Tobias (Toby) Lewis 1918-2020
这篇讣告回顾了统计学家托比·刘易斯的一生，他贡献了统计方法、方向数据分析等领域的开创性工作，并以其教学和指导影响了许多学生。适合对统计史或统计教育感兴趣的读者。
Toby Lewis, who died in London on 20 November 2020 at the age of 102, will be remembered as a statistician who made wide-ranging contributions to statistical methodology, as a gifted teacher who inspired those he taught and supervised in their research, and as a warm and generous friend to professional and social colleagues alike. Many of his past students went on to become his lifelong friends. He continued to be immersed in statistics long after formal retirement and was still active in this respect in the final months of his life, reflecting a professional life of more than seven decades. Toby was born in Great Yarmouth on 19 October 1918. His family were Jewish, originally from Eastern Europe. His father was a herring exporter, his mother a gifted musician and artist (she invented a fish paste and was sent to Paris as a fashion artist) but the family was not well-off and Toby and his brother and two sisters had to succeed in terms of their own natural talents. His brother, Maurice, read Mathematics at Cambridge and went on to teach; his sisters, Naomi and Edina, excelled in the arts, as a distinguished writer, critic and poet, and as a gifted linguist, respectively. Toby attended Great Yarmouth Grammar School until the age of 12, when the family moved to London and Toby won a scholarship to the renowned St Paul’s School in Hammersmith. Toby read Mathematics on a Scholarship at Balliol College, Oxford (there was little prospect of reading statistics in those days) and he graduated with a First in 1938 at the age of 20. As was common for all young men at the time, he was called up for military service for 6 months, but he ended up serving 10 years in the army and attaining the rank of Major. He served partly in Borneo, on attachment to the Australian army, having worked ‘in the back room’ on logistical tasks, and from 1945 to 1948 in the AORG (Army Operational Research Group), It was at the Anti-Aircraft Artillery School in Manorbier in Wales in 1948, while giving some lectures on mathematics ‘to anyone who turned up’, that he met his future bride, ATS Officer Catherine McEachern. She was ‘the only one in the class to get all the right answers’ to his questions. They were happily married for nearly 60 years until Catherine died in June 2007. From 1948 to 1952, Toby worked at the National Physical Laboratory and then moved to the Statistical Advisory Unit of the Ministry of Supply for a further 5 years, clearly by this time having identified an interest and skill in statistics. His wish to enter academic work was realised in 1957 when he obtained a post as Lecturer in Mathematical Statistics in the Statistical Laboratory of the Mathematics Department at Manchester University, working for the renowned Maurice Bartlett. In the 5 years he was there that Department featured many mathematical luminaries, including Max Newman, Charles Lighthill, Bernard Neumann, Kurt Mahler, and eminent statisticians such as John Bather, Maurice Priestley and Peter Whittle, as well as many others who went on to lead their own statistics groups. Toby transferred with Maurice Bartlett to University College, London, in 1962. He stayed there for 6 years, spending interludes elsewhere, including time at Churchill College, Cambridge, and a year’s sabbatical in 1966, when he took up the offer of a visiting professorship at the University of Western Australia in Perth. This renewed his love of Australia, to which he returned many times throughout his career. As well as returning to the University of WA, he also later worked for the CSIRO in Canberra. Toby’s early research interests were catholic and included distributional theory, closed queues, conflict models and—a persisting interest—the analysis of directional data (circular and spherical), on which he published widely in research papers and books up to 2007. His collaboration on this subject with Nick Fisher lasted almost 20 years, and their book (with Brian Embleton) on spherical data analysis, first published in 1987, remains a standard and practical text. Toby became a Fellow of the Royal Statistical Society in 1950. In the 1960s, he served on the RSS Council, as well as on the Research Section and General Applications Section Committees and on the Editorial Panels of the RSS Journals, Series B and C. He was also a member of the Council of the Institute of Statisticians from 1980 to 1985. With the amalgamation of the RSS and the IoS, Toby became a Chartered Statistician. He was also elected to Fellowship of the International Statistical Institute. Throughout his career, Toby retained an abiding interest in the practical side of statistics and in statistical education. In the 1960s and 1970s, he was part of an influential group concerned with teaching statistics in schools. He was himself renowned as a gifted and caring teacher of the subject, maintaining close personal and professional links with many colleagues he had initially taught or supervised in their research. In 1968, Toby was appointed to the Chair of Statistics in the newly created department at Hull University (living next door to Philip Larkin so ‘not totally bereft of culture’). It was during the 11 years at Hull that Toby jointly published with Vic Barnett the first edition of their renowned book on Outliers in Statistical Data which, now in its very much extended third edition, is widely regarded as the authoritative treatment on the subject. In 1979, Toby was awarded a DSc by Manchester University. In the same year, he also took up the first Chair in Statistics at the Open University, and this provided new challenges and interests from the teaching standpoint in terms of TV preparation and presentation. He lived locally, in Husborne Crawley, where Catherine was headmistress of the village school. While at the Open University Toby fought a heartfelt and protracted, but ultimately unsuccessful, campaign to help a Mathematics student who was accused of cheating in his Final exams and was refused an Honours degree. Toby was determined to help the fight against what he believed to be the injustice of the case. The arguments ranged through a University tribunal, an appeal to the Visitor and a case in the High Court, which was eventually dismissed since the large sum required as security against future costs could not be met. Toby retired from the Open University in 1986. He and Catherine moved to Diss in Norfolk to go ‘back to his roots’, and he was appointed to an Honorary Professorship at the University of East Anglia, a post which was extended until the summer of 2010. He continued to be active professionally, receiving an Honorary DSc from UEA in 2000 in recognition of his work as ‘an inspirational teacher and a leading light in the area of multi-level modelling’. Research interests and publications over these periods encompassed new fields for Toby, principally multi-level statistical models, longitudinal studies and educational testing and assessment, often working with Harvey Goldstein and Ian Langford. Toby continued to take a close interest in professional matters until into his 90s, among other things travelling back and forth from his home in Diss to the Offices of the RSS in London, where he was involved with activities of the Society’s General Application Section. With the help of Debra Hurcomb at the RSS and others, Toby successfully campaigned to get statistical topics into the mathematics programme of the annual festival of the British Association for the Advancement of Science (BAAS; now the British Science Association). One of his last professional writings was a proposal for a session on statistics in sport for BAAS. That exemplifies several lifelong aspects of Toby’s statistical work; a recognition of the importance of building on their existing experience to motivate learners’ interest in statistics, an ability to contribute in an application area that was new to Toby at the time, and a talent for making enduring contributions. (The British Science Association continues to present statistical topics in its events, and at the RSS, Statistics in Sport has since become a Section.) He is survived by his daughters Gina and Rae, and grandson Alexander, and will be remembered with great warmth by the very large number of professional colleagues whom he taught, helped and worked with over an impressive range of statistical topics.