Statistical Reasoning and Scientific Error
tatistical reasoning pervades experimental research, but how to apply it is a longstanding issue of debate in philosophy and science. Recent studies that reveal a high prevalence of error and lack of reproducibilityin published research highlight the urgency of developing sound foundations of statistical reasoning, and finding techniques for detecting and correcting scientific error. So much the more as scientific error undermines the epistemic authority of science, and the degree to which policy-makers trust scientific experts.
The conference brings together an interdisciplinary group of researchers interested in issues of statistical reasoning and scientific error. It focuses on the foundations of statistical inference, how statistical reasoning is applied in the sciences, how statistical inference can correct scientific error and which proposals for reforming scientific method (including restructuring the peer review and publication system) can increase the reliability of published research.
The conference is a joint event of the Munich-Sydney-Turin conference series in philosophy of science and the workshop series “Perspectives on Scientific Error”. Previous editions of these workshops can be consulted here (PSE1, Tilburg 2017) and here (PSE2, Groningen 2018).
The topics of the conference include, but are not limited to:
- the explanatory and inferential function of statistical models in the sciences
- the foundations of statistical inference
- the relationship between statistical reasoning and scientific error
- the place of statistics in scientific method
- the replicability of research findings
- techniques for detecting and correcting scientific error
- reform proposals for scientific method (e.g., Bayesian vs. frequentist statistics, publication practices, incentive structures)
- reliable methods of aggregating and interpreting evidence
- values in science, and their impact on scientific error