Obituary for Thomas Augustin
Posted on January 23, 2026 by Christoph Jansen, Julian Rodemann and Georg Schollmeyer[ go back to blog ]
It is with great sadness that we have learned of the death of Professor Thomas Augustin, from Ludwig-Maximilians-Universität (LMU) in Munich.
Thomas Augustin was one of the founders of our society, being already present at the very first ISIPTA in Gent in 1999 and contributing decisively to the creation of the society in 2002. From the very beginning, Thomas intensively supported the development of the society and the spread of the ideas of imprecise probabilities. He contributed to and participated in every edition of ISIPTA until the last ISIPTA 2025 in Bielefeld.
After completing his diploma studies in statistics at LMU in 1992, Thomas decided to pursue a PhD, initially focused on the historical roots of imprecise probability. Then, during the preparation for the (unfortunately never published) third volume of Kurt Weichselberger’s book Elementare Grundbegriffe einer allgemeineren Wahrscheinlichkeitsrechnung, which was devoted to statistics with interval probability and particularly the Huber–Strassen theory on robust testing of hypotheses described by neighbourhood models, he changed his topic to support this project and developed a Neyman–Pearson theory under general interval probability, where the hypotheses are described by F-probability instead of two-monotone capacities known from the famous Huber–Strassen theorem for capacities. In 1997 he finished his dissertation (Optimale Tests bei Intervallwahrscheinlichkeit, optimal tests under interval probabilities) under the supervision of Kurt Weichselberger. This early achievement shows Thomas’ commitment to make the core idea of imprecise probabilities (here in the guise of interval probabilities) fruitfully applicable to statistics.
Following a habilitation in 2002 on survival analysis under measurement error, supervised by Hans Schneeweiß, and a brief stay at the University of Bielefeld he became professor at the department of statistics at LMU in 2003. There, he was the head of the working group Methodological Foundations of Statistics and Their Applications within the Department of Statistics at LMU Munich.
Moreover, he served as the Dean of Studies (Teaching Coordinator) at the Department of Statistics since 2006, being in charge of all teaching activities in the undergraduate and graduate courses. Thomas was deeply devoted to teaching and left a lasting impact on generations of students. He was known for always having a sympathetic ear to the wishes and concerns of students; and he was never too proud to care for outsiders or less privileged people.
As one of the founders of SIPTA, Thomas was a member of the Executive Committee between 2002 and 2009, 2017 and 2019 and since the last elections held in the summer of 2025. In addition, he was in the Program Committee Board of the ISIPTA conferences in 2009 in Durham and 2015 in Pescara. In all these endeavours, his presence was always calming and reassuring, and the final result benefited greatly from his input.
All this administrative work did not detract him from making important contributions to imprecise probabilities and related fields: he published over 100 contributions on very diverse topics such as Interval and Imprecise Probabilities, Statistics with Imprecise Probabilities, Measurement Error Modelling, Decision Theory with Imprecise Probabilities, Partial Identification and Analysis of Deficient Data, Official Statistics, Machine Learning with Imprecise Probabilities, (IP in) Applied Statistics (Biostatistics, Social Sciences, …), …
He was also one of the editors of the book Introduction to imprecise probabilities, published in 2014 that gave a general overview of the theory at the time, and of the Festschrift honouring Prof. Teddy Seidenfeld in 2022. His work contributed decisively to establishing imprecise probabilities as a viable and sensible alternative to classical models in situations of ambiguous, contradictory or imprecise information. The impact of his work is attested by almost 9000 citations and by his regular inclusion in the ranking of the 2% most influential researchers by the Stanford University.
At LMU, not only at the scientific level, but also with respect to administrative duties, Thomas was always very involved and interested in the general positioning of the institute in terms of strengthening statistics and in particular also imprecise probabilities as a means to an indispensable and helpful methodological ingredient of a reliable and well-founded approach to statistical data analysis.
Thomas always worked together with his PhD students very intensively. He did not only introduce several generations of PhD students to the world of science, its demand of intellectual integrity and the role of criticism of ideas as a main pillar of science. He also lived up to this aspiration. Generally, working in Thomas’ group was not so much like working within a hierarchy of PhD’s, research assistants and a boss, it was more like caring for each other within some kind of a ‘family’. This became especially evident when Thomas regularly invited PhDs from other working groups to participate in his research seminar, since their supervisors left the university and they were left with no clear ‘academic home’.
The intellectual vision of Thomas was always that of statistics as an interdisciplinary science of responsible data analysis with all its implicit and explicit relations to philosophy, mathematics, history of science, computer science, as well as the political and social sciences. Therefore, he had many cooperations and joint events, including courses embedded in the master curriculum, for example with partners from official statistics or from philosophy. Joint research seminars together with the colleagues from the Munich Center for Mathematical Philosophy (MCMP) or a Workshop on the History of Statistics show the broad research interests of Thomas. Regarding SIPTA related activities, it is also worth mentioning the Workshops on Principles and Methods of Statistical Inference with Interval Probability (WPMSIIP) initiated by Thomas together with Frank Coolen, where participants partly from the core SIPTA community but also from other related communities came together.
But above all these curriculum considerations, there was one aspect of Thomas that always stood up above the others: his understanding that mathematical rigor does not need to exclude gentleness and generosity in the interactions. As anyone who was fortunate to meet him can attest, Thomas was always willing to help, particularly young students who are often in need of orientation in their research path. This is one of the reasons why his loss is felt even more acutely: it is not only the research that will not be done, but also the atmosphere that he helped create.
The greatness of a man can be measured by his positive impact on those around him; and Thomas was indeed a great man.
