Reviewer Charles Seife considers my book Big Data, Little Data, No Data: Scholarship in the Networked World alongside two popular books on big data in commerce: Steve Lohr's Data-ism and Bruce Schneier's Data and Goliath (see Nature 518, 480–481; 2015). Yet the review does not make clear that mine is aimed at a very different audience.
I wrote the book for scientific researchers, scholars, librarians, publishers, policy-makers and other stakeholders, for whom the subtle uses of data as evidence in research are being swamped in the hype about big data. With case studies exploring how the idea of data varies in and between domains, I show how one researcher's data can be someone else's noise. Therein lies the rub.
Because so much about research data is open to personal interpretation, the information can be difficult to describe, represent and manage — and to share or reuse. The failure to understand these complexities leads to misguided policies for data management and to a lack of investment in both the workforce and the infrastructure for data curation. Ultimately, it can mean that no data survive for research.