Workshop participants will also have the opportunity to attend tutorial sessions at the beginning of the workshop:
An Introduction to the Epistemic and Decision-Theoretic Foundations of Statistical Inference.
Instructor: Jan Sprenger (University of Turin)
Scientists usually have a high degree of methodological savvyness, but they are less aware of the philosophical foundations of statistical inference. This is notable since choosing an inference method, such as a significance test or a Bayesian analysis, entails various (implicit) commitments: to epistemic principles, to conceptualizations of scientific hypotheses, and to dispositions to act in a certain way. These commitments may be hidden, but they are of the utmost importance to the interpretation of an experimental result. The tutorial will point out the nature of these commitments and explain how they are entrenched in the most widespread methods of statistical inference. In particular, I will show how the decision-theoretic elements in statistical research relate to the ideal of value-freedom: that is, the (contested) thesis that scientific research should be as free of personal values as possible.
Jan is Professor in the Department of Philosophy at the University of Turin (2017–). After completing a mathematics degree, he gained a Ph.D. in philosophy in 2008 at the University of Bonn, Germany. Then, he was Assistant Professor in the philosophy department at Tilburg University (2008–2014). From 2014 to 2017, he was Professor of Philosophy of Science at Tilburg University and Scientific Director of the Tilburg Center for Logic, General Ethics, and Philosophy of Science (TiLPS). Jan works mostly in philosophy of science, in particular the foundations of statistical inference, formal epistemology and decision theory. He publishes in journals such as Philosophical Review, Mind, Philosophy of Science, and British Journal for the Philosophy of Science and he is currently finishing a research monograph “Bayesian Philosophy of Science” (with Stephan Hartmann).
An Introduction to (Mis)applied Statistics.
Instructor: Daniël Lakens (Eindhoven University of Technology)
Researchers are often criticized for misusing statistics. Common approaches to statistics are criticized for not giving researchers ‘what they really want to know’. What do scientists want to know? What which questions can statistics answer? Are statistics related to knowledge generation in a meaningful manner? In this brief introduction, we will discuss three main approaches to statistical inferences, and think about which, (if any) tell us meaningful things to know.
Daniel is an assistant professor at the School of Innovation Sciences at Eindhoven University of Technology. In addition to his empirical research in cognitive psychology on conceptual meaning, he is interested in applied statistics, reward structures in science, improving research practices, and reducing publication bias. He created a highly rated and popular MOOC on Coursera, “Improving Your Statistical Inferences”, writes on methods, statistics, and open science at his blog the 20% statistician, and received an award for best teacher at Eindhoven University of Technology in 2014. He is currently working on a NWO VIDI funded project ‘Improving the reliability and efficiency of psychological science’ and is interested in helping researchers make optimal choices when designing studies, based on their resources, values, and objectives.