The Quantitative Bias: How to Look at Lots of Books in Context

Abstract: This article discusses the phenomenon of “quantitative bias,” or the tendency to believe that numbers and data are more trustworthy due to their inherent empiricism, in the context of the quantitative turn in manuscript studies. It argues that all data is inherently constructed by the contexts and circumstances of its production and curation, and that in order to combat this bias, it is necessary for scholars to contextualize their data before presenting it. The first section of the article defines and identifies the quantitative bias, employing the example of a recent article that used biostatistics to measure manuscript loss rates, while engaging with recent scholarly work on implicit bias and its relationship with empiricism. The second section describes the interaction between a particular quantitative method, network analysis, and how quantitative bias can occur when constructing and analyzing data for networks. Using the example of the bibliographic and provenance networks relating to Syon Abbey (1415–1539), the sole English Birgittine house, this section illustrates not only the risk of quantitative bias, but also offers solutions for data contextualization that can be applied across any kind of quantitative bibliographic project dealing with lots of books.