Accepted Paper:

Power: statisticians, authority and the genetics of common complex diseases  

Author:

Catherine Heeney (University of Edinburgh)

Paper short abstract:

Lack of replication of gene/phenotype associations in the early 2000s threatened funding for research into common complex diseases. Frustration with existing approaches coupled with technological optimism opened the door to a different type of statistical expert to enter the genetics community.

Paper long abstract:

2007 was considered the year of the Genome Wide Association Study (GWAS). This paper considers how frustration with both the candidate gene approach and family linkage studies drew a new type of statistician into the field of genetics and common complex disorders. Epidemiologists trained in looking at predisposing factors for disease in populations, were key in providing solutions for some of the difficult problems being identified by scientists in the field. The Wellcome Trust Case Control Consortium (WTCCC) which published a highly influential GWA Study in 2007 will be used as an empirical focus to trace key epistemic debates. Providing the statistical authority or power calculations to demand funding for large sample sizes to look for genes of small effect, was one part of the role of epidemiologists. However, in order to ensure the conditions for GWAS, this type of statistician also had a number of other maintenance tasks. These included convincing statisticians in already in the field of genetics that they had the tools to deal with the various types of bias that were presumed to exist in a population wide study and shoring up the gaps in performance of microarray chip technologies. Drawing on ANT, scientific literature and interviews with key players, it is argued that in the WTCCC assemblage, scientific authority ultimately rested on any number of alliances between technologies, funders and scientists but central to all of these was that between epidemiologists and geneticists.

Panel T033
Who is in, who is out? Exploring collectives in health research