Newsletter 64 Index

Australian &
New Zealand Journal of Statistics

Newsletter Archive

Join the NZSA

Feedback to Editor

New Zealand Statistical Association Newsletter 64

September 2006

Local News

Department of Conservation

The statistical staff at DOC is back up to two full-timers. Carla Meurk has left our shores for redder pastures to undertake a PhD at the University of Queensland, but I am back on the force on a one-year contract based in Wellington. I submitted my MSc thesis to the University of Auckland in July, the subject of which was a multivariate evaluation of the Marine Environment Classification for coastal reef fishes, supervised by Marti Anderson. My new project with the department is to assess DOC's needs for methods of longitudinal data analysis, and to boost our capacity in this area. I invite discussion and collaboration from anyone with expertise in this branch of statistics. Meanwhile, Ian Westbrooke has spent a couple of months living in Auckland, where he attended the 2006 ASC/NZSA conference, and returns to Christchurch soon. He recently received quite a surprise when he opened the Dominion Post one morning to find some equations he had written some years ago featured on page two, in the illustration for a news article on the Maori Electoral Option.

Adam Smith

University of Otago

Professor Mark Meerschaert, Chair of Applied Mathematics, University of Otago has left the department to return to America to take up the position of Chairperson, Department of Statistics and Probability, Michigan State University.

John Harraway attended ICOTS 7 in Salvador Bahia, Brazil in July. The conference was very worthwhile with 550 registered but about 15 not able to attend because of the collapse of the Brazilian Airline, Varig. A further 100 high school teachers from Bahia attended workshops in the two days before the conference. John presented a paper on Item Response Theory, took part in an invited debate about the teaching of statistics in context and organised the session on multivariate statistics within the tertiary education topic. He was Scientific Secretary for ICOTS7 and has been asked to Chair the International Programme Committee for ICOTS 8 which will take place in Slovenia in 2010. John also visited the Federal University of Santa Catarina in Florianopolis where he is working with Dalton Andrade on a project in Brazil.

David Fletcher is enjoying all the delights of sabbatical this year, and is currently visiting Byron Morgan at the University of Kent in England. He has already been to CNRS in Montpellier, France, to work with Jean-Dominique Lebreton, and will visit there again before returning to Dunedin in September. He will be giving a seminar at Kent at the end of August entitled "Mark-Recapture Models and Population Dynamics". This will be broadcast via video-link to St. Andrews and Cambridge, as statisticians at these three universities run the National Centre for Statistical Ecology. David arrived in England just after the heatwave across Europe was coming to an end, so felt quite at home arriving to find summer consisted of rain and 12 degrees Celsius temperatures (much like Dunedin!). It was quite a change from the 35-degree climate he had earlier enjoyed in Montpellier.

Irene Goodwin

Crop & Food

Once again, Crop & Food is farewelling a biometrician - but this time we're hopeful she will come back. Esther Meenken is off to the UK for seven months, to do a Masters in Biometry at the University of Reading, with funding from QEII Technicians Award, Allan & Grace Kay Overseas Scholarship, and Crop & Food.

Duncan Hedderley

Fonterra

With fewer arrivals (Hayley Knox) than departures (Maree Luckman, Dongwen Luo and Hayley Knox) over the last year or so our current statistical staffing sits at three (or three and a half depending on how it's counted) - Rob Crawford and Roger Kissling (Hamilton) and Barbara Kuhn-Sherlock and Catherine Lloyd-West (Palmerston North). A recent restructuring has resulted in the formation of a Calibration and Statistics team, the first time statistics has received any structural recognition within Fonterra. Rob Crawford has taken over management of the combined group, which includes some 20 calibration staff, as well as continuing as a half-time statistician. Subject to negotiations, we are hopeful of recruiting at least one more statistician in the near future.

Roger Kissling

University Of Auckland

We have had some comings and goings recently. We have said a fond farewell to Mik Black, who has departed for his native South Island and is now a member of the Bioinformatics Group in the Biochemistry Department of the University of Otago. We are delighted to welcome Associate Professor Catherine Loader, who has joined our department from Case Western Reserve University in Ohio. Yuichi Hirose, who completed his PhD last year, has taken up a lectureship in Applied Statistics at Victoria University of Wellington. Carl Donovan has successfully defended his PhD thesis and continues as a lecturer at the University of St Andrews in Scotland. And a warm welcome to Jacob Miller: the newest family member for Arden and wife Mel, born on May 8th.

2006 has been the Year of the Workshop. Russell Millar spent a week in St John's, Newfoundland in May, participating in a workshop on Bayesian methods for stock assessment. Russell stopped off in Seattle on the way, enjoying temperatures as high as 25C. On the Atlantic coast he endured temperatures below freezing. Rachel Fewster noted similar temperature fluctuations in April while attending a workshop on Uncertainty in Ecological Analysis, organised by Noel Cressie at Ohio State University. Rachel also co-presented an Auckland workshop on distance sampling, with Steve Buckland and David Borchers of the University of St Andrews. Paul Murrell gave a workshop on R Graphics following the ASC/NZSA conference in Auckland in July, and the department also hosted two further workshops by Bill Venables (S Programming) and Robert Gentleman (Bioconductor).

Continuing the workshop theme, Marti Anderson gave a workshop on multivariate analysis for ecologists at the University of Western Australia in July. The workshop was the inauguration of commercial versions of Marti's multivariate software, which will soon be available as an add-on package to the existing PRIMER software (standing for Plymouth Routines in Multivariate Ecological Research), developed by a company based in Plymouth, UK. Marti's next workshop will be at the University of Pisa in Italy in September, followed by another at the National Centre for Ecological Analysis and Synthesis in Santa Barbara, California. The official release of the program is scheduled for early 2007.
PhD student James Russell, whose failure to catch an absconding rat shot him to fame (and a Nature publication) in October 2005, was selected to represent the university for the state visit of the Crown Princess of Thailand, Her Royal Highness Maha Chakri Sirindhorn, in March. James gave one of two talks to the Princess and diplomatic delegation (right), aimed to showcase research at the University of Auckland on the theme of conservation and environmental science. The star of the show, athletic Norway rat Razza who swam across 400m of ocean to escape from James's research attentions, declined to attend for personal reasons.

With more ratty tales, this time about a genetic study of rat invasions among New Zealand islands, PhD student Steven Miller won the NZSA student prize at the conference in Auckland in July. He then went to the International Biometric Conference in Montreal, also attended by Chris Wild and Chris Triggs. And finally, two Statistics grad students had a memorable stroke of luck in March. After parking in Parnell and walking to campus, they were aghast to see their own car driving past them in central Auckland. At times like this, what better phenomenon to rely upon than the trusty Auckland gridlock. The car was halted at a traffic light long enough for the owner to sprint the length of the street and jump into his car next to the thief. In his stoned state, the thief took a moment to catch on, but then his jaw dropped and he `sort of fell out the driver's door.' Having calculated the odds of such good fortune as being somewhat less than those of winning Lotto, the student Dougal Greer was then faced with the less welcome challenge of having to find another parking space in central Auckland.
Rachel Fewster

AgResearch

In April our group had its annual 3-day retreat/advance in Hamilton, together with guests Esther Meenken, Andrew McLachlan and Duncan Hedderley from Crop & Food. The first day was spent flying in during a break in the fog, having morning tea and lunch amid speculation as to what other planes had landed where, then a late start to our in-house presentations by those folk whose planes had landed. At 6.30pm the last, largest contingent landed, and “Little India” was visited for the first full gathering. Next day got off to a much better start, and was spent at Geoff McLachlan’s workshop on microarrays followed by dinner at The Station”. On the final day, the remaining presentations from the first day were given, with our team-building exercise, petanque, being sacrificed in the interests of science!

AgResearch Bioinformatics, Mathematics and Statistics Group, April 2006, at its annual gathering at Ruakura.

Re comings and goings. With the relocation of the AgResearch Wallaceville staff to the soon-to-be-created Hopkirk Institute (Palmerston North) and Invermay, Lilian Morrison, biometrician at Wallaceville, is finishing with us mid December 2006 after many years of service to AgResearch and MAF (Lilian is interested in any job offers in Wellington or Hutt Valley!). All the best, Lilian!

Re conferences, it all happened in July. Six of us attended the ASC/NZSA 2006 conference in Auckland. Martin Upsdell talked on “Comparison of methods for analyzing fluorescent spectral data”, Dave Saville on “Replication and randomisation – lost arts?” and Roger Littlejohn on “Analysis of bite force time series”. An even luckier trio (Peter Johnstone, Fred Potter and Roger Littlejohn) went further afield, to the International Biometric Conference in Montreal, Canada, where Fred spoke on “Segmented regression: some methods and case studies” and Roger on “A semi-markov model for biting time series”. Fred celebrated his 60th birthday by crossing the International Date Line at 10 pm, thus making the 'day' last for 42 hours, then having a swim in the Pacific Ocean at Manhattan Beach, Los Angeles (photo available, but censored by Fred!).

Dave Saville

University of Canterbury

Jennifer Brown presented two papers at ICOTS7 in Brazil in July. One was with Richard Penny and Marco Reale on the links between Statistics NZ and UC in developing the department's teaching programme. The other paper was on teaching environmental statistics to environmental managers.

The NZIMA programme on Modelling Invasive Species and Weed Impact is underway. A postdoctoral fellow has been appointed and scholarships have been offered to 2 PhD students. A 3rd PhD student is starting soon. The students will be jointly based at UC and Lincoln University, working on developing population models for weed spread and for designing monitoring and management strategies. More information is at http://www.math.canterbury.ac.nz/bio/NZIMA/.

Marco and Carl recently presented papers at the IWSM2006 in Galway, Ireland, They were kindly hosted by John Newell of the National University Ireland, who recently spent a few months on sabbatical with our department. The conference organisers tried to persuade us that the 28-30oC temperatures during the conference were entirely as expected, rather than a statistical outlier. After many adventures enroute, the poster presented by Marco (joint work with Jennifer and Bill Rea, their joint Masters student) won second prize in the poster competition.

The Merry Adventures of a Poster – A True Story. Marco volunteered for the taxing duty of flying to the other side of the world to give the presentation. However, on checking in at Christchurch International Airport he realized he’d left the poster in the carpark, so a friend was dispatched to find it. Then, at Changi Airport in Singapore, he lost the poster again…this time no luck (presumably taken away by the bomb disposal team). The poster was reprinted enroute in Naples, courtesy of the University of Naples Federico II. Unfortunately, when Marco left Rome for Ireland, he lost the poster at Fiumicino Airport! The poster was finally reprinted in Galway, the day before the presentation. At which point Marco left his memory stick with the printers, thankfully returned later that day, shame we can’t say the same about his main memory bank!

John Newell made such an impression on the dept that we are in the process of setting up an adjunct position for him, which is part of plans to foster greater links and cooperation between our respective departments.
Easaw and his PhD student, Issarest Weeraprajak, attended the International Conference on Time Series, Econometrics, Finance and Risk in Perth from 29 June to 1 July. They presented a paper entitled “New Learning Algorithm for Adaptive Network-based Inference System in Application of Forecasting Chaotic Time Series”.

As part of an on-going study with researchers based at Christchurch Women’s Hospital, Dominic will present a paper entitled “Developing hidden Markov models for aiding the assessment of preterm babies’ health” at the International Conference of Biomedical Pharmaceutical Engineering in Singapore from 11-14 December 2006.

Professor Christian Robert recently visited us for 6 weeks hosted by Dominic Lee. Christian is an internationally renowned expert in Bayesian and Computational Statistics from the Universite Paris Dauphine in France, and has made a substantial contribution to our Bayesian statistics course.

Carl Scarrott

Wellington Statistics Group

The Wellington Statistics Group (WSG), a local group of the NZSA, continues to meet regularly (although a little less frequently so far this year than we would have liked). The Group receives regular sponsorship from the Ministry of Social Development, Statistics New Zealand, Statistics Research Associates Ltd, and Victoria University of Wellington. There have been recent WSG talks given by:

• July 2006: Edith Hodgen, Rachel Dingle and Hilary Ferral, New Zealand Council for Educational Research, "Statistics: a growth area for NZCER"

• June 2006: Ian Westbrooke, Department of Conservation, Research Development and Improvement Division, Christchurch, "Meeting statistical needs in a conservation management organisation"

• March 2006: Geoff Chambers, Cell and Molecular Biosciences, VUW, "Out of Taiwan? Genetics sheds new light on Maori origins"

There are a couple of 'promised' talks in the pipeline, but with no confirmed dates yet I won't drop any names! Anyone who does not presently receive WSG announcements and who wishes to be informed of future events is welcome to contact the WSG Convenor, John Haywood: John.Haywood@mcs.vuw.ac.nz.

John Haywood

Victoria University

The most exciting news from the Statistics and Operations Research (STOR) Group at VUW is that we have a new staff member, who arrived (with perfect timing!) just before the deadline for this newsletter, at the end of August. Yuichi Hirose came to us from Auckland, where he completed his PhD under supervision from Alan Lee. His PhD research considered the efficiency of semiparametric maximum likelihood estimation in a variation of case-control sampling that was originally proposed by Alastair Scott and Chris Wild. Currently Yuichi is adapting this theory for GEE and Bayesian estimating function situations. Yuichi has filled the previously mentioned new position with a focus on 'Applied Statistics', that we created following our external review in 2005.

In other news, Dong Wang is overseas on sabbatical, from July 2006 to April 2007. Richard Arnold gave a talk at the 8th Valencia Bayesian statistics meeting in Spain in July, and Richard jointly ran a course in longitudinal data analysis for Statistics New Zealand in May, along with Ivy Liu. Ivy and Richard also presented work at the ASC/NZSA 2006 conference in Auckland in July, as did John Haywood and Estate Khmaladze. Victoria University sponsored Hira Koul's invited talk in the "Modern Goodness of Fit Methods" session that Estate organised, and in which Estate also spoke. Hira (Michigan State University) then spent some time working at VUW with Estate following the conference. At the end of June, John Haywood presented a talk at the Time Series Econometrics, Finance and Risk conference in Perth (UWA). John also gave an invited presentation to Statistics New Zealand at the end of August, and was an invited speaker at a workshop organised by the Reserve Bank of New Zealand in March.

Stefanka Chukova visited Japan and South Korea in August. In Japan Stefanka worked with Yu Hayakawa (in Tokyo), and gave an invited seminar on some joint work of hers with Srinivas Chakravarthy and Harry Perros. In Busan (Korea), Stefanka organized a special session on Advanced Warranty Modeling at the 2006 Asian International Workshop on Advanced Reliability Modeling (AIWARM 2006). Stefanka also visited Bulgaria in June, where she attended the 12th International Summer Conference on Probability and Statistics in Sozopol and gave a talk on Warranty Repair Strategies, which is joint work with Mark Johnston. Mark has been busy since joining us last year, and he's having an exciting time: in the same week in April he got engaged (to Emily) and bought a house. In addition to his work with Stefanka, Mark gave a talk on adding rewards in combinatorial optimization at the ANZAM Operations Management Symposium held at Victoria University in June.

We've had quite a few visitors to the group too. Bhramar Mukherjee (from University of Florida) visited Ivy Liu for a month (in May) and worked with Ivy and Dong Wang. Estate Khmaladze hosted Jon Wellner (University of Washington) for three weeks in August/September, in addition to Hira Koul for a fortnight in July. Daryl Daley did some work with David Vere-Jones here in Wellington after the ASC/NZSA conference, where Daryl received his (surprisingly heavy!) Pitman Medal. Christian Robert (Universite Paris Dauphine and Visiting Erskine Fellow at Canterbury) visited the group briefly in early August. Not surprisingly, Christian's talk drew a large and diverse audience, with several computer science representatives, along with those from stats/OR and maths. Harry Perros (North Carolina State University) visited Stefanka Chukova from February to April. Stefanka and Harry co-taught a graduate class on Computer Networks and Reliability, and they did some research on issues related to computer networks. Stefanka is also hosting Dimitar Christozov (American University in Bulgaria) from August to October. Stefanka and Dimitar are jointly researching warranty issues linked to malfunctioning and misinforming.

John Haywood

Massey University, Turitea

Mark Bebbington was peacefully minding his own business, giving a short presentation on the modelling of eruption intervals at Mt Taranaki, when he incautiously advanced the estimate of a 1/3 - 1/2 probability of an eruption within the next 50 years. Within the hour he was being interviewed on Radio New Zealand, and was tracked down by a TVNZ crew the following day. He says that the most interesting experience was doing Radio Live with Graeme Hill - having words like "pyroclastic flow" tossed at you by an interviewer at 7.20am requires more than the one cup of coffee beforehand. Jonathon Godfrey also made the news, appearing very briefly on TV1 news in March (already on the NZSA web site) and a Radio NZ item on Morning Report around the same time. Interestingly neither Jonathan nor Mark was credited in the media with being a statistician: Mark was designated as "vulcanologist" and Jonathan as "blind man". Clearly "statistician" isn't sexy enough.

Our new Professor, Martin Hazelton, was at the ASC/NZSA conference in Auckland to give an invited presentation on density deconvolution, and was pleased to able to meet a number of members of the NZ statistical community for the first time. He has recently been awarded, jointly with Nigel French, professor of veterinary epidemiology, a Massey University award to fund a postdoctoral research fellow for 2007 & 2008.

Quite a number of us made, one way or another, the long trek to Auckland for the ASC/NZSA conference in July. Ganesh hired a van for the weekend roadtrip along with Ganes, Greg and Alasdair. They met with some interesting weather conditions on the desert road (pictured below). Chin Diew and Geoff opted for the potentially quicker method of flying up on the Monday morning, but after circling the fog around Auckland airport a few times were diverted to Rotorua, to arrive by bus just in time for lunch.

Chin Diew Lai's new book "Stochastic Ageing and Dependence for Reliability" with M Xie of Singapore has just been published by Springer. Chin Diew delivered an invited talk at the ASC/NZSA conference in Auckland, on "Compliance inspection and conformity testing in the presence of measurement or inspection errors".

Doug Stirling attended ICOTS7 in Salvador, Brazil in early July. He has recently started developing experimental design material for CAST as part of a contract with Nestlé.

Geoff Jones spent two weeks in the UK and one in California in July, first attending MOLS 2006 at the University of Essex to learn something about longitudinal surveys, then meeting with veterinary epidemiologists at the University of Liverpool, and finally spending a week with Wes Johnson at the Centre for Animal Disease Monitoring of the University of California, Davis. We hosted Wes briefly at Massey in August, and he will be returning to New Zealand next year for the Conference in Honour of John Deely in Christchurch.

Ganes and Alasdair are organizing another one-day Stats Forum at Massey in conjunction with AgResearch Grasslands. The keynote speaker this year will be David Baird from AgResearch Lincoln.

Geoff Jones

Return to top