Although there is a vibrant discussion on the “replication crisis” among statisticians, this discourse has been largely going on in parallel with STS research on the same issue. Unlike applied scientists and STS researchers, statisticians are less interested in structural/incentive issues and the difficulty in conceptualizing different types of replications. Instead, their main concern is widespread misconceptions and misuse of statistical tools that result in invalid and unreliable research findings, which statisticians argue are the underlying cause of the “replication crisis”.

While the “replication crisis” seems like a recent development, similar corrective efforts in statistics can be traced to at least the 1960s. Such efforts tend to center around Null Hypothesis Significance Testing (NHST), whose dominance in applied sciences is reflected in method courses across disciplines. Pointing to human cognitive limitations in understanding uncertainty as the root of the problem, statisticians have proposed ways forward through statistical education and changes in professional communication protocols. However, due to its technical nature, this discourse has received little attention outside of the statistical circle and its potential for sparking system-wide changes remained unfulfilled.

For statistical reformers, the challenge is to recruit allies across technoscience while steering around thorny disciplinary-specific issues. In this light, STS has much insight to offer on the interplay between the social and technical aspects of technoscience. This paper reviews the relevant literatures and identifies potential areas of cross-fertilization across these two bodies of research.