To view the entire ISI Newsletter, please download the
PDF
.
Please first choose your favorite format:
Default Style |
Large fonts |
No Graphics
Editors: Ms. Ada van Krimpen and Ms. Shabani Mehta,
Graphic Designer: Mr. Hans Lucas
100th Issue Front Cover
Top
President's Message
Dear Members of the ISI and the Associations,
Let me use this opportunity to wish all of you a very successful 2010.
Allow me to add a number of points to the message that has been sent by the Permanent Office. In that e-mail, I asked for your help and suggestions for the first World Statistics Day 20-10-2010, for the 125th anniversary of the ISI and for suggestions for the Short Courses in Dublin next year. (Moreover, the ISI members have been asked to give their agreement on a procedure to gain the immediate support of 2 Member Auditors.)
1. Progress with the next World Statistics Congress in Dublin, August 21-26, 2011, is going well. Please consult the website www.isi2011.ie for up-to-date information. We of course expect a very large group of participants, in particular from developing countries. The ISI will do its utmost to reduce as much as possible the financial burden for the developing country participants.
2. The Executive Committee (EC) of the ISI met in The Hague, December 7-8, 2009. The agenda of the meeting was very extensive. I would like to inform you about a number of items that were discussed.
(a) After preparation by the EC, the Council was asked to make a number of decisions in connection with the ISI Committees. As a result, the following has been decided.
(i) Since the ISI Glossary Committee, under the chairmanship of René Padieu, has finished its assignment, the Committee has been abolished. The safeguarding of the content and operating facilities of the Glossary have been taken up by Council member Jens Ledet Jensen, who has been added as an ex-officio Council representative to the ISI Publications Committee.
(ii) The Council has agreed to the establishment of an ISI Committee on Astrostatistics. The Committee aims to foster collaborative research efforts between astrophysicists and statisticians with an interest in the analysis of astronomical data. The Chair of the Committee is Joseph M. Hilbe who started the initiative. We invite interested members to contact Joe at hilbe@asu.edu. For more information on the Committee, read the Chair’s contribution further in the issue of the ISI Newsletter.
(iii) In order to help Ali Hadi, the new Editor-in-Chief of the International Statistical Review, in his attempts to improve the quality of our flagship journal of our Institute, the Council has agreed to establish an Advisory Committee for International Statistical Review. A summary of the tasks for the Committee will be announced soon.
(b) Just before the ISI Session in Durban, the members of the Council had been asked to help the ISI, and in particular the EC, by taking up the responsibility for one Portfolio. I am pleased to say that the response was overwhelmingly positive. At its December 2009 meeting, the EC had already received four extensive reports that will be followed up on over the next year. All of these portfolios need thorough discussion before they can be implemented into decisions by the Council. We hope to report on a number of them after the next Council meeting that will be held in Iceland, July 22-23, 2010.
(c) One of the main tasks of the current ISI Governance is to prepare a serious revision of the ISI Statutes and By-Laws. I am sure that everybody who has ever dealt with such an exercise knows how difficult this assignment is. Both the EC and the Council are committed to fulfil this task by the end of their term of office, namely the World Statistics Congress in Dublin. It goes without saying that the membership of the ISI and of the Associations will get full insight into the proposed texts well ahead of that date.
3. The Executive Committee envisages to regularly spend part of every meeting discussing important ISI issues in full detail. For example, in its next gathering in Singapore, May 3-5, the EC will brainstorm about the Quality of the World Statistics Congresses. Also, the Council is expected to go through similar exercises but for problems with an even wider scope. We invite the membership to offer us their suggestions for such brainstorming topics.
I really hope that all of you are convinced of our intentions to greatly intensify the interaction between the EC, the Council and the membership of the ISI and the Associations. We all have a responsibility towards an institute that needs to rejuvenate, even at (or perhaps because of) its age of 125.
Jef Teugels,
ISI President
Director's Message
My very best wishes to all members of the ISI and the Sections for a Happy New Year.
The year 2010 will be a special year in many respects. To begin with, it will be marked by the 100th issue of the ISI Newsletter. Secondly, the ISI - founded in 1885 - will celebrate its 125th anniversary. And lastly, the ISI will participate in the first World Statistics Day scheduled for 20th October 2010.
The 100th issue of the ISI Newsletter was a good reason to explore our archives and to have a closer look at all preceding 99 issues. On the cover of this Newsletter, you will see a compilation of the covers of the issues over the years. The Newsletter in its present form will cease to exist at some point this year. News items will appear with a more regular interval on our new website, which we hope will go live by the middle of this year. One of the advantages of this is that the ISI members will be informed about the developments occurring within our organisation more frequently. On the other hand, we are also aware of the fact that some of you would like to continue receiving a paper version of the Newsletter. Therefore, the Publications Committee wishes to investigate the possibilities for a proper follow-up of the present Newsletter.
Looking into the past issues of the Newsletters shows that since 1977, when the first issue was published, a lot has changed within our organization. However, many items are still on our agenda. The first Newsletter was intended for the members of the ISI, and its three Sections at the time: Bernoulli Society, IASS and the then existing International Association for Regional and Urban Statistics (IARUS, now SCORUS). Among the topics discussed were the 41st ISI Session in Delhi, India, the World Fertility Survey and the 100th Anniversary of the ISI. Over the years, the new Sections were covered and more general topics were discussed. Some of the subjects discussed at that time still concern us today. For instance, I have learned about the discussions in the nineties to increase the ISI membership, the possibility to have all Sections’ members also as associate members of the ISI, and the need for an increase in the ISI’s human as well as financial resources. These are still relevant questions today! One of the ambitions of the present Executive Committee is to make progress in all these questions - with the support of our members.
One of the new features is dedicated to the ISI’s LinkedIn Group. LinkedIn is a professional networking website that you can register for via http://www.linkedin.com/. Through this group, information can be exchanged and discussions can be initiated among our members. I invite all of you who are active on the Internet to join the group and help us explore the ways of using this new method of communication for a more active exchange of ideas. A discussion has started about what ISI’s contribution could be to the World Statistics Day on 20th October 2010 among the few members who have already joined the group at the time of writing this message. A delegation representing the Executive Committee of the ISI will participate in the UN Statistical Commission in February where this topic will be discussed and it would be great if the ISI could contribute to this event. ISI President Jef Teugels has already invited you, our members, to present your ideas to us.
The 125th anniversary of the ISI is a remarkable event as well. The ISI is one of the oldest scientific organisations still alive in the world today; undertaking a lot of activities and new initiatives are definitely worth congratulating. During the course of this year, you will be informed about the activities we are planning to organize. Coincidence or not, but the fact that I am invited to give a presentation about the ISI in St. Petersburg this month was a good reason to give the history of our organisation and its predecessors, the 19th century statistical congresses, more attention. There are many lessons to be learned from this rich history and experiences, and I hope that we can find new ways for making them available to you, our members.
In 2010, news ways of communication will be explored and implemented, while keeping the good things from the past and learning lessons from our history.
Ada van Krimpen
Director ISI
58th ISI World Statistics Congress, Dublin, Ireland
21-26 August 2011
Call for Special Topic Session Proposals
Planning for the Scientific Programme of the ISI World Statistics Congress in Dublin is well underway and the Programme Coordinating Committee (PCC) is currently finalising the organisers for the Invited Sessions (please read article below).
The next stage of planning is to choose topics and organisers for the Special Topic Sessions.The PCC is now calling for proposals for organising Special Topic Sessions from potential organisers, members of the ISI or of one of the Sections. Proposals should be submitted by 30th April 2010 to a relevant member of the PCC, see http://www.isi2011.ie/pages/scientific-programme/programme-coordinating-committee.php, or to the Chair of the Local Organising Committee, John Haslett, John.Haslett@tcd.ie.
An organiser of a Special Topic Session proposes the topic and, if that topic is selected, he or she finds speakers who are interested in providing a comprehensive and coordinated presentation of it. The Session lasts for 135 minutes and there are generally 3 to 5 speakers. More information is available at http://www.isi2011.ie/pages/scientific-programme.php.
John Haslett
Local Organising Committee ChairProgramme Coordinating Committee - Progress Report
The PCC has now a set of 78 Invited Sessions that have been agreed and organisers are confirmed for 56 of them. We anticipate that there may be some minor changes in organisers or in titles of sessions. Some sessions on 'late breaking topics', together with sessions on the Water Theme, are to be added. The provisional list is given below. The next tasks are (a) for organisers to identify speakers and their topics and (b) for the PCC, through Sections and Committees, to develop a similar sized list of Special Topic Sessions as mentioned in the above article.
Murray Cameron &
Silvia Regina Costa Lopes
Programme Coordinating Committee Chairs
News of Members
Director General of Statistics Finland receives OECD recognition
“The Secretary General of the OECD has awarded a special diploma to Heli Jeskanen-Sundström, Director General of Statistics Finland, in recognition of her contribution to the development and strengthening of co-operation between Finland and the OECD.
“Heli Jeskanen-Sundström has participated widely in OECD's statistical co-operation for over two decades, and been Bureau Member of the OECD Committee on Statistics since 2006. To celebrate Finland's 40 years of membership in the OECD, more than 30 officials of Finland's state administration were awarded the diploma. The presentation ceremony took place at the Finnish Ministry of Foreign Affairs on 12 January 2010.”
Mrs. Jeskanen-Sundström is a member of the IAOS.Source of announcement:
http://www.stat.fi/ajk/poimintoja/2010-01-14_oecdpalkinto_en.htmlPhoto: Statistics Finland
ISI Membership Elections 2009
We would like to congratulate the 54 new ISI members, who were elected in the second round of the 2009 ISI membership elections. For those who wish to contact any of these individuals, please note that the ISI website contains a component including the names and addresses of all ISI members (see http://isi.cbs.nl/isimembers/isimembers.htm), and these new members will be added to this list in the coming weeks.
Elected Membership
(
We regret to announce the deaths of the following members
Born Elected Deceased Professor Wilfrid J. Dixon 1915 1975 20 September 2008 Professor Barry H. Margolin 1943 1977 28 January 2009 Professor Dr. Elsa Servy 1928 1985 20 April 2009 Dr. John Clement Koop 1919 1978 9 June 2009 Professor Erich Leo Lehmann* 1917 1954 12 September 2009 Mr. Tom Griffin 1937 1999 31 January 2010 Dr. Hirotugu Akaike 1927 1972 Date unknown Mr. Peter Reginald Fisk 1929 1971 Date unknown Dr. Stephen William Lagakos 1946 1980 Date unknown Professor Dr. Masashi Okamoto 1923 1974 Date unknown *Correction to the previous ISI Newsletter: The date of death was noted as 13 September 2009.
If you know the date of death of any of the above-mentioned members whose dates are unknown at the time of publishing this ISI Newsletter, we would appreciate your passing the information to Mrs. Margaret de Ruiter-Molloy at @cbs.nl.
In Memoriam
Hugh Daniel Brunk*: Aug. 22, 1919 - July 19, 2009
Hugh Daniel Brunk, 89, died Sunday, July 19, 2009, at his home in Corvallis.
Born August 22, 1919, in Manteca, California, he was the eldest son of Hugh and Velma Benson Brunk. Dan received a bachelor’s degree from the University of California, Berkeley, and a doctorate from the Rice Institute.
In 1942, he married Jean Young. After serving in the U.S. Navy during World War II, he began his career as a professor of mathematics and statistics. He taught at Rice; the University of Missouri; the University of California, Riverside; and Oregon State University. He spent sabbaticals as a visiting professor in Denmark, England and Wales.
Professor Brunk was a man of diverse interests and accomplishments. He turned his hand to many things, and excelled in all of them. Dan was the author of numerous mathematical books and papers. He was a Fellow of the Royal Statistical Society, and was an active elected member since 1973 in the International Statistical Institute, the Mathematical Association of America and the American Mathematical Society.
He loved music and sang in choirs, choral groups and light opera. He also enjoyed nature and liked to hike.
He married Jean Sherk in 2004 in Corvallis.
Dan was preceded in death by his parents; his sister, Ruth; and his wife, Jean Young Brunk, who died in 1997.
Survivors include his wife, Jean; his daughters, Bridget Glidden and her husband, Derald, of California, Gretchen Armacost and her husband, David, of Massachusetts, and Heidi O’Riley of Colorado; three grandchildren; five great-grandchildren; one great-great-grandchild; his sister, Betty Kemph of Oregon; and his brother, Joe Brunk, of California.
A memorial service took place at 1:30 p.m. Thursday, July 23, at McHenry Funeral Home, 206 N.W. Fifth St.
Memorial donations in his name can be made to one’s favorite charity. Online condolences can be made via www.mchenryfuneralhome.com.
*This obituary was originally published in the Corvallis Gazette Times and reprinted with permission.
Daniel Schwartz (1917-2009)
Daniel Schwartz passed away on 6th September 2009. He was an ISI elected member for 41 years, since 1962, and he resigned for health reasons in 2003. He is legitimately considered as the founder of the French School of Medical Statistics and of the main stream of French epidemiology.
His parents came from Alsace, where they left after its annexation by Germany following the 1870 war. His father was a well-known hospital surgeon. He, as well as his two brothers, did not follow their father’s footsteps in the medical field since the three of them were trained in high-level mathematics: the eldest, Laurent, became a universally known mathematician (rewarded with the Fields Medal), Daniel and Bertrand, the youngest, entered Ecole Polytechnique, one of the best higher education establishments in France. Mobilized when the Second World War began, Daniel, after the armistice was signed in 1940 between the Nazi regime and the French Vichy Government, had to take refuge in the south-west of France where he entered the French Resistance during a very difficult period for his family and an awful one for the family of his wife Yvonne.
After the end of the War, Daniel Schwartz started his professional career in Bergerac, a research centre belonging to the French monopoly in charge of tobacco and matches (SEITA) where he had been nominated as an engineer. Encouraged by Professor Georges Darmois (a former ISI President and the first French mathematician who understood in those days that statistics was not only one of the branches of “pure” mathematics), Daniel Schwartz and another young SEITA engineer, André Vessereau (one of the first French members of the ISI), developed statistically-based researches on the illnesses of the tobacco plants, in particular those induced by specific viruses. Some people think they dreamt to transform the Bergerac centre into a little French “Rothamsted”, the famous British experimental station where Sir Ronald Fisher played such an important role.
However, what should happen happened. The memory of his father’s profession, the influence of his uncle, Robert Debré (a famous paediatrician who, later, created the French integrated hospital-university system), and his personal taste for the medical field pushed him towards a new and major orientation: Using his statistical knowledge to study the diseases that might be related to smoking. He always underlined that he had no strict opposition from the Director-General of SEITA to enter this field and to publish the results. Schwartz launched the first large-scale French survey on aero-digestive, lung and bladder cancers, taking into account detailed smoking habits, including inhalation of cigarette smoke. He could demonstrate that inhalation played a major additive role, but only on “deep” cancers’ incidence and not on mouth cancers’one – which indeed was a major contribution to the results obtained in the UK, mainly by Sir Richard Doll, and the US in favour of a causal relationship between smoking and cancer.
At the end of the fifties, Daniel Schwartz met Professor Pierre Denoix, the Director of the main French anti-cancer institute (Institut Gustave-Roussy), and Louis Bugnard, the Director of the National Institute of Hygiene (which became, a few years later, the National Institute of Health and Medical Research - INSERM). Denoix, Bugnard and Schwartz decided to create a research laboratory in medical statistics (called unité de recherches statistiques), located in the Gustave-Roussy Institute. This creation was the starting point of a genuine explosion of epidemiological researches in a great variety of fields: Cancers, cardiovascular diseases, diabetes, perinatology, genetics, environmental exposures, human reproduction, sexuality, sexually transmitted diseases, infectious diseases, mathematical models, etc. In the following years, at least ten research units started from this training and research centre, which contributed to the worldwide reputation of French epidemiological research.
One main original feature of the French approach of these issues must be underlined. In the UK and US, “statisticians” usually have a basic training in mathematics while “epidemiologists” have a medical one. The French situation is quite different, and this is mainly due to the personal vision of Daniel Schwartz. Notwithstanding, his personal basic high-level training in mathematics, he thought that mathematicians working with him should focus on using a statistical language sufficiently simple to be understood by medical doctors, and, vice-versa, that his medically trained colleagues should be able to understand the rationale of statistical methodology, which clearly meant that the boundaries between medical doctors and mathematicians could progressively be softened with the help of a specific teaching.
This is a way in which to recall the exceptional teaching influence of Daniel Schwartz and his team. Schwartz started his teaching very early (in the early fifties), in a building that was paradoxically “the temple of pure mathematics” in Paris, the Institut Henri Poincaré. He created in the sixties the world-famous CESAM[1], which still exists more than forty years later and has trained tens of thousands of medical students or doctors, obviously in Paris but also in the rest of France by correspondence, and frequently in Europe and elsewhere in the world. Two main basic principles structured the highly diversified set of CESAM trainings: To help students understand how it is possible to establish the significance of an observed relationship and how to progress towards its (causal) interpretation - all this with the minimum of mathematical formalism. These guidelines were perfectly compatible with the strong support Schwartz gave to those of his colleagues working in theoretical statistics and its applications to various fields of medical and health research.
Schwartz also dealt with the methodology of therapeutic trials. In the fifties, nobody in France dared to “decide at random” what treatment should be given to a patient if two possible options existed. Here also, the British researchers were one step further than the French and it is from their experience (B. Hill and P. Armitage) that Daniel Schwartz introduced the first randomized trial into our country - allowing, precisely, to interpret an observed difference in causal terms. Many years later, this way of comparing the efficiency of various treatments was finally recognized as the only one to be ethically acceptable through a statement of the French Comité consultatif national d’éthique pour les sciences de la vie et de la santé[2] (1984) It was finally codified by law in 1988.
However, Schwartz went much further in a very original way. He invented the dual concept of trials based either on an explanatory attitude or on a pragmatic one. The classical statistical analysis of a trial was considered as an application of an ordinary significance test. The chosen threshold 5% is usually “small” in order to prevent an incorrect conclusion that the null hypothesis should be rejected. For instance, it is a procedure well adapted to the comparison of a new molecule to a placebo. However, said Schwartz, it is possible to adopt quite a different attitude if the goal is different; for instance, if one wishes to compare two complex therapeutical strategies for which it is quite unlikely that the null hypothesis is true! Then the goal is to make the best decision and to minimize the risk in order to choose the best treatment. The differences between both types of trials are important. Some previous statisticians already tried to define clinical trials as a decision-making problem. However, Schwartz showed that the difference was not only in the number of subjects to be included and the rules of conclusion of the trial, but in all the aspects of the protocol, the precise definition of the disease, the nature of the patients to be included, the treatments to compare, the way to deal with the patients “lost” during their follow-up, etc. This was described in a seminal paper[3] and then extended in a book[4].
What strategy should be adopted in a specific situation? The choice is not necessarily obvious: Most of the trials wish both, to answer an explanatory need and to help make correct decisions… Schwartz’s conceptual progress was nevertheless a major step forward since it helped ask the right questions.
Philippe Lazar
Former Director-General of INSERM[5] (1982-1996)
[1]CESAM: Centre d’enseignement de la statistique appliquée à la médecine et à la biologie médicale -Teaching Centre for Statistics Applied Medicine and Medical Biology
[2]National Advisory Commission for Ethics in Life and Health Sciences.
[3]Explanatotory and Pragmatic Attitudes in Therapeutical Trials, Daniel Schwartz and Joseph Lellouch, Journal of Chronic Diseases 1967; 20; 637-648.
[4]L’essai thérapeutique chez l’Homme, Daniel Schwartz, Robert Flamant and Joseph Lellouch, Flammarion Médecine-Sciences, 1970; English translation: Clinical Trials, Academic-Press, Michael Healy, 1980.
[5] INSERM: Institut National de la Santé et de la Recherche Médicale – National Institute for Health and Medical Research.
John Hajnal (1924-2008)
Professor John Hajnal was an economic historian, a demographer, and a mathematical statistician; a deep and original thinker with an interest in social and educational issues, and religion. His publications were numerically few, but had profound influence. In demography, the Hajnal line is named after him. This is an imaginary geographical boundary connecting St. Petersburg, Russia, and Trieste, Italy, distinguishing the historic marriage patterns of Eastern and Western Europe. In the (demography-related) mathematical theory of products of non-negative matrices, two papers, published in 1958 and 1976, were fundamental for the revival of the theory of inhomogeneous Markov chains; and for the use of coefficients of ergodicity as a tool for addressing the convergence of non-negative matrix products. My work on this theory, from its early stages in the early 1970’s, was greatly influenced by his, and we soon established written contact and later became friends. Another who became a friend and whom he influenced in both demography and matrix theory was Joel E. Cohen of Rockefeller University, New York.
John Hajnal was born in Darmstadt, Germany, on 26th November 1924 to Kálmán and Eva Hajnal-Kónyi. His Jewish family had a strong Hungarian identity, and a passion for German high culture. John was intellectually gifted and a linguist. With the onset of the Nazi regime, Jewish children were discriminated against in school, and in 1936, while his parents prepared to leave Germany for England, John was sent to a Quaker school in The Netherlands. He described his time there as one of the happiest of his life. After a year, he was reunited with his parents in London, before The Netherlands was overrun by the Nazis. He remained a fluent Dutch speaker and enthusiastically attended reunions of the school in the later years of his life.
His secondary school education was at University College School in London. He went to Balliol College, Oxford, at age 16. John’s school had steered this brilliant young man towards classics, although his interests were more in science and mathematics. He switched to politics, philosophy and economics in his second year and gained a First Class Honours degree in 1943. However, he regretted missing out on early formal mathematical education for the rest of his life, believing that mathematics was his true vocation, and that mathematicians did their best work early in their lives. As a mathematician, John was largely self-taught, but had some contact with the probabilists William Feller (during John’s period at Princeton), as well as Maurice Bartlett and Walter Ledermann during his time at Manchester. Later, one of John’s Ph.D. students, Bill Hamilton, became eminent through applying mathematics towards the understanding of altruism in theoretical biology.
John was at the Royal Commission on Population, 1944-1948, and developed what became a lifelong friendship with his boss, Bryan Hopkin. Based on the strength of his work, he was recruited by Frank Notestein to work on demography at the UN in New York; he stayed there from 1948-1951. In New York, he met Nina Lande of Jewish stock and they married soon after. Her father, Lev Lande, a prominent Menshevik politician, left Moscow in 1917 and Nina was born in Berlin.
John gave up his well-paid and secure job at the UN and moved to Princeton. Here at the Office of Population Research at Princeton University, 1951-1953, he continued occasional work for Frank Notestein while taking courses in mathematics. The family then returned to England, where, during 1953-1957, John was at Manchester University at the Department of Social and Preventive Medicine.
For family reasons, the Hajnals moved to London in 1957, when John secured a lectureship at the London School of Economics (LSE). John started at LSE as a demographer, gradually making a transition to theoretical statistics, then a promotion to Reader in 1966 and to Professor of Statistics in 1975. He became an elected member of the ISI in 1961 and a Fellow of the British Academy in 1966. By 1964, he was sufficiently recognized to be asked to join the galaxy of stars who constituted the Editorial Board at the founding of Joe Gani’s Journal of Applied Probability. He retired from LSE in 1986.
His periods of study leave included Visiting Fellow Commoner, Trinity College, Cambridge, 1974-1975 (during this time he wrote the 1976 paper); Visiting Professor, Rockefeller University, 1981; and one or two brief visits to the Demography Department, Institute of Advanced Studies (IAS), Australian National University (ANU - one such occasion was in 1988).
P.A.P. (Pat) Moran (1917-1988) was Professor and Head of the Department of Statistics at the IAS, ANU. In 1988, a group of visiting statisticians including John and me (there was a conference in Canberra, timed for the national bicentenary) visited him at his home. He was already wheelchair-bound and in fragile condition, having suffered a stroke (he died later that year). John was very moved by the gift of a small book on statistical genetics (an interest they shared) that Pat gave him at the time. When it was time for us to go home, we all packed into the car, which someone had rented, outside Pat’s home. The car’s battery was dead. Pat’s home was on a slightly sloping road, however, and when I offered to try a rolling start, the engine fired, which was met by a gratifyingly loud “Bravo!” from the back seat, unmistakably in John’s precisely-dictioned voice. Later, when John stayed at our home in Sydney for a few days, we had many walks and discussions motivated by our similar European origins on social and religious issues, especially, on the need (or not) to belong to some tribal group.
In 1991, there was a conference (titled Doeblin and Modern Probability) to honour the memory of the brilliant young Jewish mathematician, Wolfgang (Vincent) Doeblin, who was killed while in the French army in World War II, and who played an instrumental role in the development of Markov chains, amongst other probabilistic areas. The conference was held in Blaubeuren, Germany, with the participation of esteemed colleagues (such as J.L. Doob, K.L. Chung, M. Iosifescu, J. Hajnal, and J.F.C. Kingman) and other well-established colleagues, including many from Communist bloc countries. It was thematically inclined towards inhomogeneous Markov chains. John Hajnal’s presentation, “Shuffling with Two Matrices”, was published in the conference proceedings in 1993 in the series Contemporary Mathematics (149). The paper was inspired by an example in a paper of Pat Moran in Sankhya, Ser. A. in 1962 on a similar topic, both in lucidity and content, is a worthy successor to his two seminal papers, also as a tribute to Doeblin.
I first visited John and Nina’s home in Golders Green in London in 1985. I spent an afternoon on Hampstead Heath with John, who remembered at each of our subsequent meetings my admiration for the lovely avenue of lime trees. During one visit, John was happily preparing for his role as guest of honour and lecturer at a historical demography symposium in Iceland, which took place in 2003. The last time I stayed with the Hajnals was in 2007. Nina drove us to Highgate Cemetery, by then John was using a walking stick, though we walked back to their home. During each of our visits to Highgate Cemetary, and this one was no different, John liked to point out to me the memorials to Karl Marx and Sigmund Freud, and the alcove with a chair and Bible, where John used to sit and read.
Both John and Nina were diagnosed with cancer later in that year. Nina, who had been my contact through e-mail, died unexpectedly on 25th April 2008. John kept reasonably well for several months, but eventually his condition deteriorated and he died on 30th November 2008.
A commemorative article entitled “John Hajnal on the Prospect for Population Forecasts” leads off on pages 189-192 with a brief biographical sketch of John’s life and outstanding contributions to demography. The rest is a reprinting of his landmark paper “The Prospect for Population Forecasts” in full from Volume III of the Proceedings of the 1954 World Population Conference held in Rome. The paper, in its longer version, appeared in the June 1955 issue of the Journal of the American Statistical Association. The most influential paper (on the Hajnal line, still said to be a stimulus of research and debate) was entitled “European Marriage Patterns in Perspective”, published in the 1965 volume of Population in History, edited by D.V. Glass and D.E.C. Eversley. A sequel to this article was published in Population and Development Review in the September 1982 issue.
To this list, to expand on John’s creativity and movements, should be added the paper “Concepts of Random Mating and the Frequency of Consanguineous Marriages”, Proceedings of the Royal Society, B159 (1963), 125-174. This was a discussion meeting and a great honour. The paper is reproduced (without the discussion) in Demographic Genetics (1975) edited by Kenneth M. Weiss and Paul A. Ballonoff, in the series Benchmark Papers in Genetics. In this, John is in the company of the esteemed of statistical demography and genetics, including J. Graunt, A.J. Lotka, V. Volterra, E.G. Lewis, C. Darwin, R.A. Fisher, J.B.S. Haldane, H.T.J. Norton, S. Wright, A.J. Bateman, and L.L. Cavalli-Sforza. In this paper, he cites three of his own early papers from 1950, 1953 and 1960. A footnote states that part of the work was done at Dartmouth College, New Hampshire, USA, supported by an NSF grant to the Dartmouth Mathematics Project, as well as by the Population Investigative Committee, London.
John and Nina, who are survived by four children, are buried together in Golders Green Cemetery. The headstone says of John: Demographer and Statistician. Lover of knowledge who followed his own path.
Eugene Seneta
ISI Committee Matters: Astrostatistics Committee and Network
An ISI Astrostatistics Interest Group was founded in 2008 with the purpose of exploring the feasibility of creating a standing astrostatistics committee under the auspices of the ISI. There had never been such a committee formed under a statistical nor an astronomical organization. Some thirty statisticians and astronomers having an interest in astrostatistics met at the ISI Session in Durban, in August 2009, to discuss the possibility of such a committee, and voted unanimously to propose this committee to the ISI Executive Committee and Council. I drew up the proposal that was presented to the ISI EC and Council by President Jef Teugels in December 2009. I learned of the approval from President Teugels on December 24th.
Astrostatistics is perhaps the oldest of the applied statistical disciplines. Although we know little of the details, it appears that elementary descriptive statistics were being applied by Babylonia astronomers to planetary events as long ago as the second millennium BCE. The Venus Tablet of Ammisaduqa, for example, records the heliacal risings and settings of Venus for a 21-year period from 1646 to 1625 BCE. We know that the mean was used by later Chaldean astronomers, in the 7th and 9th centuries, who claimed to obtain their knowledge from these earlier periods. Hipparchus, in the 2nd century BCE, appears to have employed the median, and used the mean and a primitive form of confidence interval to describe various surprisingly accurate measurements relating the positions of the Sun, Moon, and Earth.
Inferential statistics was also heavily influenced by astronomy in its early implementations. Gauss engaged the first serious application of least squares regression for calculating the likely orbit of the asteroid Ceres behind the sun in 1801. Legendre then applied least squares to cometary orbits (1805). After this auspicious beginning, however, advancements to statistical methodology tended to come from applications to agriculture and gambling. Advances in mathematical astronomy did not come from statistics as much as from differential equations.
In recent years, astrophysicists have come to realize that the new statistical methodologies that have been developed as a consequence of powerful computers can be applied to understanding astronomical data. It was not until the 1990’s that a true renaissance of astrostatistics began, with the emergence of various centers and associations devoted to the application of statistics to astronomical data. To be sure, there were some astrophysicists who employed statistical techniques to their data before this, but there were no consistent efforts to conjoin statistics and astronomy.
The most influential astrostatistical affiliations today are perhaps the California-Harvard Astrostatistics Collaboration (CHASM), founded in 1997, the Pennsylvania State University Center for Astrostatistics (PSCOA), founded in 2003, the International Computational Astrostatistics Group (InCA), and the CosmoStat conferences that have been held the past several years in Europe. A number of short-term educational programs in astrostatistics have appeared throughout the U.S. and Europe, as well as in coordination with PSCOA in India. However, until now, there has been no global association of astrostatisticians, nor a central committee or group that can represent the discipline. The ISI Astrostatistics Committee and Network aims to serve that purpose.
A driving force behind the renewed interest in astrostatistics is the type of data being gathered by both land- and space-based instruments. Astronomers began to realize that the basic statistical tools they used in their research would not be sufficient for the analysis of the mass of data generated from the new surveys, e.g. the Sloan Digital Sky Survey, which mapped some 930,000 galaxies and 120,000 quasars in the first eight years of its existence. Forthcoming surveys will be creating a terabyte or more of information per night, requiring sophisticated statistical methods that are outside the purview of most astronomers. In order to satisfy the need for collaboration between the statistical and astronomical communities, the ISI Astrostatistics Committee will coordinate with the International Astronomical Union, Jet Propulsion Laboratory, European Space Agency and other astronomical organizations in the fostering of collaborative efforts.
Members of the ISI Astrostatistics Committee represent some of the leading figures in the discipline. The Committee will serve as the executive committee of the Network and will serve to support existing astrostatistical organizations and groups, as well as sponsor colloquia and papers’ sessions, prepare astrostatistical educational tools, and sponsor a proposed Journal of Astrostatistics. I believe that the creation of the ISI Astrostatistics Committee and Network of astrostatisticians will help to advance the discipline, and contribute to both the fields of statistics and astronomy.
I encourage anyone with an interest in the application of statistics to astronomical data to join the ISI Astrostatistics Committee and Network and to contact me at hilbe@asu.edu or j.m.hilbe@gmail.com. Please list your name, e-mail address, website if applicable, and your major statistical or astronomical area of expertise.
Members of the inaugural ISI Astrostatistics Committee are: G. Jogesh Babu (US), Bruce Bassett (South Africa), Kirk Borne (US), David van Dyk (US), Eric Feigelson (US), David Hand (UK), Joseph Hilbe (US), Steffen Lauritzen (UK), Gary Miner (US), Nazeen Mustapha (South Africa), Victor Panaretos (Switzerland), Christian Röver (Germany), Anand Sookun (Mauritius), Simon Vaughan (UK) and Evdokia Xekalaki (Greece).
Joseph M. Hilbe
Chair
ISI Astrostatistics Committee
Historical Anniversaries: Thomas Simpson
Thomas Simpson was born in Market Bosworth, Leicestershire, England, on 20th of August 1710. His father, who was a weaver, wanted Thomas to take-up his profession. This is why Thomas received little formal education and taught himself mathematics.
To avoid family conflicts, he left home at age 14 and went to a neighbouring village to the lodging house of Mrs. Swinfield, the widow of a tailor, and his senior by about 30 years with two children more or less of Thomas’ age. He continued to work as a weaver and to educate himself. A fortune teller who was lodging in the same house lent him a book on arithmetic and another on astrology leading him to become a fortune teller himself. A short time after this, he married Mrs. Swinfield. Due to an unfortunate instance of fortune telling, Thomas and his family moved to Derby where, besides his work as a weaver, he became an instructor at an evening school. At the age of 25 he moved to London still working as a weaver and teaching mathematics.
From the time he came to London, he contributed (using a variety of pseudonyms) many mathematical problems to the Ladies’ Diary, answering problems raised in this popular annual publication. He even became Editor of the magazine in 1754 for a period of six years. He also contributed to other periodicals like Gentleman’s Magazine, Miscellanea Curiosa Mathematica and Gentleman’s Diary. His mathematical skills were demonstrated in these publications and allowed him to become known by other mathematicians.
In 1743, he left London to become a teacher of Mathematics at the Royal Military Academy of Woolwich. Two years after his appointment, Simpson was elected a Fellow of the Royal Society.
He became a successful writer and published 11 books between 1737 and 1757: On the theory of fluxions (1737), on the laws of chance (1740), speculative and mixed mathematics (1740), annuities and reversion (1742), on a variety of physical and analytical subjects (1743), algebra (1745), geometry (1747), trigonometry (1748), on fluxions (1750), select exercises in mathematics (1752) and miscellaneous tracts on mechanics, physical astronomy, and speculative mathematics (1757). These publications where successful in England − with numerous editions − but also in America, France and Germany; they also raised controversy and accusations of plagiarism. In 1758, Simpson was elected a Fellow of the Royal Swedish Academy of Science.
Simpson used the numerical method known today as Simpson’s rule in his work on interpolation and numerical methods of integration, but he learned this method from Newton as he himself acknowledged. Two main contributions are due to Thomas Simpson. The first one, in the field of actuarial science, concerns the calculation of joint-life annuities based on the earlier works of De Moivre. The second one is a contribution to statistical error theory.
In 1755, Simpson published A Letter to the Right Honourable George Earl of Macclesfield, President of the Royal Society, on the Advantage of Taking the Mean of a Number of Observations, in Practical Astronomy in the Philosophical Transactions of the Royal Society of London. Simpson’s study was limited but enterprising. The development was limited, because he showed only that a mean is better than a single observation when the mean is calculated from six measurements and when one takes into account a limited hypothesis about the probabilistic features of the observations. However, it was also enterprising because Simpson chose to focus not on the observations themselves, but on the errors made on the observations. Specifically, he assumed that each of the independent observations was prone to errors having a known symmetric distribution. His calculation was made for the case where this distribution was triangular. Inspired by De Moivre’s earlier study of the distribution of the sum of the points on the sides of several dice, Simpson used the generating function that later became an essential tool for Lagrange and Laplace.
Thomas Simpson remained active until the end of his life. For example, it is interesting to note that he and Roger Joseph Boscovich met in 1760 and that the latter discussed a problem with him in least deviations regression; Simpson proposed an analytical solution to this problem.
He died on 14th of May 1761 in Market Bosworth, the village of his birth.Jean-Jacques Droesbeke
Chair
Christiaan Huygens Committee on the History of Statistics
Announcements
Call for Papers for a Special Issue of the International Statistical Review on
Energy Statistics
The ISI plans to publish a special issue of the International Statistical Review (ISR) with the theme of Energy Statistics. I am very pleased to announce that Dr. Carol Joyce Blumberg (U.S. Energy Information Administration, USA) and Dr. Abdel H. El-Shaarawi (National Water Research Institute, Canada) have agreed to be the Guest Editors.
It is hoped that the issue will contain papers covering different aspects of energy statistics. Some of the manuscripts selected for publication may be supplemented by a discussion by appropriate experts and a rejoinder by the author(s).
Some general areas for papers are:
* reviews/surveys of significant developments in theory, methodology, statistical computing and graphics as related to energy statistics
* relationships between energy statistics and statistics education
* expository papers on emerging areas of research or application related to energy statistics
* papers describing new developments and/or challenges in energy statistics
* papers on the history of energy statisticsSome general topics within energy that would be of high interest for papers are:
* measuring the supply and/or consumption of energy
* forecasting future supply and/or consumption (demand) of energy
* forecasting future energy prices
* data quality issues
* issues related to renewable energy sources statistics
* measuring environmental effects of energy usePapers will be refereed. Do not submit papers that have previously been published or accepted for publication in refereed publications. Submit papers by 9th January 2011 to Liliana Pinkasovych at the ISI Permanent Office (e-mail: @cbs.nl) as an attachment with a cover letter. The cover letter should indicate that the manuscript is for the Special Issue of ISR on Energy Statistics. Manuscripts should be submitted as either Word or PDF files. On acceptance of a manuscript, the author will be asked to send an electronic version of the paper, preferably in LaTeX. Further Instructions to Authors are available on the ISR websites at http://isi.cbs.nl/ISR/isr-authors.htm and http://www3.interscience.wiley.com/journal/117998483/home.
2010 International Methodology Symposium
Statistics Canada
October 26-29, 2010 Ottawa, ON, Canada
Social Statistics: The Interplay among Censuses, Surveys and Administrative DataCall for Contributed Papers (PDF
, in English and in French)
A Successful European Establishment Statistics Workshop
The European Establishment Statistics Workshop (EESW09), gathering practitioners and researchers active in the various areas of the value chain in establishment statistics (i.e. statistics on businesses, organisations or on economy, produced from surveys or from administrative data), was held from 7-9 September 2009 in Stockholm, Sweden.
Short papers and/or presentations prepared for the workshop are posted on the workshop website, www.eesw09.eu, since November 2009.
ISOSS Conference in Cairo Reinforces Statistics for Development and Good Governance
The 10th Islamic Countries Conference on Statistical Sciences (ICCS-X) was successfully held at the American University in Cairo (AUC), New Cairo, Egypt, with over 300 participants from all over the world. The biennial Conference was jointly organized by the Islamic Countries Society of Statistical Sciences (ISOSS), Egyptian Cabinet Information and Decision Support Centre (IDSC) and AUC. The Conference was financially sponsored by the Islamic Development Bank (Jeddah, Saudi Arabia), the Egyptian Ministries of Tourism and Investments, IDSC and AUC. The Conference was dedicated to the late Professor Mir Maswood Ali, a prominent Statistician of Bangladeshi origin and key founder of the Department of Statistical and Actuarial Sciences, University of Western Ontario, Canada.
Standing (from left): Kaye Bashford (Australia), Sharif A. Abdelhalim (IDB, Saudi Arabia), Magued Oslam (Chair, LOC, Egypt), Jef Teugels (ISI President, Belgium), Zeinab Amin (Co-Chair, LOC, Egypt), Mohammad Hanif Mian (ISOSS Vice-President, Pakistan), Edward Wegman (USA), Jim Berger (USA).
Sitting (from left): Shahjahan Khan (ISOSS President, Australia), Wafik Younan (Treasurer, LOC, Egypt), Ali S Hadi (ISOSS President-Elect, USA/Egypt).In the opening session, Dr. Lisa Anderson, Provost of AUC, spoke about the importance and diverse applications of statistics from the ancient need of the states to the modern day public affairs. The ISOSS President, Dr. Shahjahan Khan of the University of Southern Queensland, Australia, emphasized the essence of engagements between the statisticians from the developed countries with those in the developing countries, especially with those of the IOC member states for improving the quality of government statistics and statistical research as well as enhancing its state of the art applications. Dr. Magued Osman, Chair of the Local Organizing Committee and Head of IDSC, spoke about the role of statistics for development and good governance. Dr. Ali S. Hadi, Vice-Provost of AUC, current Chief Editor of International Statistical Review, Chair of the Scientific and Program Committee, welcomed the participants while also covering different scientific and cultural activities of the Conference.
Sitting (from left): Edward Wegman (USA), Jef Teugels (ISI President, Belgium), Mohammad Hanif Mian (ISOSS VicePresident, Pakistan), Jim Berger (USA), Shahjahan Khan (ISOSS President, Australia).
Standing (from left): Sharif A. Abdelhalim (IDB, Saudi Arabia), Rasheed AlSuwaidi (DG, NBS, UAE), Rashid Salaria (Pakistan), Magued Osman (Chair, LOC, Egypt), Zeinab Amin (Co-Chair, LOC, Egypt), Ali Hadi (ISOSS President-Elect, Egypt/USA), Motiur Rahman (Bangladesh), Wafik Younan (Treasurer, LOC, Egypt), and Ishaq Bhatti (Australia).The highlights of the Conference included the participation of four keynote speakers. First, Dr. Jef L. Teugels, President of the International Statistical Institute (ISI), presented a keynote address on the extreme value distributions with applications. He analyzed data of natural calamities/disasters with extremes coming from the 1970 cyclone in Bangladesh and Hurricane Katrina in southern USA. Dr. Jim Berger covered the Bayesian adjustments of multiplicity in the testing regime of huge number of tests coming from multi-disciplinary scientific studies. Dr. Edward Wegman, who testified in the US Congress twice on the scientific aspects of climate change issues, discussed the rapid changes in the data science, and took the audience to the outer universe of huge datasets and the associated challenges to analyze them for scientific applications. Dr. Kaye Bashford, former President of the Statistical Society of Australia Inc., presented her keynote address on some applications of multivariate data analysis for determining the best quality of wheat production. The data presented in the talk came from an international team of experts working on the project around the world.
Standing (from left): Jef Teugels (ISI President), Shahjahan Khan (ISOSS President, Australia), Magued Osman (Chair, LOC, Egypt), Mohammad Hanif Mian (Pakistan), Edward Wegman (USA), Jim Berger (USA), and Ali S. Hadi (ISOSS President-Elect, USA/Egypt).
The theme of the Conference was Statistics for Development and Good Governance. The following series of Panel Discussions related to this theme were presented. Panel I: Public Opinion Polling & Good Governance (Speaker: Magued Osman, Moderator: Dina El Khawaja, and Discussants: Jennifer Bremer and Hafez Al Mirazi), Panel II: Measuring the Unmeasurable (Speaker: Anis Yusoff, Moderator: Mostafa Kamel El Sayed, and Discussants: Nadia Makary, Ghada Moussa and Andrew Stone), and Panel III: Indicators & Politics, The Ibrahim Index for African Governance (Speaker: Ali S. Hadi, Moderator: Lisa Anderson, and Discussants: Lisa Anderson, Stephen Everhart and Nabil Fahmy).
Opening session of ICCS-X (left to right): Jef Teugels (ISI President), Shahjahan Khan (ISOSS President, Australia), Lisa Anderson (Provost of AUC), Magued Osman (Chair, LOC, Egypt), and Ali S. Hadi (ISOSS President-Elect, USA/Egypt).
Another salient feature of the Conference was the presentation of invited sessions on topics such as statistics education, demography and aging, small area sampling, medical meta-analysis, statistical inference, astrostatistics, directional data analysis, etc. These sessions attracted leading scholars and researchers of these areas which benefited the participants enormously. Young and new researchers found the sessions stimulating for their future research. More details on the Conference Program and other activities of ISOSS can be found on the website: http://www.isoss.com.pk/.
The Nile Cruise Gala Dinner was held on a luxurious boat and the performance by the young Egyptian men was outstanding. Some participants of the Conference and the accompanying family members also participated in the interesting dance and performances. The trip to the pyramids in Giza was a once in a lifetime experience for many participants.
ICCS-X participants in a scientific session
Relating to the history of the invention of paper from the papyrus and the collection of essence from the lotus were popular items of purchase. Many went to the Egyptian Museum to see the mummies of the Pharos, including Ramses II, many consider to be the Pharaoh of Moses. The rich and colorful history of the ancient Egypt includes the first ever census conducted by Moses to count the men of Egypt at the time.
ICCS-X participants in a scientific session
The business session of the Conference was held on the evening of 21 December at the Marry Cross Hall of AUC. The President of ISOSS chaired the session and reported the main activities and achievements of ISOSS during the last two years following the Conference in Kuala Lumpur, Malaysia, in December 2007. The participants noted that in recent years ISOSS has become more visible in the international community of professionals and various statistical bodies along with receiving global recognition. The session re-elected Dr. Shahjahan Khan as President of ISOSS for 2009-2011, and elected Dr. Ali S. Hadi as President-Elect for 2009-2011. The session endorsed the call of Dr. Munir Ahmed, founding President of ISOSS, for donations of funds to the ongoing ISOSS House Construction Project in Lahore, Pakistan. A number of participants committed to contribute US$1,000 and US$500 during the session.
Full report (
Full report (Shahjahan Khan, President of ISOSS
Department of Mathematics & Computing
University of Southern Queensland, Australia
Memories of the Past
Delegates mixing at the 41st ISI Session in New Delhi, India, in 1977,
with the Honourable Mr. Morarji Desai (centre of the photograph).
Did you know the Indian Government issued a special commemorative stamp to celebrate the New Delhi Session?
Should any of you know of its whereabouts, please share it with all of .
100th Issue Back Cover
ISI Director Ada van Krimpen and ISI President Jef Teugels pose with Statistics Netherlands’ Director-General Gosse van der Veen (centre) after the unveiling of the ISI nameplate. ISI Executive Committee, from left to right: Vice-President Louis Chen, President Jef Teugels, President-Elect Jae C Lee, Vice-President Hallgrímur Snorrason.
Missing-in-Action: Vice-President Vijay Nair.
President Jef Teugels opens the champagne in the ISI hall while Director-General Gosse van der Veen speaks with other Executive Committee members in the background. The social gathering was organised by Statistics Netherlands for the ISI Executive Committee and ISI Permanent Office staff. ISI Permanent Office staff can barely be seen from behind the new nameplate.
Back to Home Page
News from ISI Sections Volume 34, No. 1 (100) 2010
