Embedded Files
EPIX
Epistemic Foundations and Limitations of Statistics and Science - Explaining the Replication Crisis?
An OSIS@OSC Workshop
Munich, Germany
The aim of the workshop is to bring together statisticians, philosophers of science, and other interested researchers to discuss aspects of Open Science and the replication crisis from an epistemic perspective on the foundations and limitations of statistics.
The replication crisis is usually explained by misaligned incentive structures in the scientific system ("publish or perish"). Another important, but underrated, aspect is that researchers -- even statisticians -- tend to disregard important epistemic foundations of statistics. This lack of principle (or knowledge) effectively contributes to questionable research practices, statistical rituals, delusions and fallacies about what statistics and science can achieve. In the age of big data and foundation models, the importance of these topics is greater than ever before.
With keynotes by select speakers and ample time for discussion, we will address the question why the inappropriate use of statistical methods (most prominently, statistical testing) is still so prevalent, despite warnings that have been voiced for decades.
At a glance
At a glance
Who Statisticians, philosophers of science, and interested researchers (registered participants only)
Where LMU Munich, Schellingstr. 3, Room S 007, in person
What Series of talks (each circa 20 minutes + 15 minutes discussion). Find the program here.
When Friday, 22 November 2024. Check-in is 8:45 - 9:15. The workshop will end around 17:30.
!!! A synopsis of the workshop can be found in this post on the MCML Blog !!!
Thank you to all speakers and attendees who braved the sudden onset of winter for a productive and fun workshop!
Invited talks
Invited talks
Sabina Leonelli
TU MunichSabina Leonelli
What Reproducibility Can’t Solve (Keynote)
Sabina Leonelli was recently appointed Professor for Philosophy and History of Science and Technology at TUM. She works on the philosophy of data-intensive science; in particular, the impact of big and open data on research and society in general.
Uwe Saint-Mont
HS NordhausenUwe Saint-Mont
How Feynman Predicted the Replication Crisis (Keynote)
Uwe Saint-Mont holds a chair at the Department of Economics and Social Sciences. He works on statistical theory, philosophy of statistics, and philosophy of science.
Jürgen Landes
LMU MunichJürgen Landes
Data Aggregation of Big Data Is Not Enough
Jürgen Landes is a PostDoc at the Munich Center for Mathematical Philosophy and member of the LMU Open Science Center. His general interests lie in all things uncertainty, including uncertain reasoning, inductive logic, philosophy of probability and statistics, uncertain inference and inference from imperfect data, as well as Bayesian inference in general and in medicine in particular.
Moritz Herrmann
LMU Munich, MCMLMoritz Herrmann
When Measures Become Targets
Moritz Herrmann is a PostDoc at the Institute for Medical Information Processing, Biometry, and Epidemiology (IBE) and part of the Munich Center for Machine Learning, where he serves as the Reproducibility and Open Science Transfer Coordinator. He is also a member of OSIS and the LMU Open Science Center. His research interests include empirical machine learning, epistemology of statistics and machine learning, meta-science, and data literacy.
Sabine Hoffmann
LMU MunichSabine Hoffmann
How Foundational Assumptions about Probability, Uncertainty, and Subjectivity Jeopardize the Replicability of Research Findings
Sabine Hoffmann is a trained statistician, psychologist, and epidemiologist, and head of the statistical consulting unit at LMU Munich. She is member of the scientific board of the LMU Open Science Center and chair of the Open Science panel of the STRATOS initiative. Her main research focus is on the modeling of complex uncertainty structures using a Bayesian hierarchical approach and on methodological aspects in meta research.
Walter Radermacher
Advisory Board on Ethics of the International Statistical Institute and LMU MunichWalter Radermacher
Epistemology and Sociology of Quantification based on Convention Theory
Walter J. Radermacher is a Professor of Statistics at LMU. Additionally, he serves as the Advisory Board Chair of the International Statistical Institute (ISI). Furthermore, he has experience in public administration at national and European levels, having served as President of the German Statistical Office (destatis), Director General of Eurostat and Chief Statistician of the European Union, and Chair of the destatis Commission Future Statistics. He works on statistics, sustainable development, governance, ethics, and data analysis.
Michael Schomaker
LMU MunichMichael Schomaker
Replicability When Considering Unconditional Interpretations and Gradations of Evidence
Michael Schomaker is a Professor for Biostatistics working under the German Research Foundation's Heisenberg Program. He works on the appropriate use of modern causal inference methods.
Rudolf Seising
Deutsches MuseumRudolf Seising
An Interwoven History of AI and Statistics
Rudolf Seising currently chairs the BMBF research project "Eine Geschichte der KI in der BRD". He works on the history of science and artificial intelligence.
Directions
Directions
Arrival by public transport
Take subway trains U3 or U6 to stop "Universität" and take the exit "Schellingstraße". You will be at the intersection of Ludwigstraße and Schellingstraße.
Bus lines 153 and 154 as well as 58/68 also stop at the intersection Ludwigstraße/Schellingstraße.
For planning your trip, please use https://www.mvg.de/verbindungen.html and https://www.mvv-muenchen.de/en/index.html.
Arrival by car
Finding a (paid) parking space for your car on site is theoretically possible, but practically very unlikely. Park & Ride options are described here and here.
Room S 007 in Schellingstraße 3
The room is located on the ground floor. The entrance to the room is next to a big glass façade.
The room's location in the "LMU Roomfinder" can be found here.
The map below shows the building and room S 007 (yellow X), the main entrance (M) and side entrance (S) of the building, the exit "Schellingstraße" of subway station "Universität" (blue U) and local bus stops (red boxes).
Local food options
Local food options
Those who did not sign up for lunch will find many cafés, restaurants, and to-go places nearby. Veg options can be found here.
For employees and students of LMU, TUM, and HM in particular: the StuBistroMensa Schellingstraße is exactly one floor above the workshop room. (Open 10 am to 2 pm. Don't forget to bring your ID card. Payment options are described here.)
Registration
Registration
Registration is closed.
To all who signed up: Looking forward to seeing you at EPIX!
Acknowledgments
Acknowledgments
The organizers are grateful to the Munich Center for Machine Learning (MCML), the GraduateCenterLMU, the LMU Open Science Center (OSC) as well as LMU Munich and TU Munich for their support.
Contact
Contact
If you have any questions, please do not hesitate to contact epix.workshop@stat.uni-muenchen.de.
Moritz Herrmann
Moritz Herrmann
MCML Reproducibility and Open Science Transfer Coordinator,
LMU
Patrick Schenk
Patrick Schenk
PhD Candidate Social Data Science and AI Lab,
LMU
Lisa Wimmer
Lisa Wimmer
PhD Candidate Statistical Learning & Data Science,
LMU
David Strieder
David Strieder
PhD Candidate Mathematical Statistics,
TUM
Page updated
Google Sites
Report abuse