Barbara Osimani and Jürgen Landes gave talks at Issues in Medical Epistemology Cologne (14-16 Dec, 2017): Barbara Osimani delivered the opening keynote lecture on ”Nature, noise and evidence in medicine”. Jürgen Landes presented on “Variety, Reliability, Confirmation and Drug Safety”.
Abstract of Osimani:
With respect to other scientific ecosystems medicine is characterised by the joint interaction of the following factors: 1) medicine is an intrinsic interdisciplinary science: it does not investigate a specific level of reality like physics, or biology, but rather works across levels; 2) this leads to a stronger error propagation with regard to causal inference and extrapolation, as well as to explanation and intervention, and therefore to stronger epistemic uncertainty; 3) with therapeutic interventions being associated with an eliminable amount of risk (“residual risk”), medicine is affected by ethical dilemmas regarding not only conflicting goods, but also the same kind of good: health; 4) high stakes are involved in most decisions at various levels (not only regarding health and well-being, but also implicating existential, psychological, and financial dimensions, as well as policy-making at the societal level); 5) vested interests held by (at least some of) the producers of medical knowledge strongly impact on the processes and norms regarding the production, interpretation and evaluation of evidence. (strategic behaviour).
Various institutional instruments have been developed in order to address these issues: evidential standards (such as the evidence hierarchies proposed within the evidence based medicine paradigm), decision-rules (e.g. the precautionary principle), and deontological norms: reproducibility requirements and bias detection tools. On their turn, scientists have advocated for a more comprehensive view of evidence, which may incorporate these issues explicitly, and keep track of them in the inferential process (Gelman, 2015, Marsman et al. 2017).
This talk presents a research program where these different issues are investigated in their joint interaction, by distinguishing different dimensions of first and second order evidence (Landes, Osimani, Poellinger; EJPS, 2017, Osimani, 2018). In particular, I will present how our theoretical framework can address problems of causal assessment of harm in pharmacology (Osimani 2013, Poellinger , 2018, forthcoming), and other meta-evidential questions such as the analysis of random error vs. bias, and the reproducibility crisis (Landes and Osimani, forthcoming).
1. Gelman, Andrew. Working through some issues. Significance 12.3 (2015): 33-35.
2. Landes J. Osimani B. Poellinger R. (2017) Epistemology of causal inference in pharmacology. Towards a framework for the assessment of harms. European Journal for Philosophy of Science.
3. Landes J. Osimani (forthcoming) Varieties of Error and Varieties of Evidence in Scientific Inference.
4. Marsman M., Schönbrodt F.D., Morey R.D., Yao Y., Gelman A., Wagenmakers EJ. (2017) A Bayesian bird’s eye view of ‘Replications of important results in social psychology’. R Soc Open Sci. 4(1): 160426.
5. Osimani B. (2018) Epistemic games and epistemic gains. A multilayer approach to causal inference in medicine. In: Osimani B., La Caze A. (eds.) (2018) Uncertainty in Pharmacology: Epistemology, Methods and Decisions”. Springer: Boston Series in Philosophy of Science.
6. Osimani B. (2013) Hunting side effects and explaining them: should we reverse evidence hierarchies upside down? Topoi (Special Issue: Evidence and Causality in the Sciences) October, Volume 33, Issue 2, pp 295–312.
7. Poellinger, R. (2018) Analogy-Based Inference Patterns in Pharmacological Research. In: La Caze, A. & Osimani, B (eds.) (2018): Uncertainty in Pharmacology: Epistemology, Methods, and Decisions. Boston Studies in Philosophy of Science. Springer.
8. Poellinger R. (forthcoming) On the Ramifications of Theory Choice in Causal Assessment: Indicators of Causation and Their Conceptual Relationships.
Abstract of Landes:
In this talk, I investigate the notions of varied evidence and reliability, their interplay and their contributions towards hypothesis confirmation within the framework of Landes et al. (2017). In particular, I shall show how one can explicate the notion of varied evidence, how too much positive evidence leads to a sharp drop in assessed reliability (too-good-to-be-true evidence) and whether the hypothesis
of interest or biases are more likely given the available evidence.