Please first choose your favorite format:
Default Style | Large fonts | No Graphics
Published three times a year, the ISI Newsletter provides a broad overview of the Institute's activities, and also includes additional information of interest to statisticians. The Newsletter is sent to all members of the ISI and its Sections (approx. 5,000) as part of their membership.
Editors: Mr. D. Berze and Ms. S. Mehta, Graphic Designer: Mr. H. Lucas
In this online Issue Message from the President Message from the Director News of Members Statisticians share Nobel Peace Prize In Memoriam Survey of ISI Members Regarding the Lisboa Session Awards, Prizes and Competitions Honorary Member Interviews: Professor George Box ISI Committee Matters: Khawarezmi Committee ISI Membership Elections 2007 Historical Anniversaries: Max Planck (1858-1947) Memories of the ISI's Past Calendar of Events News from ISI Sections Volume 32, Number 1 (94) 2008
Message from the President
The first few months of my ISI Presidency have sped by so fast. I have enjoyed a number of exciting visits and am grateful to those organisations that have paid my travel expenses, mindful of the need to keep ISI expenditure as low as possible.
I was invited to South Africa by the Statistician General, Pali Lehohla. The week I was there marked the publication of the first results from the South African community survey and I was fortunate enough not only to be present at the press conference and the celebratory party for Stats SA staff (somewhat different from other office parties I have experienced and I hope a foretaste of the fun we will have at the ISI Session), but also to accompany the Statistician General and the Chair of the Statistics Council at a meeting of the Cabinet where they presented the results. I have never attended a Cabinet meeting before and was pleased to see the interested and enthusiastic reception given to the survey. I was also amused to find that the Cabinet members were dressed in rugby strip and that I was especially welcome coming from England, the losing side in the rugby!
Statistics South Africa is adopting some schools in order to help them improve the quality of their mathematics education and encourage more young people to study mathematics (and to consider statistics as a career). I was so privileged to attend the ceremony to open a new mathematics resource centre and prize giving at a rural girls’ school in KwaZulu-Natal. The commitment of the teachers together with the bright optimism and determination of the girls was wonderful. I look forward to reporting further on this programme of Statistics South Africa as it progresses.
My visit incorporated meetings in Durban about the planning for the ISI Session – I hope the dates – 16-22 August 2009 – are firmly in your diaries.
The following week saw the South African Statistical Association’s annual conference, which was a lively and stimulating meeting, set in a lovely country location. Ben Kiregyera, past recipient of the Malahanobis Prize, was one of the keynote speakers talking about the new Centre for African Statistics, which he directs, located in Addis Ababa. Ben and I led some workshops for attendees from national statistical offices across sub-Saharan Africa. We are organising a meeting of African statisticians later this year to share experiences in establishing national statistical societies, and I hope this will be jointly led by the ISI, RSS and ASA. If other statistical societies are interested to assist with this, I would be pleased to hear from them.
In December, I attended my last Board meeting of the American Statistical Association, where I have just completed my three years as the international representative. I was very pleased to be asked to talk about my personal views on the ways in which the ASA might – in collaboration with the ISI – provide greater support for statisticians in developing countries. I look forward to a discussion of this topic in the ISI, most particularly with statisticians who are themselves from developing countries.
In November, the Arab Institute for Training & Research in Statistics organised, with the collaboration of the Jordanian Statistical Office, the First Arabic Statistic Conference. Its theme was “No Development without Statistics”. The Conference was opened by Dr. Maroof Al-Bakheet, the Jordanian Prime Minister. I was very honoured to represent the ISI.
More than 300 researchers and experts from 19 Arabic countries participated in this Conference. There were lively debates on a wide variety of topics within four themes: the importance of capacity building in statistics; the current position and future prospects of Arabic statistics; the relationship between theoretical and applied statistics; and the role of Universities and Research Centres in the development of statistics. An open meeting of the Khawarezmi Committee of the ISI was held in order to explore ways in which this Committee might forge stronger statistical liaisons within the Arab region and across to other regions of the world. We have included a report from this Committee (more).
Over the last few months, I have been very happy to speak at two events in the UK arranged by and for young statisticians. One was an evening meeting of the Royal Statistical Society and the other a two-day conference of official statisticians in the early stages of their careers. The ISI Executive Committee is keen to gather ideas as to how we (the ISI and Sections) might organise activities and services of especial interest to young statisticians. Len Cook (ISI Vice-President) arranged a very successful reception for young attendees at the ISI Session in Lisboa and this was followed up by a consultation held with young statisticians by President-Elect Jef Teugels. If you have views on our involvement of young statisticians, please contact Jef.
I represented the ISI at two other conferences since the previous Newsletter, both on very important topics – one in Brussels on the topic “Beyond GDP” (you can view web-streamed presentations on http://www.beyond-gdp.eu/) and the other in Luxembourg, a Eurostat meeting, entitled "Modern Statistics for Modern Society". The Prime Minister of Luxembourg, Mr. Jean-Claude Juncker, and European Commissioner, Mr. Joaquín Almunia, were among the speakers presenting their views on future developments and challenges for statistics.
As always, I welcome hearing directly from ISI and Section members and can be reached at email@example.com.
My best wishes for a happy and healthy 2008.
Message from the Director
In the course of my professional life, I come into regular contact with people who are very curious about the ‘raison d’être’ of our association. ‘Does your organisation compile, retain and disseminate statistical information?’ No - although we do guide persons to the appropriate sources of statistical information they require. ‘Well then do you act as statistical judges, blessing some statistical output while condemning others?’ No again, although we do have a Professional Ethics Committee that examines various contemporary statistical issues, providing guidelines for statisticians regarding appropriate ethical practices. Unless one actually gets into our organisation as an active member, taking advantage of all we have to offer, it is difficult to fully appreciate that the ISI and its Sections are essentially ‘network hubs’, linking statisticians with each other and organising them according to their specific interests, thus enabling the profession and its practitioners to interact and progress. This is essential work, as we somewhat ironically are experiencing a dramatic increase in the use of statistics throughout society (matched by a public hunger to understand more about them), while on the other hand many people responsible for generating these statistics (or the methodology behind them) do not see themselves as being statisticians. Statisticians must unite, on a regional, national and international basis, for the benefit of each other and the society in which we live. There is ‘power in numbers’. In the coming year, the ISI will be working to increase its interactions with national statistical societies, an essential link in our efforts to build stronger international ties between statisticians. We will also be investing time and resources in the improvement of our communication outputs, including our website, this Newsletter, and other communication resources.
I am pleased to announce that in the second round of the 2007 ISI membership elections, 32 new candidates were elected as ISI members. We include the names and nationalities of these new members here, as well as the 68 persons who were elected in the first round of the 2007 membership elections, a few of whose names were mistakenly omitted in the previous edition of the Newsletter. Prof. George Box was also elected to ISI Honorary Membership, and we have included a special interview with him. (more)
I would like to thank all of you who participated in the ISI Post-Session Survey last year. We include a brief summary of the report kindly prepared by Prof. Gilbert Saporta and his assistant Dr. Emmanuel Jakobowicz, with the assistance of ISI President Prof. Denise Lievesley. (more)
Please note that an announcement of the availability of the Lisbon Session proceedings (on DVD) will be made in the next edition of the ISI Newsletter. These proceedings will eventually be placed on the ISI website for all to access.
I would like to thank all of you who have so diligently processed your membership dues payment for 2008. If you have not yet had the opportunity to do so, we would be grateful if you could process your dues payment in the coming weeks. Should you have any questions about your invoice, please contact Mr. Michael Leeuwe (@cbs.nl) or Mr. Sieriel Hoesenie (@cbs.nl).
The ISI History of Statistics Committee Chair Prof. Jean-Jacques Droesbeke has again provided an insightful article, this time celebrating the 150th anniversary of Max Planck’s birth. (more)
Please take note of the ISI and IASS co-ordinated competitions that we are again launching for the coming biennium, as announced here. Please help us to circulate news about these competitions to potential participants.
Finally, I would like to bring your attention to the announcement that we have received from Dr. Juana Sanchez, the dynamic Chair of the IASE’s International Statistical Literacy Project, regarding the ‘International Statistical Literacy Competition’ and the call for nominations for the ‘2009 Best Cooperative Project Award in Statistical Literacy’. (more)
News of Members
ISI elected and IASS member Professor Sir Roger M. Jowell received a knighthood for ‘services to Social Science’. Professor Jowell is the Co-Founder and Director of the European Social Survey, as well as a research professor at City University in London, UK. The announcement was made on 29th December 2007 to mark the beginning of 2008 with the New Year’s Honours. The ceremony with Queen Elizabeth II will take place on 13th February 2008.
Professor Fionn Murtagh
ISI elected and IASC member Professor Fionn Murtagh has been appointed Director of Information and Communication Technologies at Science Foundation Ireland.
Professor Jianqing Fan, ISI elected member, received the Morningside Gold Medal of Applied Mathematics “for his ground-breaking and seminal work in nonparametric modeling and inferences, for his fundamental contributions to high-dimensional statistical learning, nonlinear time series and biostatistics, and his achievements toward the development of novel statistical techniques in finance and molecular biology”. The Morningside Gold Medal honours outstanding mathematicians of Chinese descent under age 45, and is awarded once every three years at the International Congress for Chinese Mathematicians (ICCM). It is the highest honour of the ICCM and carries a gold medal, a citation and a cash award. Professor Jianqing Fan is the only recipient of the Morningside Gold Medal of Applied Mathematics this year, presented at the 4th ICCM on 17th December in Hangzhou, China. Please see http://www.cms.zju.edu.cn/iccm2007/ for more details.
Dr. Shahjahan Khan
President of the Islamic Countries Society of Statistical Sciences (ISOS), ISI elected and ISBIS member, Dr. Shahjahan Khan, presented a keynote speech at the First Arab Statistics Conference (FASC) organised by the Arab Institute for Training and Research in Statistics (AITRS) and was held from 12-13 November 2007 in the Radisson SAS Hotel, Amman, Jordan. The AITRS was established in 1971 to raise competencies and skills of the Arab Statistical Organisation. The Conference is organised under the patronage of his Royal Majesty King Abdullah II Ibn Al-Hussain of Jordan, and the session in which Dr. Khan presented his keynote speech was chaired by the Minister of Planning of Jordan, Dr. Suhair Al-Ali. An Australia-based academic, Dr. Khan spoke on the Importance of Statistics in Development in line with the theme of the Conference No Development without Statistics.
News from the ISI Permanent Office
Mr. Hans Lucas, ISI Webmaster, celebrated his five-year anniversary at the ISI on 20th January 2008. We congratulate Hans on this occasion.
Statisticians share Nobel Peace Prize
Can you name the statisticians who received a 2007 Nobel Prize? In the past, only a few statisticians working in econometrics have been Nobel Prize winners. However, this year was different. At least nine of our colleagues, including some ISI members, are now Nobel awardees.
The 2007 Nobel Peace Prize, awarded by a committee from the Norwegian Parliament, went to former US Vice-President Albert Gore and to the Intergovernmental Panel on Climate Change.
"This is an honour that goes to all the scientists and authors who have contributed to the work of the IPCC, which alone has resulted in enormous prestige for this organization and the remarkable effectiveness of the message that it contains", says Mr. Rajendra Pachauri, the Chairman of the IPCC. The IPCC has its headquarters in Geneva. It was established in 1988 by the United Nations Environment Programme and the World Meteorological Organization. The main product of IPCC is its Assessments of Climate Change, the fourth of which (Climate Change 2007) was recently published.
The assessments result from contributions written by hundreds of scientists. Various articles are written especially for the IPCC. These are edited and put together by section editors, carefully peer reviewed and discussed in public meetings. The executive summaries of the assessments are analyzed line-by-line by representatives of over 100 governments. The resulting assessment is as close as one can possibly get to a consensus of the scientific community.
Are there any statisticians involved in the IPCC? Yes, but only a few. The most prominent is Canadian statisitical climatologist Francis Zwiers of Environment Canada in Toronto. He has participated as reviewer, contributing author, lead author or coordinating lead author in all four assessments. There are other contributors and reviewers in the statistical community. Among them are ISI members David Brillinger (USA/Canada), Peter Guttorp (USA/Sweden), and Dennis Trewin (Australia). In addition, Knut Alfsen (Norway), Mark Berliner (USA), Peter Bloomfield (USA), K. Brekke (Norway), Torstein Bye (Norway), Riccardo De Lauretis (Italy), Reinhard Furrer (USA), Tom-Regel Heggedal (Norway), Jim Hughes (USA), Richard Katz (USA), Robert Kaufmann (USA), Riitta Pipatti (Finland), Daniela Romano (Italy), Knut Einar Rosendahl (Norway), Andrew Solow (USA), Claudia Tebaldi (USA) and Angelica Tudini (Italy) have participated in IPCC assessment.
At a recent workshop organized by the American Statistical Association in Boulder, Colorado, climatologists and statisticians discussed how statisticians can contribute to climate change research. There are some extremely intriguing, and highly statistical, issues, such as:
How can one assess whether a given weather event is due to climate change?
How does one estimate historical sea surface temperatures from sparse data mainly obtained from ships travelling only in shipping lanes?
How does one combine different kinds of measurements of weather phenomena, when instruments change, locations are moved, or a rural site becomes highly urbanized with time?
The issue of climate change is real. There are huge data sets available, and some intriguing methodological problems. The answers to these problems will seriously affect the lives of our children and grandchildren. Let's get more involved!
Professor Peter Guttorp
The ISI regrets to announce the death of our colleagues:
Born Elected Deceased Professor Em. Fred Frishman 1923 1979 8 September 2007 Professor Anders H. Hald 1913 1951 11 November 2007 Professor David G. Kendall 1918 1955 23 October 2007
Professor Anders Hald (1913-2007)
Professor Anders Hald was a central figure in establishing mathematical statistics in Denmark through his teaching, textbooks and organization. He was instrumental in introducing and implementing Fisherian statistics into Denmark. He took part in the structural development of European collaboration in statistics in the 1960’s, thereby contributing significantly to the prehistory of the Bernoulli Society. After retirement, he contributed an impressive series of careful readings of classical historical texts in statistics.
Anders Hald was born 1913 as son of a schoolteacher in a village in Jutland, Denmark. He trained in actuarial science and statistics, and graduated from the University of Copenhagen in 1939. Already as a student, he had won the university gold medal for his prize paper ‘Mathematical exposition of R.A. Fisher’s theories’. In 1948, he obtained the degree of DPhil with a dissertation on identifying a trend, seasonal variation, and noise in a time series.
After his graduation in 1939, Hald had various consulting jobs, the most important for his scientific development being at the State Serum Institute in Copenhagen, where he was an assistant to G. Rasch (1901-80), who had spent the year 1935-36 with Fisher in London. During the 1940’s, he developed his detailed and well-documented textbook on statistical methods with many carefully explained concrete examples from his own experience (and re-used by later textbook authors with less practical experience): Danish version 1948, English version Statistical Theory with Engineering Applications, 1952, Russian edition 1956. In 1948, he succeeded H.C. Nybølle as Professor of Statistics in the Social Sciences Faculty. As so many other young European post-war scientists, he soon benefited from an extended visit to the USA (Chicago, Berkeley, New York), in his case sponsored by the Rockefeller Foundation. Hald was an ISI member since 1951 and soon became Fellow of the Institute of Mathematical Statistics (IMS) and of the American Statistical Association.
The Natural Sciences Faculty at the University of Copenhagen had no statistics programme or statistics department until Hald, in 1960, accepted an invitation to assume a new chair in mathematical statistics there. Hald spent the next years building up the special Danish brand of statistics based on a strong mathematical background, but always with concrete statistical consulting as an ultimate purpose of the programme. His detailed experience in university negotiation and the political trends of the 1960’s allowed the establishment of a department of mathematical statistics of 12 permanent posts at its height – then the recession of the late 1970’s turned the tide. Hald retired as Department Chair around 1971 and from his professorship in 1982.
Hald’s research covered three quite separate themes. His gold medal paper and a couple of smaller contributions with Rasch aimed at further mathematical development of Fisher’s theories. Of lasting importance is Hald’s detailed development of Fisher’s approach in his textbook and as basis for the programme in mathematical statistics in Copenhagen.
Hald had developed an interest for sampling plans in industrial quality control already in the 1940’s, and in 1952-53 he taught Statistical Quality Control at a UN supported programme at several universities in India. His many obligations in teaching and administration in the 1950’s left little time for research, but in 1960 Hald published an important paper in Technometrics on the mixed binomial distribution and Bayesian sampling plans for alternative variation. Hald decided to develop the area of attribute sampling plans in considerable detail in a long series of journal articles over the next 20 years, culminating with a monograph in 1981.
Starting from his contribution to the History of the University of Copenhagen at the occasion of its 500-year anniversary in 1979, Hald (now in his late 60’s) then embarked on a series of studies of the history of statistics that were to literally occupy the rest of his life. These were published in three monographs (1990, 1998, 2007), and in journal articles, some of which were reprinted in the exposition of Thiele’s major statistical works published by S. Lauritzen (2002). The emphasis is on scrupulous studies of the texts and detailed explanations of these in modern mathematical and verbal language.
Hald put prime priority on establishing and maintaining a scientific network in statistics with international colleagues, and followed up on his own early travels by systematically inviting researchers for shorter or longer visits and encouraging and facilitating younger colleagues to travel. It was only natural for Hald to engage himself in the development of European collaboration in statistics. Hald organized the second European meeting in statistics in Copenhagen in 1963 and chaired the European Regional Committee for statistics 1961-64.
At that time, this was a committee under the IMS, which in 1960 had picked up the increasing need for a formal structure for the European activities in (mathematical) statistics. As chronicled in detail by J.F. Teugels in Bernoulli News 2000-01, the committee later separated from the IMS to become one of the building blocks of the Bernoulli Society, a Section of the ISI.
Hald received several honours, including honorary memberships of the Royal Statistical Society and the Danish Society of Theoretical Statistics. In 1983, the Danish Technical University bestowed upon him an honorary doctorate.
Professor Anders Hald died peacefully on November 11, 2007.
Survey of ISI Members Regarding the Lisboa Session
The ISI Executive Committee is keen to explore ways in which the ISI Sessions can be made more useful and enjoyable for ISI and Section members. To this end, we decided to conduct a simple electronic survey on the views of the participants of the Lisboa Session. We ran this alongside an even briefer survey for ISI and Section members who did not participate in Lisboa on the reasons why they were unable or chose not to participate.
About 27% of the Session participants (603 of 2,109 participants, coming from 89 countries) and 24% of the ISI/Section members (795 of 3339 questionnaires issued), who did not attend the Lisboa Session, responded to our questionnaire.
I wish to express my sincere thanks to Professor Gilbert Saporta and his student Dr. Emmanuel Jakobowicz, who designed and analysed this survey, and to our staff at the ISI Permanent Office, who administered the survey. We intend to repeat the survey for future Sessions and are grateful to those of you who have suggested ways in which it can be improved.
I have received a full report on the survey though only the headlines are reported below. We will place a more complete report on the ISI website shortly. I personally have also read the large number of verbatim responses, which will be extremely valuable to us in planning the Sessions in South Africa, Ireland and thereafter.
Broadly, the survey of participants showed overwhelming support for the choice of location with approval ratings of over 95% (for Portugal in general) and over 90% (for Lisboa more specifically). Overall, more than 80% of participants were satisfied by the 56th ISI Session. More than 45% have a clear intention to participate in the next Session in South Africa and more than 88% were prepared to recommend to their colleagues that they do so. Participants were generally happy with the scientific programme with 70% expressing their satisfaction, and within this the invited sessions receiving the most praise. However, concerns were expressed about the location of the poster sessions. Many participants experienced some problems with the website and with the online payment system, and half of all respondents expressed their view that the registration and social programme fees were too high. Many of the verbatim responses expressed concerns about the high costs for registered accompanying persons.
The survey of non-participants also reflected concerns about the costs with more than 40% saying that problems with funding was the main cause of their lack of participation. The next most frequent response related to problems with the dates of the Session (this was mentioned by 29% of the respondents). In many countries, university classes have already begun by late August, the time of the ISI Session in Lisboa. Professional activities and personal reasons were the third and fourth reasons for non-participation. It is interesting to note that 38% of non-participants who responded to the survey have never attended an ISI Session, but also that of the 11% who had been to the previous Session in Sydney there were very high levels of satisfaction with that meeting.
May I reiterate my warm thanks to those who planned and worked hard on the survey, but also to those of you who committed your time to completing the questionnaires.
Awards, Prizes and Competitions
Jan Tinbergen Awards: Competition for Young Statisticians
from Developing Countries 2009
The International Statistical Institute announces the fourteenth Competition among young statisticians from developing countries who are invited to submit a paper on any topic within the broad field of statistics, for possible presentation at the 57th Session of the ISI to be held in Durban, South Africa, in August 2009.
Participation in the Competition is open to nationals of developing countries who are living in a developing country, and who were born in 1977 or later (see https://www.isi-web.org/404?tinbergen/2009papers.htm). Developing countries will be defined as countries with an annual GDP per capita of less than US$ 4,000 (U.N. 2006); see the list at https://www.isi-web.org/404?developing.htm.
Previous winners of the Award are prohibited to compete again. Papers submitted must be unpublished original works, which may include material from participants’ university theses. The papers submitted will be examined by an International Jury of distinguished statisticians, who will select the three best papers presented in the Competition. Their decision will be final.
Each author of a winning paper will receive the Jan Tinbergen Award in the amount of € 2,269 and will be invited to present his/her paper at the Durban Session of the ISI, with all expenses paid (i.e. round trip economy airline ticket from his/her place of residence to Durban, plus a lump sum to cover living expenses).
Manuscripts for the Competition should be submitted in time to reach the ISI no later than January 1, 2009.
The rules governing the preparation of papers, application forms and full details are available on request from the ISI Permanent Office. The address is as follows:
The Director of the Permanent Office
International Statistical Institute
428 Prinses Beatrixlaan
2273 XZ Voorburg, The Netherlands
Fax: +31 70 386 0025; E-mail: @cbs.nl
The Fourth ISI Mahalanobis Prize 2009
The Indian Ministry of Statistics and Programme Implementation, representing the Indian Government, is continuing its biennial initiative to award the P.C. Mahalanobis Prize in memory of this eminent Indian statistician. The Prize is to be awarded to a statistician who comes from a developing country and has worked there in recognition of his/her lifetime achievement in statistics and the promotion of best statistical practices. This initiative will serve the double purpose of keeping the memory of P.C. Mahalanobis alive and of recognising and stimulating progress in statistics in developing countries. The previous winners of the Prize are Prof. C.R. Rao, Prof. Benjamin Kiregyera and Dr. Isidoro David.
At the request of the Minister of Statistics and Programme Implementation for the Government of India, the International Statistical Institute has again established a Committee to propose a candidate for this Prize, according to the Memorandum of Understanding. The ISI administers a Fund of € 52,000 provided by the Indian Government, of which the interest will be used to award the Prize, which will consist of an economy class return airplane ticket to the ISI Durban Session, a per diem payment for accommodation and other living expenses while in Durban, and a Euro currency equivalent prize of US$ 5,000. The prize will be awarded at the ISI Session in Durban (August 16-22, 2009).
The Mahalanobis Committee Jury invites you to propose a candidate to the Committee, with arguments supporting the proposed candidate. Proposals can be sent to Ms. Shabani Mehta at the ISI Permanent Office (@cbs.nl) before August 1, 2008.
Cochran-Hansen Prize 2009: Competition for Young Survey Statisticians
from Developing and Transitional Countries
In celebration of its 25th anniversary, the International Association of Survey Statisticians established the Cochran-Hansen Prize to be awarded every two years to the best paper on survey research methods submitted by a young statistician from a developing or transitional country.
Participation in the competition for the Prize is open to nationals of developing or transitional countries who are living in such countries and who were born in 1969 or later.
Papers submitted must be unpublished original works. They may include materials from the participant's university thesis. They should be in either English or French. Papers for consideration should be submitted to the IASS Secretariat at the address given below and to arrive by December 29, 2008. Each submission should be accompanied by a cover letter that gives the participant's year of birth, nationality, and country of residence. The cover letter must also indicate if the work submitted is the result of a PhD thesis and, in the case of joint papers, the prize candidate must state clearly what his/her contribution to the paper is.
The papers submitted will be examined by the Cochran-Hansen Prize Committee appointed by the IASS. The decision of the Committee is final.
The winner of the Prize will be invited to present his/her paper at the 57th Session of the International Statistical Institute to be held in Durban, South Africa, August 16-22, 2009, and the name of the winner will be announced at the ISI General Assembly in Durban.
The author of the winning paper will receive the Cochran-Hansen Prize in the form of books and journal subscriptions to the value of about € 500, and will have reasonable travel and living expenses paid in order to present the paper at the ISI Session in Durban.
For further information, please contact:
Madame Claude Olivier
International Association of Survey Statisticians
CEFIL-INSEE, 3 rue de la Cité, 33500 Libourne, France
Tel : +33 5 57 55 56 17, Fax : +33 5 57 55 56 20
E-mail : Claude.firstname.lastname@example.org
2008 Advertising Information
The ISI Newsletter is published three times a year: February, June and October.
Advertisements in the ISI Newsletter may cover a half or full page.
Potential advertisers should note the following:
Advertisement Rates (per issue) & Sizes:
Per issue rate Length (cm) Width (cm) 1/2 page € 350 13 18 1 page € 550 26.5 18 inside back page € 650 26.5 18 back cover € 750 26.5 18
Honorary Member Interviews:
Professor George Box in conversation with Shabani Mehta
Professor George Box
There are many variations of the following aphorism that has been attributed to you:
Essentially, all models are wrong, but some are useful.
Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful.
All models are false but some models are useful.
Do these still correctly represent your intentions and would you still agree with any of these maxims?
What I said was ‘all models are wrong, some models are useful’.
You originally trained as a chemist and spent time during World War II working for the British Army on biochemical experiments on the effects of poisonous gases on small animals. Do you remember your time conducting these experiments and how did you feel about that type of work?
About my war service, I had been studying chemistry when WWII started in 1939, so I joined the army and went through the usual training. Sometime later, someone at the defence department must have been looking at my file. I found myself transferred to a secret experiment station in the south of England. They were bombing London every night. It was feared that some night it might be gas and not explosives, and the question was what you would do then. There were, I suppose, some of the best brains there and I wasn’t one of them. I worked for Prof. H. Collumbine, a professor of physiology, dressed up as a colonel. I was his lab assistant dressed up as a staff sergeant.
I understand that you only turned to the field of statistics when you were not able to find a statistician to advise you on how to analyze the results of the aforementioned experiments. In how much is this correct? As a result of this, you taught yourself statistics from the available textbooks. How do you feel this experience impacted your life?
I talked to Prof. Cullumbine about the extremely variable results I was getting from my biochemical determinations. I said what we really need is a statistician to look at these results. He said, ‘I know, we can’t get one. What do you know about it?’ I replied that I didn’t know anything about it, but I tried to read a book on it by someone called R.A. Fisher once and I didn’t get very far with it. Prof. Cullumbine said, ‘Well, you read the book, so you better do it’ and from then on I did it for three or four years. In the course of that time, there was nobody else around who knew anything about it so I spent a lot of time studying statistics during the evening. I designed and analyzed dozens of experiments. I remember one day I must have looked perplexed at some problem. A senior scientist suggested that I should go and see Fisher about it. He might just as well have told me that I should go and see God. In the end, I wrote to Fisher and he said to come and see him. I was attached to an artillery unit for administrative purposes. They couldn’t figure out how to send a sergeant to see a professor at Cambridge. In the end, they made up a story that I was delivering a horse. Fisher was very gracious and kind to me. He worked with me all day and we solved the problem.
Near the end of the War, it was discovered that the Germans had new chemical warfare agents (which for some reason were never used). These were nerve gases Tabun and so on. They were orders of magnitude more lethal than anything we knew about. It was decided that it was too dangerous to bring them to England. So I was part of a large team that went to the German research station to study these new gasses. When the War was over, I knew that I did not want to go back to chemistry but to study statistics.
After the War, you studied and received your bachelor’s degree in mathematics and statistics from University College London (UCL), and then obtained your PhD from the University of London in 1953. How was the field of statistics perceived in those days? Also, what was it like studying statistics at that time in the UK?
I never had any money and came from a very poor family. However, because of my service in the Army, they paid my expenses. When I went to UCL for an interview with Egon Pearson, I spent the whole time saying how wonderful R.A. Fisher was. He was very kind to me and said to come and study, but that there are one or two other people in statistics besides Fisher. I eventually I got a first class honours degree in statistics and also a PhD. The kind of statistics that I was doing concerned the design of experiments, which actually started in England with R.A. Fisher and his associates. The work I had been doing at the experiment station used these ideas and I found it was a field that a number of people were interested in. The Royal Statistical Society (RSS) used to have four research meetings a year and they would print the paper that was to be discussed in advance and send copies to anyone who wanted to participate. The majority of the time was given to discussion. If you look back at the journal of the RSS in those days, you’ll see discussions were extremely interesting and lively. You had Fisher, Frank Yates, George Barnard, Henry Daniels, Pearson, and a lot of other people, even including me.
Although you studied and received your degrees in England and worked for Imperial Chemical Industries (ICI); at some point in your career you moved to the United States, when exactly was this and what was your reason for moving across the Atlantic? You spent time in the United States at different universities, such as the University of North Carolina at Chapel Hill and Princeton University, before you settled in 1960 at the University of Wisconsin-Madison. It is at this University that you have spent the most time researching and contributing to the University itself. You participated in creating the Department of Statistics, and along with Bill Hunter co-founded the Center for Quality and Product Improvement (CQPI) in 1984. What led you to Wisconsin and to co-founding the Center at the University? What were your experiences as the director of such a department?
I got a job with Imperial Chemical Industries, a very large chemical company. I was a statistician at the Dyestuffs division in Manchester. There was a thing called the statistics panel, which involved statisticians from several different divisions of the company. The panel got together periodically at some central place to discuss statistics. One of the things we did was write a book for chemists and engineers called Statistical Methods in Research and Production. Later, there was a second book The Design and Analysis of Industrial Experiments. Both were edited by O.L. Davies and written by a number of people from different divisions.
There was a lot of chemical and engineering research going on and I found that I could be useful to these people by designing experiments for them. Before very long, I got involved in working on the problem of obtaining optimal conditions by improving yields and so on. That was when what is now called responsive surface methodology was invented. I worked with a chemist called K.B. Wilson and we produced a paper in 1951 that was read to the RSS. J.S. (Stu) Hunter was a student at that time at Raleigh, North Carolina. He read our paper and he got very excited about it. There was a lady called Ms. Gertrude Cox who ran two or three departments in North Carolina. Stu Hunter went to her with our paper and said, ‘You’ve got to get George Box over here’. They got a research grant from the Army Research Office. The next thing, to my surprise, I received a letter inviting me to become a visiting research professor at Raleigh for a year. I had no intention of any sort of academic career. I thought of myself as a statistician working in industry and it was what I expected to do the rest of my life.
In 1953, I went to Raleigh and was told by Ms. Cox that I had only one duty, which was to help Stu Hunter get a PhD, which I did. Moving from industry to academia, I had to take my invitation letter from Ms. Cox to the ICI directors. ICI let me go but wanted me to return, so they sent me to the US on the Queen Mary. I had a wonderful time at North Carolina and at various other departments I visited in the US. After a year, I went back to England and then I started getting phone calls from John Tukey who was at Princeton and wanted me to come over to be the Director for the Statistical Techniques Research Group there – an offer I resisted for some time, but in the end I decided to accept. The Group had a number of permanent members and visitors. For three years it went very well, I greatly enjoyed it. The only problem was that I wanted to do research with graduate students. That was not possible at Princeton.
Around 1960, I got an offer to come to Madison. When I arrived, they asked me to give two seminars; one on ‘What a statistics department should be like’, and the other one on a technical subject, the exact topic I’ve forgotten. I wanted the department to have a central group with offshoots to engineering, medical school, business school and so on. My idea was that these people would nourish each other and that worked to some extent. Later, Bill Hunter and I co-founded the Center for Quality and Product Improvement.
So from 1960-1966, I was Chairman of the Statistics Department. During that time, we were building our staff and getting some more senior people. Years later, when we started CQPI, Bill Hunter was the Director and I was Director of Research. One day Bill greatly surprised me. Previously, all our work had concerned engineering and science. Bill said, ‘I’ve just been to see the Mayor of Madison. I think this quality-thing should be used to improve the city and I went to talk to the Mayor about it and he got interested and gave me a project to work on’. It was to find out why it took such a long time to get police cars and other vehicles belonging to the city of Madison into the repair shop and out again. They were blaming the mechanics. However, it turned out that the main reason for the delay was that the repaired vehicles spent a long time sitting waiting to be picked up. These were very simple ideas but they were helpful. So the Center did get involved in a number of things of this sort.
You married Joan Fisher, the second of Sir Ronald A. Fisher’s five daughters. How did you meet your wife? Is it a coincidence that your father-in-law was a famed statistician or did you meet Mrs. Box through her father?
When I returned to the US for Princeton in 1956, I met Joan Fisher to whom I was married for a number of years.
Mrs. Box is also a published author; she wrote a biography on her father entitled ‘R.A. Fisher: The Life of a Scientist’. Has she influenced your statistical thinking?
I don’t think Joan has influenced my statistical thinking, but her father has to a considerable extent.
Do you have children and have any of them followed you into the field of statistics?
I have children; my daughter is a physician and my son is a cameraman in Hollywood.
There is so much experience and advice you could share with the young statisticians of today. Is there any specific advice you would like to impart to them?
I don’t think you can be a decent statistician without knowing some science and without doing some science. I would say that if students are studying statistics, they should try to find some problems in science that they could be involved with at the same time they are studying statistics.
Is there anything in your life/career that you wish you had done differently? Is there any part of your life/career that you most cherish?
Here at Madison, I have had many joint projects with other departments such as engineering and economics. So Madison is place that you can do those kinds of things quite easily. I have been happy here.
There are several theories named after you and those you worked with, including the Box-Jenkins Models, the Box-Cox Transformations and the Box-Behnken Designs; is there any one of these that you would single out as the one you are the most proud of creating? Which do you find the most useful to present-day statisticians?
I like them all but it is Gwilym Jenkins, my co-author on Box-Jenkins, I think of most warmly, it was strange the way we got involved in time series analysis. It began with building a self-optimizing reactor with the chemical engineering department. Jenkins came over and he knew a lot about time series and dynamic systems and helped enormously with the analysis. It was in the process of getting this optimiser to work that we got involved in all this stuff about non-stationary processes, and then it seemed gradually as things went on we found that we got involved in forecasting. These things are very closely related to process control. This project went on for several years. Gwilym was quite ill at the time with Hodgkin’s disease, so I used to go to England over the summer to work with him on this book. To answer which is most useful, certainly Box-Jenkins methods have been used very widely in economic modelling and business forecasting.
Though you have many research papers and books published, it is said that ‘Statistics for Experimenters’ is considered one of the most important books with regard to contributions made to the field of experimental design. Do you have any special recollections regarding this work that you would like to share?
The story about Statistics for Experimenters is that Stu Hunter, Bill Hunter and I used to teach short courses for industry, mostly for engineers and chemists. They would last about a week and we got some notes together. In the end, we got to know how to teach these people some useful things. This book did appeal to people who wanted to know how to design experiments. It approached things from their point of view. The first edition of the book sold 146,000 copies.
Many honours have been bestowed upon you, such as being appointed the Vilas Research Professor of Statistics (1980) and Professor Emeritus (1992) both at the University of Wisconsin-Madison, as well as Fellow of the Royal Statistical Society in 1979. You are also considered one of the most influential statisticians of the 20th century and a pioneer in areas of quality control, time series analysis, design of experiments and Bayesian Inference. Is there any particular area that you preferred working in? Which accomplishment are you most proud of?
I became a Fellow of the Royal Society some years ago and that is a great honour. The Royal Society was formed over three hundred years ago and when you become Fellow of the Royal Society you sign a big book; there are signatures in it of Isaac Newton and Charles Darwin, for example. Also, I did receive the Gold Medal of the Royal Statistical Society and honorary membership of the American Society of Quality Control.
It always seems to me that one thing in one area inspires work in another, so I would not go off and shut myself away and say I’m just going to do time series analysis or Bayesian analysis or whatever. Very often a lot of light is shed on problem A by looking at what’s going on in problem B. From the inception of the Department of Statistics in 1960 till I retired in the 1990’s, there was something called the ‘Monday-night beer session’. It was in my home and was a voluntary group where people from different departments got together. What we did was to get someone who had a problem in some area of statistics. We had talent scouts going around finding out who should be invited to come and present their problems and we would discuss it with them and with each other. A lot of people that I’ve seen in the last few years have said the thing they most remember about Madison are those Monday nights.
What advice do you have for the future of the ISI and in what direction would you like to see the ISI go?
One of my pet things that I go on about is the connection between theory and practice – that is that one inspires the other. If the ISI or any organisation sets itself up to make that interaction easy, then they will flourish and come up with all sorts of novel ideas.
You will turn ninety years of age on October 18th, 2009. To what do you attribute your longevity and your high level of activity?
I was really rather surprised when I started getting old, I didn’t expect to get this old. The only clue that I have comes from my cardiologist who gave me two years to live ten years ago. When I didn’t die ‘on time’ he said, “I know what it is, it’s his attitude!’ I really believe that you need to ‘accent the positive and eliminate the negative, and don’t mess with the Mr. Between’.
ISI Committee Matters: Khawarezmi Committee
A meeting of the Khawarezmi Committee of the International Statistical Institute was held on Tuesday, 13 November 2007, in conjunction with the first Arab Statistical Conference sessions that took place in Amman, Jordan, from 12-13 November 2007. The Khawarezmi Committee is chaired by Professor Dr. Hilal Aboud Al-Bayati and the meeting took place in the presence of the ISI President, Professor Denise Lievesley, who was invited to Chair the meeting.
To start the meeting off, the Khawarezmi Committee Chair provided some general information about the ISI, its Sections and Committees, as a number of the participants were not ISI members. Dr. Al-Bayati described efforts to create the Arabic translations for the Glossary of Statistics Terms, which was translated by the Arab Institute for Training and Research in Statistics (AITRS) and is now a component of the ISI website. Then, the ISI President responded to a large number of questions posed by the attendees.
The objectives of the Committee were subsequently explained, providing insight into what can be achieved by Arab statisticians, both inside and outside the Arab region. A number of action points have been agreed upon, in particular:
The opportunity to encourage young Arab statisticians to join the ISI family and to participate in international conferences organised by the ISI or meetings and conferences organised by the ISI Sections.
The identification of potential sources of financial support to cover the costs of Arab statisticians (particularly young statisticians) intending to participate in the 57th ISI Session, which will be held in Durban, South Africa in August 2009, and the 58th ISI Session, which will be held in Ireland in 2011.
The possibility that the Khawarezmi Committee, as well as the Arab Institute for Training and Research in Statistics, may help to assess visa applications from persons intending to attend statistical conferences held in developed countries in order to determine if they are legitimate.
It is possible that the Khawarezmi Committee may act to coordinate relations between the Union of Arabs Statisticians and the ISI to ensure the participation of Arab statisticians in international activities and foreign ISI members in regional activities.
Prof. Dr. Hilal Aboud Al-Bayati
Khawarezmi Committee Chair
ISI Membership Elections 2007
We would like to congratulate the one Honorary member and the one hundred elected ISI members, who were elected in the first and the second round of elections in 2007. 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 https://www.isi-web.org/404?isimembers/isimembers.htm), and these new members will be added to this list in the coming weeks.
First Round, 2007
Box, George E.P. - United Kingdom
Al-Omari, Amer Ibrahim F. - Jordan
Baker, Stuart G. - United States
Ben-Gal, Irad - Israel
Betensky, Rebecca - United States
Bisogno, Enrico - Switzerland
Bühlmann, Peter - Switzerland
Camdeviren Ankarali, Handan - Turkey
Congdon, Peter Douglas - United Kingdom
D'Ambra, Luigi - Italy
Datta, Susmita - United States
Del Moral, Pierre - France
Eilers, Paul H.C. - Netherlands
El Melhaoui, Saïd - Morocco
Elkum, Naser - Saudi Arabia
Farrell, Patrick John - Canada
Fischer, Matthias - Germany
Fraga Alves, Maria Isabel - Portugal
Fried, Roland H. - Germany
Gamrot, Wojciech - Poland
Gladun, Oleksandr - Ukraine
Grover, Lovleen Kumar - India
Hamasaki, Toshimitsu - Japan
Haughton, Dominique - United States
Hilbe, Joseph M. - United States
Hu, Xiaoquiong Joan - Canada
Hussain, Shakir M. - United Kingdom
Kafadar, Karen - United States
Karlis, Dimitris - Greece
Kelmansky, Diana Mabel - Argentina
Kenett, Ron S. - Israel
Kondylis, Athanassios - Switzerland
Konno, Yoshihiko - Japan
Kursa, Liliana Barbara - Poland
Lebreton, Jean-Dominique - France
Lee, Seung-Chun - South Korea
Linde, Peter - Denmark
Liu, Junfeng - United States
Looney, Stephen W. - United States
Ma, Shuangge - United States
Malgarini, Marco - Italy
Marchette, David - United States
Martinez, Wendy - United States
Martínez, Elena Julia - Argentina
Mendes, Mehmet - Turkey
Mishura, Yuliya Stepanivna - Ukraine
Müller, Marlene - Germany
Núñez-Anton, Vicente A. - Spain
Nyongesa, L. Kennedy - Kenya
Okrasa, Wlodzimierz - Poland
Pan, Jianxin - United Kingdom
Priebe, Carey E. - United States
Reiter, Jerome P. - United States
Said, Yasmin H. - United States
Sankaran, Paduthol G. - India
Smith, Rognvald Innes - United Kingdom
Sun, Jianguo (Tony) - United States
Sweetland, Mary - United Kingdom
Symanzik, Juergen - United States
Thomas, Neal - United States
Wang, Yan - Australia
Wasserstein, Ronald L. - United States
Westlake, Andrew J. - United Kingdom
Wong, Weng Kee - United States
Xue, Lan - United States
Yan, Ting - United States
Yeh, Arthur B. - United States
Zadlo, Tomasz - Poland
Ziegler, Andreas - Germany
Second Round, 2007
Alqallaf, Fatemah Ali - Kuwait
Bécue-Bertaut, Monica Maria - Spain
Bini, Matilda - Italy
Brito, Isabel - Portugal
Das, Kalyan - India
Epler, Margit - Austria
Fruewirth-Schnatter, Sylvia - Austria
Gong, Xiaoning - China
Gopalakrishnan, Asha - India
Heckman, James J. - United States
Kanichukattu, Jose Korakutty - India
Lee, Yonghee - South Korea
Lehohla, Pali J. - South Africa
Li, Hongzhe - United States
Lin, Xiaodong - China
Lo, Gane Samb - Senegal
López-Fidalgo, Jesús - Spain
Markatou, Marianthi - United States
Nakanishi, Hiroko - Japan
Narita, Masahiro - Japan
Nsowah-Nuamah, Nicholas N.N. - Ghana
Panaretos, Victor Michael - Greece
Papadopoulos, Savas - Greece
Rabe-Hesketh, Sophia - Germany
Sherman, Michael - United States
Vannucci, Marina - Italy
Vishwakarma, Gajendra Kumar - India
Walugembe, Fredrick Edward - Uganda
Wang, Huiwen - China
Wang, Li - China
Wu, Colin - United States
Yansaneh, Ibrahim Sorie - Sierra Leone
Historical Anniversaries: Max Planck (1858-1947)
Born in Kiel, capital of Schleswig-Holstein (Germany), on April 23, 1858, Max Karl Ernst Ludwig Planck played an important part in the physics of the 20th century. His father was Professor of Constitutional Law at the University of Kiel.
A short time after the Danish-Prussian war in 1864, Planck moved with his family to Munich where he became a good mathematician and learned about the law of conservation of energy, coming through this door into contact with the field of physics. He graduated early at the age of 16.
The meeting with music also played an important part for him, having a predilection for singing, piano, organ and cello. Between music and physics, he chose the latter one to enter into a professional life.
After three years at the University of Munich, Planck went to Berlin in 1877 for a year of study, in particular with Gustav Kirchhoff, Hermann von Helmholtz and Karl Weierstrass. During this time, he undertook a programme of study of Clausius' writings that justify the use of kinetic theory of gases through probabilistic arguments. Interested in thermodynamics, he defended his dissertation in 1879: Über den zweiten Hauptsatz der mechanischen Wärmetheorie (On the second fundamental theorem of the mechanical theory of heat). In June 1880, he presented his habilitation thesis entitled Gleichgewichtszustände isotroper Körper in verschiedenen Temperaturen (Equilibrium states of isotropic bodies at different temperatures).
Appointed in 1885 by the University of Kiel as Associate Professor of Theoretical Physics, he became Full Professor at the University of Berlin, as successor to Kirchhoff, in 1892. He retired from Berlin in 1926, and was succeeded by Erwin Schrödinger.
At the end of the 19th century, Planck was interested in the radiation processes and, more particularly, in the problem of black-body radiation.
In 1900, he proposed that the energy emitted for a resonator of frequency ν is equal to h ν, where h — chosen as the first letter of Hilfsgröße (auxiliary quantity) — is an universal constant, called since that time Planck's constant. The central assumption behind his new theory was the supposition that the electromagnetic energy can be emitted in quantized form, and is a multiple of an elementary unit. This assumption, incompatible with classical physics, is considered as the birth of quantum physics.
Evocating the discovery of Planck's relation must interest physicists. However, why would it also be of interest to statisticians? In fact, to understand the importance of Planck's discovery, we must link this event to the development of statistical physics at the end of the 19th century.
In his paper published in 1857 on the kinetic theory of gases, Rudolf Clausius, who is already mentioned above, used probability theory to explain the transition from a seemingly chaotic movement of the individual microscopic particles of a gas to an ordered movement of sets of particles.
Clausius' work would permit James Clerck Maxwell and Ludwig Boltzmann to develop statistical mechanics. Maxwell introduced explicitly a velocity distribution for the molecules of a gas. Boltzmann went on to develop this theory and to efficiently support this new theory.
Max Planck belongs originally to Boltzmann's critics. He changed his opinion in 1900 and adopted Boltzmann's ideas that, already in 1877, discuss the possibility that energy states of a physical system could be discrete. The relation introduced by Planck would be developed and interpreted by Albert Einstein in 1905.
A distribution with density function
is currently called Planck's radiation formula and belongs to the class of Planck distributions.
The anniversary of Planck's birth gives us the opportunity to recall the importance of probabilistic and statistical tools in the essential developments of physics at the end of the 19th century and the beginning of the 20th century. This aspect of statistics is worth recalling.
Results obtained by Planck resulted in him being awarded the Nobel Prize in Physics in 1918.
When the Nazis seized power in 1933, Planck was 74 years of age. He chose to remain in Germany and was openly opposed to some of Hitler's policies. At the end of the Second World War, his home was destroyed by bombing and one of his children, Erwin, was executed by the Gestapo for his part in the unsuccessful attempt to kill Hitler in 1944. These misfortunes added further to previous painful events given that another son, Karl, had already been killed in France in 1916, and two daughters, Margarete and Emma, died, respectively in childbirth in 1916 and 1919.
By the end of the War, Planck remained with his second wife and a last son by her. He died in Göttingen on October 4th, 1947.
Christiaan Huygens Committee on the History of Statistics
Memories of the ISI's Past
Participants in front of "Le Pré Catelan" 12th ISI Session, Paris 1909
News from ISI Sections: Bernoulli Society IAOS IASC IASE IASS IFC ISBIS
Back to Home Page
News from ISI Sections Volume 32, Number 1 (94) 2008
Other ISI Newsletters