preprint aggregators/databases may offer limited search features and/or download capabilities. Prior to this research two of the authors created a matrix for users to consult when navigating selected preprint servers as well as aggregators and databases. This living matrix focuses on the search functionality of each source to facilitate comprehensive searching.
Objectives: The authors believe investigation is needed to inform how and where to search for preprints to maximize coverage (which aggregators and databases contain which preprint servers) while balancing efficiency and cost factors. This study investigates these needs and will provide a comparison of eligible preprint-hosting aggregators and databases.
Methods: The authors applied eligibility criteria (e.g., hosts 2 or more preprint servers) to select aggregators and databases. Pre-set characteristics, features (e.g., bulk downloading), and search functionality details (e.g., ability to search by metadata field) were identified and tested for each aggregator and database in order to build a comparison table. Each criterion was designated as either essential or desired in achieving the goal of designing a comprehensive and reproducible search.
Results: Preliminary findings show there are several aggregators and databases wherein preprints are discoverable, but this coverage varies depending on discipline. The ability to design a comprehensive and reproducible search strategy in the aggregators and databases also varies. aggregators and databases differ in regard to limitations such as character or word limits of queries. Preprints are complex due to versioning and the potential existence of paired published journal articles. Analysis will be complete by the date of the presentation.
Conclusions: This study will help review teams make informed decisions about database and aggregator selection when targeting preprints. Guidance in the form of conducting and reporting standards should consider the complexities pertaining to preprint discovery.