“Background: Research is published in indexed, online scholarly journals so that published knowledge can be easily found and built upon by others. Most scholars rely on relatively few online indexing service providers to search for relevant scholarly content. It is under-appreciated that the quality of indexing can vary across different journals and that this can have an adverse effect on the quality of research.
Objective: In this short paper I compare the recall of commonly used online indexers; Google Scholar, Web of Knowledge, Scopus, Microsoft Academic Search and Mendeley Search against a selection of over 20,000 papers published in two different high-volume journals: PLOS ONE and Zootaxa.
Results: When using Google Scholar, content in Zootaxa has low recall for search terms that are known to occur in it, significantly lower than the near-perfect recall of the same terms in PLOS ONE. All other indexers tend to have lower recall than Google Scholar except Scopus which outperformed Google Scholar for recall on Zootaxa searches. I also elaborate why Dark Research is undesirable for optimal scientific progress with some recommendations for change.
Conclusion: This research is a basic proof-of-concept which demonstrates that when searching for published scholarly content, relevant studies can remain hidden as ’Dark Research’ in poorly-indexed journals, even despite expertise-informed efforts to find the content. The technological capability to do full text indexing on all modern scholarly journal content certainly exists, it is perhaps just publisher-imposed access-restrictions on content that prevents this from happening.”