For many causal claims there exist overwhelming amounts of – often discordant – evidence. As a remedy, different approaches for amalgamating such evidence have emerged . Examples are meta-analysis and qualitative reviews. Evidence amalgamation is omnipresent in the empirical sciences, especially in the psychological sciences, the social sciences and the biomedical sciences. It is often deemed the height of objectivity and it has an important impact on e.g. policy and medical practice. Still, apart from a few critical voices it has received little attention in philosophy of science so far.
The aim of this project is to analyse the epistemic status of different approaches to evidence amalgamation to see whether and when their outcomes are reliable.The following cases will serve as the starting point: the Cochrane Collaboration in evidence-based medicine (EBM), IARC in epidemiology and cancer research, the Campbell Collaboration in the social sciences, and reviews in the psychological sciences (e.g. in the Psychological Bulletin).
These cases will be analysed and compared from the point of view of three philosophical topics and related research questions. For each of these topics there exists a fascinating, topical philosophical literature which explicitly seeks contact with scientific practice in the aforementioned scientific disciplines. (a) What is the role of mechanistic evidence in these approaches, and what should it be? (b) Should we give priority to evidence from randomized controlled trials? What do we risk or gain by also including non-randomized studies? (c) Sometimes the 'Principle of Total Evidence' is used to criticize the practice of evidence amalgamation. But what does it mean and what version applies best here?
The answers to these questions will help to see how the diverse approaches to evidence amalgamation work, what they can learn from each other and how they can be improved.
Evidence amalgamation is very important, but its philosophical analysis is mostly virgin territory. The aim of this project is to contribute to its exploration.