1 Published by the New Zealand Statistical Association (Inc.), P.O. Box 1731, Wellington, New Zealand. The views expressed by contributors to this Newsletter should not be attributed to the New Zealand Statistical Association. ISSN 0112-2649 The New Zealand Statistical Association nzsa.rsnz.org Newsletter March 2006 Number 63 SKYCITY, Auckland 3-7 July, 2006 You are warmly invited to attend the ASC/NZSA 2006 Conference at SKYCITY, Auckland. SKYCITY Auckland Convention Centre is located in downtown Auckland. It is the newest and most technically advanced convention centre in NZ. Key Dates Abstracts Submission Deadline: 22 March Authors notified of acceptance: 28 March Early Bird Registration Deadline: 25 April Registration (for New Zealanders) Category Before 25 April After 25 April Member $765 $935 Non Member $1050 $1220 Student $280 $330 Accommodation Accommodation at SKYCITY can be booked through the conference webpage. Cheaper alternatives will be given in the online newsletter. Social events The welcome reception is on Monday evening, cost included in registration, and the conference dinner is on the Tuesday night ($130), both at SKYCITY. On Wednesday afternoon there will be optional tours (Auckland city, hike in the Waitakeres, Waiheke Island, sailing on the Harbour), as well as contributed papers and Society meetings. Webpage: http://www .statsnz2006.com/ Email: [email protected]David Scott, Conference Chair From the Program Chairman The theme of this year’s joint conference of the Statistical Society of Australia and the New Zealand Statistical Association is “Statistical Connections”. This theme was chosen to indicate the strong links, not only between statisticians in both countries, but also more widely in the region. The theme has also had significant impact on the program design, in which statistical theory and methodology, computational methodology and applications are linked. The program also aims to increase the connection of young statisticians with each other and existing members of both the Australian and the New Zealand societies. Putting the program together has involved a lot of work by the members of the program committee: David Scott (U. Auckland who is also the Conference Chair), Chris Carter (CSIRO CMIS), Kerrie Mengersen (QUT), Marti Anderson (U. Auckland), Beatrix Jones (Massey U., Albany), and I am very grateful for their efforts. The Scientific Program for ASC/NZSA 2006 includes Plenary Sessions, Invited Paper Sessions, Contributed Paper Sessions and Poster Sessions. Satellite workshops will also be available. The main scientific program will take place over the four days Monday July 3 through Thursday July 6. There will be a plenary session of 75 minutes duration on each day of the conference. We are fortunate to have four outstanding keynote speakers for the plenary sessions. We hope to see you all in Auckland in July! William Dunsmuir Program Committee Chair
Transcript
1
Published by the New Zealand Statistical Association (Inc.), P.O. Box 1731, Wellington, New Zealand. The views expressed by contributors to thisNewsletter should not be attributed to the New Zealand Statistical Association.
ISSN 0112-2649
The New Zealand Statistical Associationnzsa.rsnz.org
Newsletter March 2006Number 63
SKYCITY, Auckland
3-7 July, 2006You are warmly invited to attend the ASC/NZSA 2006Conference at SKYCITY, Auckland.
SKYCITY Auckland Convention Centre islocated in downtown Auckland. It is the newest andmost technically advanced convention centre in NZ.
Key DatesAbstracts Submission Deadline: 22 MarchAuthors notified of acceptance: 28 MarchEarly Bird Registration Deadline: 25 April
Registration (for New Zealanders)Category Before 25 April After 25 AprilMember $765 $935Non Member $1050 $1220Student $280 $330
AccommodationAccommodation at SKYCITY can be booked throughthe conference webpage. Cheaper alternatives willbe given in the online newsletter.
Social eventsThe welcome reception is on Monday evening, costincluded in registration, and the conference dinner ison the Tuesday night ($130), both at SKYCITY.On Wednesday afternoon there will be optional tours(Auckland city, hike in the Waitakeres, WaihekeIsland, sailing on the Harbour), as well as contributedpapers and Society meetings.Webpage: http://www.statsnz2006.com/Email: [email protected]
David Scott, Conference Chair
From the Program Chairman
The theme of this year’s joint conference of theStatistical Society of Australia and the New ZealandStatistical Association is “Statistical Connections”.This theme was chosen to indicate the strong links,not only between statisticians in both countries, butalso more widely in the region. The theme has alsohad significant impact on the program design, in whichstatistical theory and methodology, computationalmethodology and applications are linked. The programalso aims to increase the connection of youngstatisticians with each other and existing members ofboth the Australian and the New Zealand societies.
Putting the program together has involved a lot ofwork by the members of the program committee:David Scott (U. Auckland who is also the ConferenceChair), Chris Carter (CSIRO CMIS), KerrieMengersen (QUT), Marti Anderson (U. Auckland),Beatrix Jones (Massey U., Albany), and I am verygrateful for their efforts.
The Scientific Program for ASC/NZSA 2006includes Plenary Sessions, Invited Paper Sessions,Contributed Paper Sessions and Poster Sessions.Satellite workshops will also be available. The mainscientific program will take place over the four daysMonday July 3 through Thursday July 6. There willbe a plenary session of 75 minutes duration on eachday of the conference. We are fortunate to have fouroutstanding keynote speakers for the plenary sessions.
We hope to see you all in Auckland in July!William Dunsmuir
Program Committee Chair
2
ASC/NZSA 2006 Offers an Outstanding Scientific Program
Keynote Speakers
There are four plenary sessions of 75 minutes duration, one on each day of the conference. The threekeynote speakers and the Foreman Lecturer will each present a 60 minute talk at one of the plenary sessions,with the remaining time being devoted to discussion or conference activities such as opening and closing.
Professor David DonohoDavid Donoho is Anne T. and Robert M. Bass Professor in the Humanities and Sciencesat Stanford University. He works in mathematical statistics, information theory andcomputational harmonic analysis. He is currently interested in new multiscalerepresentations lying “beyond wavelets”, and in rapidly finding the sparsest solution ofsystems of underdetermined linear equations. He received his AB from Princeton summacum laude in statistics and his PhD in statistics from Harvard. He has received aMacArthur Fellowship as well as the von Neumann prize of the Society of Industrialand Applied Mathematics, and has served as Wald Lecturer for the Institute ofMathematical Statistics. He is a member of the US National Academy of Sciences andthe American Academy of Arts and Sciences. He won the 1994 COPSS Award from the Committee ofPresidents of Statistical Societies.
Professor Peter HallPeter Hall is Professor of Statistics in the Centre for Mathematics and its Applicationsin the Mathematical Sciences Institute, Australian National University. Peter receivedhis BSc degree from the University of Sydney in 1974. His MSc and DPhil degrees arefrom the Australian National University and the University of Oxford, both in 1976. Hetaught at the University of Melbourne, then in 1978 took a position at the AustralianNational University, where he has been ever since. His research interests range acrossseveral topics in probability and statistics. He is a Fellow of the Australian Academy ofScience and a Fellow of the Royal Society of London and won the 1989 COPSSAward.
Professor Xiao-Li MengXiao-Li Meng is Professor and Chair of the Department of Statisticsat Harvard University. He received a MA in statistics from Harvardin 1987 and a doctorate in statistics in 1990. Professor Meng heldprevious appointments at the University of Chicago. His researchinterests are in statistical inference under complex settings, such aspartially observed data, pre-processed data, and simulated data,quantifying statistical information and efficiency in scientific studies,particularly for scientific computation, genetic studies, andenvironmental problems, statistical principles and foundational issues and effective deterministic and stochasticalgorithms for Bayesian and likelihood computation. Professor Meng won the 2001 COPSS Award.
Foreman Lecture: Professor Ray ChambersRay Chambers is Leverhulme Professor of Social Statistics and Director of theSouthampton Statistical Sciences Research Institute at the University of Southampton.He is a leading international expert in survey statistics and has close contact with severalnational statistical offices, including the UK Office for National Statistics and the ABS.He will be moving to Australia in early 2006 to take up a research chair in statisticalmethodology at the University of Wollongong.
See the online newsletter for links to the homepages of the Keynote speakers.
3
Invited Speakers
The program committee has been delighted with the way in which the invited speaker program has developedand with the willingness of many outstanding statisticians from around the world to participate in it. Our aimwas to develop a program in which statistical and computational methodology relevant to a range of applicationswas presented and this resulted in two of the main program themes being organised. Details are given onPage 10.
The invited program would not have come about without the efforts of the Session Organizers who haveplayed a key role in bringing the program together. The Program Committee is very grateful for their efforts.
Contributed Paper and Poster SessionsContributed Papers presented orally or as a poster have always been critical to the success of past conferences.Accordingly the Program Committee has planned for plenty of Contributed Papers and encourages as manyattendees as possible to make an oral or poster presentation. We anticipate that most contributed oralpresentations will fall into one of the theme areas (outlined on Page 10) and the Program Committee willmake every effort to organize the Contributed Papers to complement and enhance the invited and plenarysessions. Contributed Paper Sessions are of 105 minutes duration allowing for 5 talks of 15 minutes with 5minutes discussion, plus 5 minutes of changeover time.
We have noticed that Poster Sessions have been increasingly popular at conferences and we expect thatthere will be a large interest in presentations of this type. They are a great way to generate direct discussionswith other participants in a less formal way than in oral presentations. There will be one Poster Session eachday during lunchtime when presenters will be expected to be available to discuss their poster.
Young StatisticiansThe conference strongly encourages participation by young statisticians and there are two important activitieson the program especially for them. As part of the Welcome Reception on Monday evening there will be a“young statisticians’ corner”. This will be a great chance to ‘break the ice’ and meet fellow younger colleagues.
A whole session is planned for young statisticians on Monday afternoon, featuring a series of 15 minutepresentations on interesting topics encountered by the speakers in their work. These sessions have beenscheduled on the first day of the conference to maximise networking among young statisticians. Full detailswill be given in the online newsletter.
If you are new to the conference and would like to make a presentation, there are options for contributedtalks and contributed poster sessions. The poster sessions are a great way to meet other people. HoareResearch Software will sponsor $1000 in student prizes.
The NZSA has some funds available to support students at New Zealand Universities travelling to theconference. For further information contact Jennifer Brown ([email protected]). Note thatfirst year membership of NZSA is free for students at NZ universities or graduates of NZ universitiesstudying overseas.
Official Statistics DayOn Tuesday the Official Statistics theme will run throughout the day, starting with the plenary ForemanLecture to be delivered by Ray Chambers. It will be followed by an invited session organized by the Surveysand Management Special Interest Section. Speakers are Kirk Wolter (sponsored by SNZ) and Alastair Scott.There will also be contributed paper sessions devoted to the theme and plans are developing to hold adiscussion forum with leading representatives from ABS and SNZ.
Satellite Workshops
Introduction to Distance Sampling Workshop - 28-30 June 2006Presented by Steve Buckland, David Borchers and Rachel FewsterFor further information, see http://www.creem.st-and.ac.uk/NZ2006/Enquiries.htm
Stochastic Processes Workshop - 7 July 2006For further information, contact Dr Ilze Ziedins: [email protected]
Additional workshops are currently under consideration. As they develop further details will appear on theconference web site.
4
EditorialThanks to all thecorrespondents who havegot together interestingand timely accounts of themany and varied activitieswithin their groups, whichhas made this issue apleasure to compile. Mymain difficulty has been inthinking of anything towrite myself, and so I willkeep it brief!
This issue focusses largely on the ASC/NZSAConference which will be held in Auckland in earlyJuly. The fact that four pages are devoted to it reflectswhat a major undertaking it is. It is usual to thank theorganizers after the event has happened, so thanksfor all the planning so far.
The online newsletter will maintain up-to-dateinformation on satellite meetings to the mainconference. There is likely to be a substantial amountof new information subsequent to the printing of thehard-copy edition.
I will also maintain a page in the online newsletterlisting alternative, cheaper accommodation options inthe vicinity of conference venue. David Scott hasprovided an inital list. I’d be grateful to receive furtherinformation from members following anyarrangements you might have made for yourselves.Email details to [email protected].
Roger Littlejohn
President’s Column
“So what do you dothen?” She was youngand vivacious; alas, tooyoung and vivacious forme! We had exchangeda few words on a half-dozen occasions whenwe were both workingout in the university gymand now this dreadedquestion.
For some reason, instead of my usual self-and-profession-deprecating easy answer, I tried to explainthe excitement of discovering new information innumerical data. I used the analogy of scientistscoming to the site of an explosion and using thedirection and extent of the rubble to draw conclusionsabout the location and size of the explosion. “Cool,”she said, “just like CSI!”
I don’t really know if she ever did get tounderstand much of statistics from our conversation,but at least she went away with a positive feelingabout statistics. I was reminded of it again whenRay Littler lent me a book called “Freakonomics” bySteven Levitt and Stephen Dubner. Here is somethingfrom an NPR blurb for and interview with Levitt:
“Mr. Levitt uses statistics to examine matters ofeveryday life: His subjects range from abortion andcrime to game shows. Some of the chapter titles fromFreakonomics illustrate Levitt’s wide-rangingcuriosity: “Why Do Drug Dealers Still Live with TheirMoms?” and “How Is the Ku Klux Klan Like a Groupof Real-Estate Agents?” are two.” Perhaps mostfamous is his contention that the falling crime rate iscaused by the increased abortion rate in economicallystressed mothers.
I find myself led to wonder if the caution that leadsus to refrain from causal speculation is not the verysame thing that gives our profession a grey publicimage. It could be significant that the book was co-written with Dubner, a New York Times journalist.Maybe we should use journalists more often to getour message through to the public. Perhaps if wekeep a store of interesting vignettes from ourconsulting experience, slightly laundered forconfidentiality, but simplified and hyped up for effect,we could stir up a bit of interest in statistics.
Murray Jorgensen
Financial Snapshot - cash at bank
NZSA accounts $30,987Campbell Bequest Fund $59,236Wellington Statistics Group $912
David Bryant, Ray Brownrigg and Pierre Aillioton top of the Rimutukas, during the wine-tastingexcursion of the Wellington Hidden MarkovModel Workshop in December (photo: Jan Bulla).
5
Newsletter on Web
An online version of this newsletter is available athttp://nzsa.rsnz.org/Newsletter63/index.htm
It will be regularly updated with information andyour letters.
Submissions to the NewsletterThe Newsletter welcomes any submissions ofinterest to members of the NZSA. News aboutNew Zealand statisticians, statistical meetings,statistical organisations, statistics in education, orstatistical curiosities are suitable for inclusion.Letters that raise issues of importance to statisticsin New Zealand are also welcomed. Photographsof recent gatherings and new appointees are ofparticular interest. Electronic submissions arepreferred.
Next deadline 25 August, 2006.
Advertising In the NewsletterThe Newsletter accepts advertising of interest tostatisticians in New Zealand. Advertising is placedsubject to space considerations. Personaladvertising by NZSA members will be publishedfree. Other advertising is $250 per page, $140 perhalf page, and $75 per quarter page. Other sizescan be quoted on request. All advertising requestsshould be directed to the editor.
EditorRoger LittlejohnAgResearch, InvermayPrivate Bag 50034, Mosgiel, New ZealandPhone: +64-3-489-9082; Fax: +64-3-489-9037Email: [email protected]
New membersA warm welcome to new members of the NZSA:Dongwen Luo, Roger Macky, Sharon Browning,Martin Hazelton, Robyn Drake, Beatrix Jones,Matthew Schofield.
Statistics Education DVDA special session on statistics education wasorganised at the Dunedin NZSA Conference last year.Seven researchers at the University of Otago spokeabout their research and illustrated the statisticalprocedures used in their work. This was filmed duringthe conference with the aim of making a DVD, andsubsequently re-recorded in a studio environment.The second takes should be completed by about theend of March, with only two remaining to be filmedat the moment.
Since then Statistics New Zealand has alsorecorded two clips, which means nine case studieswill be included in the final DVD.
The DVD of the talks is being produced by theStaff in the Higher Education Development Unit atthe University of Otago. It will be available for use inall high schools in New Zealand, to assist in motivatingthe teaching of statistics. The corresponding data setswill also be available on an accompanying CD. JohnHarraway is already using this resource in hisSTAT110 lectures!
This project is supported by a grant of $750 fromthe Campbell Fund, and is being coordinated by JohnHarraway.
John Harraway
Campbell Estate FundThe NZSA was the recipient of a very generousdonation ($48,000) from Professor Campbell’s estate.
There is roughly $1500 funding available each yearfor special projects that are in the realm of ProfessorCampbell’s interests. Refer to http://nzsa.rsnz.org/funding.shtml for more details.
The fund has now grown to over $59,000 withaccumulated interest since 2001.
Applications are received twice a year (April/October) and are invited for funding for projects in2006/07. There is no formal application process butplease supply details of your project, the full projectbudget, the amount you are requesting, a shortstatement about why your project is within ProfessorCampbell’s interests, and your full contact details.
Statistical Methods for theScreening and Classification ofMicroarray Gene ExpressionData: A One-Day Workshop
Professor GeoffMcLachlan,University ofQueensland
20 April, 2006Venue: S Block, Room S.G.02University of WaikatoThis workshop is on analyzing microarray geneexpression data. It is aimed at statisticians/biostatisticians/bioinformaticians in general, as wellas to investigators working in areas in which relevantuse can be made of microarray gene expression data.The analysis of microarray gene expression data is avery important topic in science and applied science.The latter includes the biotechnology, pharmaceutical,chemical fields, as well as animal health and foodindustries.
Traditional statistical methodology has to be re-evaluated and modified to carry out the main analysesrequired for microarray data. Such data consist ofseveral thousands of genes measured over, say, tensor hundreds of patients (tissues). The workshop willfocus on the analysis of the preprocessed geneexpressions. That is, it is to be presented with theview that although there will be rapid developmentsin new technology for the production of high-throughput data, the methodology to analyze the“cleaned data” will still be essentially the same. A
key feature of theworkshop will be thedemonstration of themethods through thereporting of several casestudies. Another keyfeature is that the focuswill be on newlydeveloped methodology,in particular that of thepresenter, some of whichmay not be available inpublication form at thetime of the workshop.
The registration fee is $75 without dinner or $120with dinner, which will be held at the University’sPerforming Arts Centre, which has a wonderfulambiance overlooking the lake. The webpage forthe workshop is via www.stats.waikato.ac.nz.
ICOTS-7: Working Cooperativelyin Statistics Education
July 2 - 7, 2006Salvador (Bahia), Brazil
The International Association for Statistical Education(IASE) and the International Statistical Institute (ISI)are organising the Seventh International Conferenceon Teaching Statistics (ICOTS-7), which will behosted by the Brazilian Statistical Association (ABE)in Salvador (Bahia), Brazil, July 2-7, 2006.
Planning is now well advanced and the ICOTS-7website at http://www.maths.otago.ac.nz/icots7 isbeing continually updated. It contains summaries ofTopics and Sessions, abstracts for all the invitedpapers, contact addresses for invited speakers,session organisers, topic convenors and organisingcommittees and much more about the conferenceorganisation.
At the present time there are 223 invited paper,123 contributed papers, 120 poster presentations, fourspecial interest groups, 12 special sessions and arange of administrative meetings all contributing to avery full six day schedule. The Plenary speakersinclude some local people who are well known to us,namely Bryan Manly, Len Cook and Chris Wild.
The website also has information about the charmof Salvador Bahia and pictures of the guest roomsand conference facilities at the Othon Hotel wherethe conference is being held. Conference participantsare encouraged to stay at the Othon Hotel, and enjoyviews like this.
More information is available from the ICOTS-7website at http://www.maths.otago.ac.nz/icots7 orfrom the ICOTS IPC Chair Carmen Batanero([email protected]), the Programme Chair SusanStarkings ([email protected]), and the ScientificSecretary John Harraway ([email protected]).
John Harraway
7
Statistics Education News
International NewsICOTS-7, WorkingCooperatively inStatistics Education,Salvador (Bahia),Brazil, July 2-7, 2006.The InternationalAssociation for StatisticalEducation (IASE) andthe InternationalStatistical Institute (ISI)are organizing theSeventh International Conference on TeachingStatistics (ICOTS-7), which will be hosted by theBrazilian Statistical Association (ABE) in Salvador(Bahia), Brazil, July 2-7, 2006. All papers have beensubmitted. Information on the conference is availableat the ICOTS web site at http://www.maths.otago.ac.nz/icots7.
Joint ICMI/IASE Study, Statistics Educationin School Mathematics: Challenges for teachingand Teacher Education. The InternationalCommission on Mathematics Instruction (ICMI)Executive committee invited IASE to cooperate in ajoint study focused on statistics. The invitation wasaccepted by IASE, which proposed to merge theStudy Conference with IASE’s next RoundtableConference to be held in 2008 in Monterey, Mexico.Carmen Batenero is chair of the InternationalProgramme Committee (IPC) of the joint study.Planning is underway for this study, which will resultin a book being published in 2010. Meetings of theIPC will be held at ICOTS-7 and ISI-56. For moreinformation see: http://www.ugr.es/~icmi/iase_study.
New Publications. The Proceedings of the 2004IASE Roundtable on Curricular Development inStatistics Education, edited by Gail Burrill and MikeCamden, is now available on the website: http://www.stat.auckland.ac.nz/~iase/publications.php.The IASE Review 2005, which reports on IASEactivities in 2005 is also available on this website.
Local newsThe School Curriculum. The draft school
mathematics and statistics curriculum will be launchedin June 2006. Joint video-conferences betweenWellington and Auckland have been held to discussthe statistics strand. Statisticians and teachers havebeen involved in these discussions. If you areinterested in participating then please email Maxine([email protected]).
The CensusAtSchool Project. This project,sponsored by the Department of Statistics, University
of Auckland, Statistics New Zealand, and the Ministryof Education, was launched on August 15, 2005 andthe funding period has now been completed. Theproject was co-directed by Chris Wild and RachelCunliffe and aimed not only to give students theexperience of participating in a census but also toprovide rich classroom resources for Years 5 to 10students using the CensusAtSchool data. Theseclassroom resources are based on the proposed newstatistics curriculum. For more information, photos,press releases, news clips and exciting new classroomresources see: http://www.censusatschool.org.nz/.The project will continue to the extent that it can onvolunteer efforts and any further funding.
Year 13 Statistics Teacher Development.Palmerston North, Dunedin and Wellington teacherswere given a workshop day on the teaching of theYear 13 course Statistics and Modelling by MattRegan recently. The materials he used were frompast annual Year 13 statistics days presented byAuckland University lecturers. For more informationand access to these resources see: http://www.stat.auckland.ac.nz/~u47510x/teachers/index6.php.
Maxine Pfannkuch
Introduction to DistanceSampling Workshop28 – 30 June, 2006University of AucklandPresenters: Steve Buckland (U of St Andrews),David Borchers (U of St Andrews), Rachel Fewster(U of Auckland)
This is a three day introductory workshop, focusingon standard distance sampling methods. It is a blendof theory and practice, and participants will learn howto use the industry-standard program Distance. Youwill gain a solid grounding in both survey design andmethods of analysis for distance sampling surveys.Participants are encouraged to bring their own datasets and can expect to do some preliminary analyseswith their data. For details see http://www.creem.st-and.ac.uk/NZ2006/Enquiries.htm.
8
AwardsMarsden Fund Awards - BeatrixJones, David Bryant
Congratulations to Beatrix Jones (Massey University,Albany Campus) and David Bryant (University ofAuckland), who received Marsden Fund Awards inthis year’s round.
Beatrix’s project is“Design of parentageanalysis experiments: acase study foru n d e r s t a n d i n guncertainty in modelswith latent (unobserved)variables”, and is a twoyear Fast Start award.Beatrix (right)welcomes inquiries from prospective domesticpostgraduates who would like to do a funded honoursproject in this area. Parentage analysis refers to agroup of techniques ecologists use to understandmating and dispersal patterns of wild organisms. In aparentage analysis, genetic data are collected from agroup of offspring, and adults that may be theirparents. The genetic data is then used to match upoffspring with their parents; questions like “do olderadults produce more offspring?” can then beanswered. Unfortunately, it’s difficult to collectenough genetic data to do this parent-offspringmatching exactly. That’s where statisticians come in.When not all offspring can be confidently assigned toparents, the parent assignments can be treated aslatent variables. Demographic parameters can thenbe estimated by sampling from the joint posterior ofdemographic parameters and parent assignments.When this method is used, the relationship betweenthe amount of data collected and the precision of theresulting parameter estimates is complex. Beatrix’sproject will develop design methods so ecologists canknow before they begin data collection how muchdata they will need to achieve a desired precision.She hopes to extend these methods to other contextsas well.
David’s project is on “The statistics of phylogeneticnetworks”. Split networks and phylogenetic networksare data representation tools that have arisen out ofresearch into phylogenetic analysis (reconstructionof evolutionary history), but have been applied to fieldsas diverse as comparative linguistics and virology. Ineffect, they can be regarded as compactrepresentations of large collections of trees. Thecentral aim of this project will be to place these
representation tools into an appropriate statisticalframework, with the development of the appropriatenetwork models, validation and sampling techniques.These will be applied to the analysis of complexevolutionary events: particularly the study of viraltransmission histories and evolutionary analysis oforganisms with unclear species boundaries. Projectmembers include Alexei Drummond (Auckland),Noah Rosenberg (Michigan), Bernd Sturmfels(Berkeley) and Paul Tupper (McGill). Software isavailable at www.splits.org. David is currentlysearching for two PhD students.
NZIMA Programme - ModellingInvasive Weed Species
Congratulations toJennifer Brown, AlexJames and David Wall(U Canterbury) whoare Directors of anNZIMA Programmeon “ModellingInvasive Species andWeed Impact”. The aim of the 3-year programme isto bring mathematicians and statisticians together withbiologists to stimulate applied research that will benefitweed control and management.
The programme will begin with a 5 day workshopin Hanmer in April 2007. Up to 6 internationalmathematicians and statisticians, along with about 35New Zealanders, will be invited. The workshop’sformat will be introductory sessions by New Zealandweed managers outlining the current issues andproblems in weed management in NZ, followed bysessions from the international invitees on the latestdevelopments in relevant mathematical and statisticaltools. Each day, in the follow-up sessions, theworkshop attendees will identify the gap between theknowledge that can be gained from the the currentmathematical models and what is needed by NZ weedmanagers. The NZIMA programme goal is to bridgethat gap by stimulating relevant research amongst NZmathematicians and statisticians.
Applications are being called for students toundertake postgraduate and postdoctoral studies inthis area. The NZIMA programme will also includefollow-up workshops and regular newsletter styleupdates on the research.
For more information or for any queries contactJennifer Brown ([email protected])at the Biomathematics Research Centre at Universityof Canterbury. The Programme webpage contains afull and updated account of all Programme activities(http://www.math.canterbury.ac.nz/bio/NZIMA/).
9
Statistical Methods for the Screening andClassification of Microarray Gene ExpressionData
Hamilton, New ZealandApril 20, 2006Presenter: Geoff McLachlanWeb: http://www.stats.waikato.ac.nz/Email: [email protected]
Introduction to Distance Sampling
Auckland, New ZealandJune 28-30, 2006Web: http://www.creem.st-and.ac.uk/NZ2006/Enquiries.htmEmail: [email protected]
ICOTS 7 - Working Cooperatively inStatistics Education
See Gordon Smyth’s Australasian conference listhttp://www.statsci.org/conf/index.html
Conference Brief
ANZJSKerrie Mengersen (Queensland University ofTechnology) has now taken over the role of ManagingEditor of ANZJS. She was an Invited Speaker atlast year’s NZSA Conference, and is well-known forher contributions in Bayesian Statistics. ANZJS haspreviously not been consistent about assigningsubmissions to Applications or Theory & Methods,pretty much allowing the author to do so. This lead tosome problems, with stronger applications often beingpublished in Theory & Methods. The submissionprocedure has now changed, so that all submissionsgo the the Managing Editor, who will decide whichsection should process it.
Steve Haslett (Massey University) was appointedas a Theory and Methods Editor last year. KenRussell is doing his usual, very careful job as the newTechnical Editor. Special thanks to Russell Millar(right - see also page13), who is currentlycompleting theprocessing ofA p p l i c a t i o n sm a n u s c r i p t sreceived prior to2006. Jeff Wood(ANU) has now succeeded him as Applications Editor.
The number of Theory and Methods paperssubmitted to ANZJS continues to be high, althoughthe acceptance rate is not, and there are to be two‘bumper’ issues published. In order to be sure ofcovering associated costs, the NZSA AGM in 2005approved an increase in the membership fee.However, members are being offered the option ofselecting electronic-only access to the journal at areduced membership cost of $5 per member. Thisoption takes effect immediately, with renewal formsenclosed with the newsletter. Details of membershipoptions are given on Page 5.
The homepage for ANZJS is now http://www.statsoc.org.au/Publications/ANZJS.htm.
Accessing ANZJS onlineBlackwell Synergy
Members were emailed (1 February, 2006)instructions and their password from BlackwellPublishing on how to log-on to the Australian andNew Zealand Journal of Statistics online directlythrough Blackwell Synergy (www.blackwell-synergy.com).
Alternatively, check with your library aboutaccess to ANZJS online through the subscriber-based providers Ingenta, Ebsco, SwetsWise orOCLC.
10
Invited Speakers
Here we give details of the Conference Themes,Sessions, Invited Speakers and Session Organizers.
STATISTICAL METHODOLOGY
• Bayesian Statistics (organized by ClaireJordan & Kerrie Mengersen). Christian Robert(University Paris-Dauphine), Robert McCulloch(University of Chicago).
• Exact Methods of Statistical Inference(organized by Chris Lloyd) Alan Agresti (FloridaState University), Chris Lloyd (MelbourneBusiness School), Ivan Chan (Merck ResearchLaboratories).
• Saddlepoint Methods in ModernStatistical Inference (organized by NevilleWeber) Ron Butler (Colorado State University),John Robinson (University of Sydney).
• Multivariate and High Dimensional Data(organized by Robert Kohn) David Donoho(see Keynote Speakers), Matt Wand (UNSW).
• Spatial Statistics (organized by AdrianBaddeley) Christian Lantuejoul (Centre deGeostatistique).
• Stochastic Processes (organized by IlzeZiedens) Adam Shwartz (Technion IsraelInstitute of Technology), Dirk Kroese (Universityof Queensland).
• Modern Goodness-of-Fit Theory andMethods (organized by Estate Khmaladze).
COMPUTATIONAL STATISTICS
• Statistical and Machine Learning(organized by Matt Wand) Geoff McLachlan(University of Queensland), Alex Smola (NationalICT Australia / ANU).
• Computationally Intensive Statistics(organized by Chris Carter) Xiao-li Meng (seeKeynote Speakers), Robert Kohn (UNSW).
• Statistical Computing (organized by PaulMurrell and Kuldeep Kumar) Junji Nakano(Institute of Statistical Mathematics, Tokyo), BillVenables (CSIRO CMIS).
• Resampling Methods (organized by MartiAnderson) Peter Hall (see Keynote Speakers),Jiming Jiang (University of California, Davis).
STATISTICS IN BIOLOGY AND MEDICINE
• Statistics in Biological Science (organizedby Harold Henderson & Simon Barry) MarkBurgman (University of Melbourne), Tony Pettitt(Queensland University of Technology).
• Bioinformatics (organized by Chris Triggs)Gordon Smyth (Walter and Eliza Hall Institute ofMedical Research), Allen Rodrigo (University ofAuckland).
• Statistical Genetics (organized by JamesCurran) Sharon Browning (University ofAuckland).
• Forensic Statistics (organized by JamesCurran, Janet Chaseling & Claude Roux)James Curran (University of Auckland), JohnBuckleton (Institute of Environmental Science andResearch Ltd)
• Medical Sciences (organized by KatrinaSharples & Peter Howley) Annette Dobson(University of Queensland), Gita Mishra (MedicalResearch Council - National Survey of Health &Development at the Royal Free & UniversityCollege London Medical School), Robert Gibberd(Health Services Research Group, CCEB).
• Statistics in Ecology and the Environment(organized by Marti Anderson) SteveBuckland (University of St Andrews), RachelFewster (University of Auckland).
• Multivariate Statistics in Ecology (organizedby Marti Anderson) Brian McArdle (Universityof Auckland), David Warton (UNSW).
SPECIAL INTEREST SECTIONS
• Industrial Statistics (organized by DavidWhitaker & Ross McVinish) NozerSingpurwalla (Institute for Reliability and RiskAnalysis, George Washington University), ChinDiew Lai (Massey University).
• Surveys and Management (organized bySteve Haslett & Robert Clark) Kirk Wolter(University of Chicago), Alistair Scott (Universityof Auckland).
• Young Statisticians (organized by JasonThomas & Ian Woods)
• Econometrics and Finance (organized byKerrie Mengersen & Chris Carter) CathyChen (Feng Chia University), Mike So (HongKong University of Science & Technology).
The program of invited sessions and speakerscontinues to be developed. For example, we areworking on a session on econometrics and time seriesand one on syndromic surveillance for diseaseoutbreaks.
The program will be updated continually at theconference web site (http://www.statsnz2006. com/invitedsp.asp), so please check there for details.
11
Postcard from Cambridge
With an eclectic mixture of emotions; eageranticipation, trepidation, excitement and sadness, Imade the ‘epic’ journey to England in order to beginmy PhD at Cambridge University. My project,provisionally entitled “Integrated Modelling of BirdPopulations”, aspires to develop Bayesian statisticalmethodology to underpin, and improve, advice givento Government and other interested parties on thestate of the UK’s bird populations.
I’m very lucky to be supervised by dual Steves –Dr Steve Freeman from the British Trust ofOrnithology (BTO), and Prof Steve Brooks in theStatistics Lab at Cambridge. I am also most fortunateto be literally overwhelmed with vast amounts of highquality data. Count data on numerous species of birdshas been collected annually using a standardised mistnetting approach by hundreds of volunteers throughoutthe UK as part of the BTO’s ‘Constant EffortScheme’, from 1983 to date.
Survival information is also available from ringrecoveries, nest success data from another BTOsurvey (the ‘Nest Record Scheme’), as well asvarious habitat and climate variables.
At this early stage I’ve concentrated my effortson modelling Sedge Warbler data (photograph below,thanks to Dawn Balmer), and have produced variousindices of abundance and productivity. The“Integrated” part is still to come. Tragically I’ve yetto meet a Sedge Warbler as they’re sensibly winteringin Africa, but come summer I’m very keen to seemist netting in operation … although the 3am start isnot so appealing.
The Statistics Lab is a very stimulating andenjoyable place to work, despite the fact I’m
A Biometrician at theMathematics in Industry StudyGroup
The Mathematics in Industry Study Group (MISG)was started in Oxford during the 1960’s as a pathwayto show industry the power of mathematics whenapplied to their particular problems. The Australasianversion of MISG has been running since 1984. Thegeneral idea is that industry representatives bringalong a problem and the workshop attendees spend aweek working together toward a solution. In recentyears it has attracted more statistical problems, andas a result greater numbers of statisticians arebeginning to attend.
This year Crop & Food Research contributed aproblem (attendees are pictured working on it below,see also page 14) relating regional and soil propertyinformation and farm management history to varioussoil health measurements, to create a decision supportsystem for NZ farmers. I was excited to work closelywith experienced statisticians and mathematiciansfrom a variety of backgrounds and disciplines, and itwas interesting to see the different ways in whichthese practitioners approached the problem.However, I won’t say there weren’t any ‘moments’between mathematicians and statisticians (all friendlyof course)! The enthusiasm of everyone involved inMISG was fantastic, resulting in some very goodresults and hopefully rewarding collaborations in thefuture.
Esther Meenken
remorselessly teased over my pronunciation of data(‘darrrrrrrrta’). My research promises to be verychallenging but extremely rewarding and wonderfullyinteresting.
Thanks to my colleagues at AgResearch for theircontinued support and encouragement, and also tomy lecturers and tutors at Otago University forfuelling my interested in Statistical Ecology.
Vanessa Cave
12
Local SceneProteus Wildlife ResearchConsultants
We are continuing our Ministry of Fisheries-fundedproject on the sustainability of seabird bycatch in NZfisheries. Our most recent work has focussed on thedifficulties of estimating bycatch, as well as on thebest methods for collecting demographic informationon the birds themselves. As usual, consulting problemssuggest interesting statistical issues that may be worthpursuing beyond the context of the original application.David Fletcher has decided to employ his new PhDstudent at the University of Otago, Peter Dillingham,to carry out some of the work. Peter’s thesis will bein the area of animal population modelling andmanagement. He is from the US and already hasexperience working in statistical ecology, on theconservation management of Stellar sealions. DarrylMacKenzie (along with his co-authors) has recentlyhad his first book published “Occupancy estimationand modeling: inferring patterns and dynamics of
species occurrence”.This is a synthesis ofcurrent estimationmethods for thepresence/absence ofspecies that aredetected imperfectly; atopic he first beganworking on in 2001while working at the USGeological SurveysPatuxent WildlifeResearch Center.While obviouslypleased with its
publication, the unfortunate thing is that readers havebegun to point out a few typos that he missed.
Darryl MacKenzieAgResearchRe comings and goings among the statisticians in oursection – a nil report! Instead, we are all lining up forAgResearch’s recently introduced “Long Service”award celebrations, to be held on the 10, 15, 20, 25,30, 40 and 50th anniversaries of joining.Congratulations to John Waller (pictured above rightwith Neil Cox, Tony Pleasants and Peter Johnstone)for attaining 30 years of service, and being the firstto receive this award – well done, John!Congratulations also to Ken Dodds and Linda Murray(our science administrator) who will be having their20-year celebrations as this goes to press. Some of
the other younger groupies (sorry, sectionies) areheading towards the 25 years of service mark, withthree still aiming for the first mark (at 10 years).Meanwhile, the other half of us are in the long slogfrom 30 to 40, with the old-age pension arriving beforeour first long-service recognition. (DS: too bad, youwin some, you lose some!) One wonders whetherthe spacing was decided upon by mathematicalmeans.... (statistical analyses are pending). (Jokingaside, congratulations to AgResearch for introducingrecognition of service awards! – a great conceptwhich is excellent for staff morale).
In September, Ken Dodds et al. gave an invitedtalk entitled “Practical aspects of a genetic evaluationsystem using parentage assigned from geneticmarkers” at the Association for Advancement ofAnimal Breeding and Genetics Conference at Noosa,Queensland. Also, David Baird, Peter Johnstone,Martin Upsdell et al. gave a talk entitled “Turning IDon its head” at the NZBio Conference on Traceabilityin Dunedin. In November, Ken Dodds attended theSymposium in Biostatistics: Statistical Genetics andGenomics in the USA, as well as two one-day courseson “Design and analysis of microarray experiments”and “Analysis of modern population genetic data”.Also, Peter Johnstone attended the workshop onexperimental design in honour of Nye John at theUniversity of Wollongong. In December, RogerLittlejohn gave a talk entitled “EDA of rhythms indeer data: feeding, biting and growth hormone” atthe Second Workshop on Hidden Markov Models andComplex Systems in Wellington.
During the December/January period, David Bairdand Peter Johnstone took to long-distance travelling.David and his family spent four weeks in England (inconnection with GenStat), and made trips to Chad,India and Thailand totalling another 4-5 weeks. Peterand his wife holidayed at exotic spots in the BajaCalifornia peninsula, “just south of the border, downMexico way”.
Dave Saville
13
Public Radio, New Scientist, and National Geographic.James reports how the rat eluded barrages of trapsand swam 400m across open sea to a neighbouringisland, presumably in search of a mate. James caughtup with it several weeks later, but by boat.
We are very proud to harbour the author of oneof the community’s most influential publications. RossIhaka’s 1996 paper with Robert Gentleman,introducing the R language, has just been announcedas the most cited paper of the last 10 years in themathematical sciences. To date, there are more than20 books based on R, with at least ten more in thepipeline.
More publicity followed the results from theCensusAtSchool programme hosted by thedepartment, including features in The New ZealandHerald, The Herald On Sunday, and Canvas. Inaddition, our Annual Teacher’s Day drew over 90participants and some very favourable feedback.
Russell Millar was back on the nation’s TVscreens with more lottery wit and wisdom, this timeabout the new game Big Wednesday. In recognitionof his new career as TV Lotteries Celebrity, Russellhas stepped down from the job of Applications Editorfor ANZJS. Not for a life of idleness, of course, ashe and Rachel Fewster are both commencing termsas Associate Editors of Biometrics this month.
Congratulations to Marti Anderson and RichardFord, who gained an $80,000 grant from the VC’sUniversity Development Fund for a project entitled“Investigating the impact of multiple stressors onbenthic communities”. Congratulations also to JamesCurran, who has obtained a 3-year $176,000 contractwith the Forensic Science Service of the UK.
And finally, it emerges that the Stats departmentis responsible for the world’s most famous Paintballcontest. Check out http://images.google.com andsearch on “paintball”. No prizes for guessing whosegruesome hand appears as the number 1 hit, butcheques and donations may be forwarded to BrianMcArdle [c/o newsletter editor...]
Rachel Fewster
University of Auckland
We welcome our new lecturer, Sharon Browning(below), who joined us in October fromGlaxoSmithKline in NorthCarolina. Sharon’sspecialisation is statisticalgenetics of human diseases.It’s a warm welcome backfor Sharon, who did herHonours degree at Aucklandbefore going to the Universityof Washington in Seattle forher PhD. By a happycoincidence, Sharon arrivedback just in time torecommend her erstwhileHonours supervisor IlzeZiedins (right) for a teachingaward. Congratulations toIlze for her well-deservedDean’s Award forExcellence in Teaching.
Unfortunately for us, there is a Law ofConservation of Statistical Geneticists at work. Weare very sorry to say farewell to Mik Black, who hasaccepted a post with the Bioinformatics Group in theBiochemistry Department of the University of Otago.Mik has been at Auckland since completing his PhDat Purdue University in 2002.
Luckily, we were able to retrieve Chris Triggs fromhis sabbatical visit to the Centre of Excellence inNutritional Genomics at UC Davis. While there, Chrisattended his first and only conference in which oneof the talks made it to the front page of the NewYork Times: completion of the HapMap project. AsChris muses, this never happens at Stats conferences.[Sample lead: “Cox announces new proportionalhazards model!”]
Congratulations to Tim Langlois, for successfullydefending his thesis “Influence of reef-associatedpredators on adjacent soft-sediment communities”,supervised by Marti Anderson. Welcome to SammieJia, who has won a university international fees bursaryto start his PhD with Russell Millar and MartiAnderson.
Members of the department have loomed large inthe public eye over the last few months. Star PhDstudent and failed rat-catcher James Russell wroteup his doomed attempts to capture a particularlyathletic rat on the Noises Islands in the Hauraki Gulf.The Tom-and-Jerry style episode was published inthe top international journal Nature in October. Thestory was picked up by over 100 news agenciesworldwide, including the BBC, Washington National
14
University of Waikato
2005 saw several changes in the department.Following the departure of James Curran, who is nowat Auckland University, and the retirement ofProfessor Nye John, we also had to farewell our longtime departmental assistant, Karen Devoy. In October,Karen left to take up a position as PA to ProfessorIan Graham, who was originally in the ComputerScience Department and then long time Dean of theSchool of Computing and Mathematical Sciences,here at the University. He now heads up his ownvery successful computing company, EndaceTechnology. Karen has been replaced by RhondaRobertson who is doing a marvellous job filling the“big shoes” that Karen left.
In honour of Nye’s retirement, a workshop onExperimental Design was organised by Ken Russell
Massey University, Albany
Towards the end of January, a number of statisticiansfrom Albany attended the Mathematics in IndustryStudy Group (MISG), coordinated by ProfessorGraeme Wake from the Albany Campus MathematicsGroup. The MISG essentially involves bringingindustrial problems in Mathematics and Statistics toUniversity researchers (with the expectation that theywill be solved!). Barry McDonald (Massey, Albany),Penny Bilton (Massey student) and Mini Ghosh(Massey, Post-Doc) examined data from the SoilSustainability Group, with a view to developing a soilmanagement index. They were assisted by fellowstatisticians Kaye Marion (RMIT), Walt Davis,Victoria Wei and Simon Leong (Statistics NewZealand), Esther Meenken (Crop & Food Ltd), alongwith mathematician Nev Fowkes (UWA) and afurther three soil experts. Work in action is picturedbelow. In addition, Claire Jordan from our Stats
Group, Henning Rasmussen (UW Ontario) andRatneesh Suri (Massey student) looked at‘Expectations for loss of supply in the New Zealandpower system, a problem provided by TranspowerNZ.
Jeff Hunter continues as the Professor of Statistics(although now only on a part-time basis). Not contentwith only 40% time in the Institute, he has got himselfinvolved (for a limited period) as the Project Managerfor the Campus Strategic Positioning Project workingwith the Deputy Vice Chancellor for the AlbanyCampus. In November he visited Professor EugeneSeneta at the University of Sydney to give a seminarand initiate some collaborative activity. He is planningan extensive leave trip to Europe and the US in themiddle of the year to participate in variousconferences. (More on that next time!)
Beatrix Jones attended the 8th New ZealandMolecular Ecology Meeting, December 3-5, held atthe YMCA Wainui Park on the Banks Peninsula. Asyou may guess from the venue, this annual conferenceis fun and informal, with a lot of students presenting.
Beatrix gave a talk called “Bayesian inference forbrood structured data”, about using genetic data tounderstand the complicated mating patterns oforganisms like fish. Although it was an ecologyconference, many of the participants use sophisticatedstatistical methods for phylogenetic, populationgenetic, and pedigree data, so Beatrix felt quite athome! The meeting provided Beatrix with a greatopportunity to interact with those collecting data andto understand the related scientific questions.
Tasos Tsoularis has recently returned from a sixmonth trip to Europe and the USA. Whilst in England,Tasos spent time in the Bradford University Schoolof Management, where he was researching the topicof stochastic learning algorithms applied to problemsin Management Science. Whilst in Miami, Tasos wentto a mathematical-biology conference and presenteda paper on “Learning strategies for a predatoroperating in variable model-mimic-alternative preyenvironments”.
Paul Cowpertwait
Nye John (left) shares a thoughtful moment withMurray Jorgensen and Peter Johnstone at theStation (U. Waikato campus)
15
University of Cantebury
There have been lots of comings and goings recentlyhere at Canterbury. Irene Hudson has moved toAdelaide to become an Associate Professor and headof the statistics group at the University of SouthernAustralia. We wish her all the best and hope the groupprospers.
We are currently advertising for a new lectureshipposition in statistics, which has a closing date of 31March.
Over summer Jennifer Brown has had 2 summerscholarship students working with her. Gavin Bellgained a UC scholarship and worked with Jenniferon developing a sequential, adaptive, 2-phase samplingdesign. Gavin was also awarded a Statistics NewZealand prize for his exceptional undergraduateperformance last year. Jason Bentley, on aBiomathematics Research Centre scholarship, beganwork on graphical models for describing phenology.The phenological study they are working with is amontane forest in Nigeria and their interest is in thetiming and intensity of flowering, fruiting and leaf setevents.
The NZIMA programme on Modelling InvasiveSpecies and Weed Impact organised by Jennifer, andAlex James and David Wall (UC) is well underway.The website for the programme has full details, http:/
/www.math.canterbury.ac.nz/bio/NZIMA/. Theyare currently advertising for a postdoctoral fellow and2 PhD/MSc students. The programme officiallybegins in April 2007 with a workshop at Hanmer.
Dominic has been working with his student JamesRoscoe on a UC scholarship to develop a hiddenMarkov model for assessing the health of pre-termbabies. His work forms part of a major study toimprove outcome forecasts with Marco Reale andGlynn Russell, a neonatal paediatrician atChristchurch Women’s hospital. Dominic alsopresented some of this work at the Second Workshopon Hidden Markov Models and Complex Systems inWellington.
Christian Robert of Ceremade-Université, Paris-Dauphine has accepted a Visiting Erskine Fellowshipfrom 8 July-20 August 2006, hosted by Dominic Lee.Christian’s areas of interest include BayesianAnalysis, Computational Statistics, Latent VariableModels and Applied Modelling.
Mike Steel has recently been hosting huge numbersof visitors following the successful annual NZPhylogenetics Conference in Kaikoura from 12-17February. A full list is available on the departmentalnewsletter on our website.
Easaw Chacko has been kept rather busy oversummer making sure the department will continue itsfinancial success in the coming year, for which weare all very thankful. The case for our replacementlecturer was ensured because of our group’s financialhealth.
Marco Reale and Carl Scarrott have beensupervising a summer project by Alethea Rea (anotherUC scholarship student), investigating spill-overeffects between international markets (namely Japanand US). They have also been hosting GranvilleTunnicliffe-Wilson from Lancaster University.
John Newell from the National University ofIreland, Galway is visiting the department (hosted byMarco) until June 2006 and is joined by his wife,Karen. John’s main areas of research involve appliedstatistics, including multivariate survival analysisproblems, computational inference, functional dataanalysis, and applications in medicine and sportsscience. He is the Consultant Statistician for the SportsPerformance Unit at Glasgow Celtic Football Cluband, on a recent visit to Auckland, had the dubiousdistinction of appearing on nationwide TV as astretcher-bearer at the NZ Knights 1-1 draw withleague leaders Adelaide!
Marco is also hosting Dominico Piccolo from theUniversity of Naples, who, most of you will be aware,is a prolific time series expert. Dominico is currentlyteaching our 3rd year inference paper.
Carl Scarrott
of the University of Wollongong, NSW. This was heldat the end of the year. David Whitaker also attendedas an invited speaker.
So far this year, things have been relatively quietin the department. Murray Jorgensen has returnedfrom sabbatical and at present we have Dave Johnson,formerly of Loughborough University, visiting thedepartment. He is helping out by teaching our firstyear Management Statistics paper, and he will be withus till Easter.
Finally, an upcoming event in the department willbe a One-Day Workshop to be held on Thursday April20, 2006. The presenter will be Professor GeoffMcLachlan from the University of Queensland. Thetitle of the workshop is “Statistical Methods for theScreening and Classification of Microarray GeneExpression Data”. Registration details will be available
at www.stats.waikato.ac.nz shortly.Recent seminars in the department:
Lyn Hunt (University of Waikato) “Using multiplechoice questions as an assessment device forstatistical thinking”David Whitaker (University of Waikato) “The stateof the art in the generation of efficient statisticaldesigns”
Judi McWhirter
16
Department of Conservation
I started at DOC in October 2005, beginning withsome line transect distance sampling analysis, whichwas part of a study of the impact of 1080 operationson tomtits. With Ian Westbrooke, I have also beeninvolved with the analysis of public favourability ofDOC, a critique of methods used for analysis of dataon crowding at busy conservation sites, and thedevelopment of a document to help DOC staff usethe SPSS classification and regression tree packageAnswerTree. Michael Ryan (from Statistics NewZealand) has been making his regular Friday morningvisits.
Ian has been busy with various kiwi and stoatstudies, the design of visitor surveys and is continuinghis crusade against the use of pie charts (especiallythe 3-D variety) in DOC. Along with kiwi scientistHugh Robertson, he successfully carried out a numberof workshops on the use of Leslie matrices andKaplan-Meier survival analysis. In December hepresented a talk at the Australasian OrnothologicalSociety conference titled “Design and analysis of birdmonitoring data – some experiences from a statisticalviewpoint”, where he stressed the importance of goodobjective setting in a study prior to data collection.He has been scarce of late due to Christmas holidays,followed by a tramping trip which resulted in ahelicopter ride (thanks to search and rescue!), andmost recently, and unfortunately, a familybereavement. We are happy to have him back in theoffice now.
Adam Smith has been “working” (aka diving) inFiordland over the past few weeks, helping to establishbaseline biological data for the marine reserves. Heis making good progress in his MSc with MartiAnderson, entitled the “Validation of the New ZealandMarine Environment Classification for shallow rockyreef fish communities” and is fitting in some statisticalconsulting in Wellington.
My contract finishes at the end of March so I canembark on a PhD at the University of Queensland. Iplan to keep up with the office gossip and hope thatIan has the inclination to mentor another recentgraduate, as he has done with myself and Adam.
Carla Meurk
Wellington Statistics Group
The Wellington Statistics Group (WSG), a local groupof the NZSA, continues to meet regularly. The Grouphas now gratefully received sponsorship from theMinistry of Social Development, in addition to StatisticsNew Zealand, Statistics Research Associates Ltd,and Victoria University of Wellington. Attendance atthe early evening meetings remains pretty good;
Massey University, Turitea
Given the amount of time we have had to spendconstructing and correcting our PBRF portfolios, it’samazing anyone has had any time to do anything else,let alone produce research outputs. Nevertheless theactivity has somehow continued.
Our new Professor of Statistics, Martin Hazelton(below), has arrived from the University of WesternAustralia. We are all very pleased, not only to haveappointed someone, but to have them actually turn
typically 20 to 30 people, with 45 or so sometimes.Since the last WSG news appeared in the NZSA
Newsletter, in reverse chronological order there havebeen WSG talks given by:
Rod Lea (Institute of Environmental Scienceand Research Ltd) “Genome Informatics andDisease Susceptibility” December 2005;
Tony Vignaux (Operations Research, VUW)“Incorporating factors such as Chance and Risk”November 2005;
Paul Jose (Psychology, VUW) “MakingModeration and Mediation Quick, Easy, and Clear”September 2005;
Phil Lester (Biological Sciences, VUW) “Usingdiscriminant analysis to predict the establishmentsuccess of exotic ant species in New Zealand”August 2005.
The next WSG meeting will be in mid March 2006,addressed by Geoff Chambers (Reader in Cell andMolecular Biosciences at VUW) on “Out ofTaiwan? Genetics sheds new light on Maori origins”.
Anyone who does not presently receive WSGannouncements and who wishes to be informed offuture events is welcome to contact the WSGConvenor, John Haywood: [email protected].
John Haywood
17
Programme. Geoff gave a talk based on this work tothe Manawatu branch of the RSNZ in February.
Jonathan Godfrey’s move from student tosupervisor has taken him places recently. While hisstudent (Deborah Brunning) has managed to do thenumber crunching, Jonathan has written the wordsaround the outside of a supplementary analysis ofthe costs of blindness data that is about to be madepublic by the Royal NZ Foundation of the Blind. Thedata is right up Jonathan’s alley both privately andprofessionally as it involves sparse responses from asmaller than desirable sample. He’s been toWellington recently to show the Ministry of SocialDevelopment and the Minister for Disability Issuesthat their plans for a Costs of Disability study need tobe considered rather carefully.
Geoff Jones
up and stay for at least a month. Particularly pleasingis the fact that Martin has arrived in time to submit aPBRF portfolio!
Now that Martin has arrived to take over theleadership of the statistics group, Mark Bebbingtonis counting down the days until the handover, althoughit is just a rumour that he is notching his door jamb.Mark did find time to be an invited speaker at the 4thInternational Statistical Seismology Workshop in Japanin January. He returned full of sushi and sashimi but,along with all the other gaijin, had wasted no time inunplugging the electrically heated toilet seat in hisroom.
Siva Ganesh was a Distinguished Invited Speakerat the 2nd Annual Asia IT Congress in Bangkok inNovember. He chaired a day-long stream on “IT andData Warehousing” and gave a talk entitled “DataMining: How data rich is your organisation?”. Thisseems a particularly pertinent question for those ofus still filling in our PBRF portfolios.
Ganesalingam was invited by the University ofJaffna and University of Peradeniya, Sri Lanka toundertake collaborative research work. Because ofongoing political unrest in the northern part of SriLanka, Jaffna University was highly understaffed andGanes’ visit was well appreciated by his formercolleagues. He gave a series of lectures in MultivariateStatistical Analysis and presented two seminarscovering his current research work in Ranked SetSampling and Discriminant Analysis.
Doug Stirling has just returned from eight monthsat the University of Reading in England. During thistime, he added a chapter about multiple regressionto CAST and developed a customised version ofCAST for a Reading University eLearning course toteach statistics to climatologists in Africa. A newrelease of CAST is now available and can be usedand downloaded from http://cast.massey.ac.nz. Duringhis sabbatical, Doug gave seminars about CAST inReading and Essen andran a workshop at theApplied Statistics 2005conference in Bled,Slovenia. He returned justin time to complete hisPBRF portfolio.
Steve Haslett, Alasdair Noble and Geoff Joneshave been commuting regularly (twice a month) toStatistics New Zealand to work with them on twoOSRDAC-funded projects using small-areaestimation techniques, one on Maori expenditurepatterns and the other on sub-regional employmentlevels. Steve Haslett paid a final visit to Nepal inFebruary to finish off the poverty estimation that heand Geoff Jones have been doing for the World Food
Statistical Methods - StatisticsNew Zealand
By the time you read this I would hope all of youhave completed your Population Census forms andreturned them to Statistics New Zealand, either bygiving it to your enumerator, posting it back, or byfilling it out on the Web.
The biggest news is that our Wellington office hasmoved from Aorangi House to a new building,Statistics House. Very recognisable! It’s the verygreen building between the waterfront and the railwaystation. The part of SNZ in the Public Trust buildingshould have moved in about the time you read this.So for the first time in living memory all the SNZpeople in Wellington will be in one building.
While the Census raises our profile, I find somepeople think that’s all Statistics NZ does and wonderwhat we do when it’s not running. You can be assuredthat we are all busy collecting other data andanalysing it. The key change over the last decadehas been the increasing use of data collected by othersto extend and supplement our outputs. For example,Statistics NZ has recently released the first resultsfrom the Longitudinal Employer Employee Database(LEED). This has been a very big job as we arelinking the PAYE data from the IRD with theinformation we have about the businesses wherepeople work. It has been very involved as manyaspects of the linking process are quite complicated.For example, SNZ needed to link employees overtime (it is L-EED) using their IRD numbers withoutactually having their IRD numbers, which had toremain at IRD for legal, confidentiality and privacyreasons. As well as the technical issues there was alarge amount of consultation with interested parties,like the Privacy Commissioner, because, as you canimagine there are major privacy and confidentiality
18
concerns that needed to be worked through andresolved.
I’ll do a quick plug for our Expert UsersNewsletter. This free SNZ newsletter aims to informusers of statistics about new statistical outputs,significant developments, survey redesigns, newclassifications and standards, and upcomingpresentations and seminars. The newsletter ispublished by email approximately every fortnight. Tosubscribe please send an email [email protected] with ‘subscribeexpertuser’ in the subject line.
OSResearch is looking for Expressions of Interest(EOI) from people wanting to do research that maybenefit the official statistics system (OSS). The OSSencompasses all statistical outputs from allgovernment departments, not just SNZ. AnyOSResearch research has to be sponsored by agovernment department, so having one attached toyour EOI helps, However it is possible for your EOIto not have one. OSResearch may be able to directyou to possible sponsors. [email protected] for more info.
As noted in my last report Statistical Methods werein the process of recruiting from the 2005 graduates.The successful candidates were Rebecca Bagma,Claire Sun, Nellie Yang and Johanna Prebble. Theyhave all recently started, beginning with a session inWellington with all the other graduates starting in otherareas of SNZ. Also joining us in Christchurch from aresearch arm of Fonterra is John Pearson. PollyStuart so liked her time here on her RSNZ teacher’sscholarship that she has joined us permanently inWellington. Karla Helgeson has come to us fromStatistics Canada for 18 months. Steve Johnston hasreturned to the fold part time while still working onhis PhD at VUW, while Frances Krsinich has beenback doing some contract work for us. ChristineBycroft is due to return soon after spending sometime working for the Office of National Statistics inthe UK. Grace Chiang is temporarily absent afterhaving a baby just before Christmas. Philippa Grahamhas recently retired and will be sorely missed, whileRachael Viles has gone to ONS in the UK for a fewyears and Chris Gianos has moved sideways to workin Injury Statistics. John Lopdell has gone to spend18 months at Westat in Washington DC, where heshould meet up with Andrea Piesse. We also have apermanent manager, Vina Cullum, who is also handlingJohn’s work while he’s away. Now that RichardPenny is relieved of his manager’s duties (relievedbeing his view as well!) he is stationed in our Aucklandoffice until July where he is hoping to build up thelinks between SNZ and the research community inAuckland. Eric Nordholt, a Senior Researcher at
Statistics Netherlands, is visiting. The primary purposeis to do a review of our confidentiality proceduresand research for OSRDAC, but while he is here weare taking an opportunity to pick his brain about thework being done in Statistics Netherlands. He is alsogiving talks for the OSS Professional StatisticiansNetwork in Wellington.
On the conference and workshop front, Tim Dukewent to a conference on data linking in Canberra runby SSAI. Allyson Seyb went to the Association forSurvey Computing conference on Maximising DataValue, Data Use & Re-use where she presented apaper, “Statistics New Zealand’s LongitudinalBusiness Frame”. Rico Namay went to the StatisticsCanada Methodology Symposium in November wherehe presented a paper, “Estimating Undercounting ofVehicle-related Injury Cases in New Zealand: AProbabilistic Data Integration and Capture-recaptureApproach”. Walter Davis, Mike Doherty, VictoriaWei, Simon Leong and Richard Penny went to therecent MISG meeting at Massey Albany. All enjoyedbeing able to work on a problem and discuss statisticsin general without being interrupted by meetings,phone calls, emails etc. At MISG Walter gave aplenary talk “Some Principles of 21st CenturyStatistics”, and Richard a talk to the students, “MakingSense of Data”. Mike Camden was down to go to aconfidentiality conference in Luxembourg inDecember, but ended up being admitted to hospital,instead.
Richard Penny
Victoria University
The Statistics and OperationsResearch (STOR) Group atVUW have had quite a fewexciting developments on thestaff front. We have two newarrivals, both in the second halfof 2005. Colleen Kelly (right)started in August, and is ourofficial “Consulting Statistician”,giving statistical advice to graduate students and staffin other disciplines around the University. The rest ofthe University view Colleen’s arrival as somethingakin to a miracle, since there is now someone who(nearly always!) has time to see them and to sit andlisten to their statistical problems. Mark Johnstonstarted in late June, as a new lecturer in OperationsResearch. Mark came to VUW after 4years at U.Essex teaching mathematics and computer science;he has a PhD in Operations Research from MasseyUniversity, on combinatorial optimization.
Now elevated to a higher plane, Dong Wang andStefanka Chukova were promoted to the rank of
19
Reader/Associate Professor at the start of 2006,where they sit alongside Shirley Pledger and MeganClark. In May 2005, on her return from sabbatical,Megan added the Deputy Head of School position toher already formidable list of administrativeresponsibilities; how’s that for a homecomingpresent?!
Stefanka Chukova took over from Shirley Pledgeras STOR Programme Director on 1 July 2005. TheGroup was very grateful to Shirley for guiding usthrough the lead up work required for the externalreview (of Maths, as well as Stats and OR) that wehad during August. The review commented favourablyon many aspects of the Group’s activities. As aconsequence of the recommendations, we haverecently advertised a new permanent position with afocus on ‘Applied Statistics’, the latter interpreted ina quite broad sense. We are also very grateful toStefanka, who has taken on the various administrativetasks required of the Programme Director and is stillseen smiling on a regular basis!
Courses have started in 2006 with an increase innumbers in just about all our courses, and in ourgraduate students. Of those graduates nearingcompletion, Nuovella Williams went part time on herPhD from March 2004, and is about to submit.Nuovella was employed on a temporary contract asa Systems Analyst by the Caribbean GovernmentAccounting Reform Project in Montserrat fromMarch 2004 to Sept 2004, and has been employedfull time as a Research Officer at the UK Office forNational Statistics from Jan 2005 to the present. SteveJohnston is also making good progress with his PhD.In January Steve went to the 4th InternationalWorkshop on Statistical Seismology, held at ShonanVillage, south of Tokyo, Japan. Steve presented aposter entitled “An accelerating moment releaseversion of the stress release model”. Others fromour group who went to the same conference includedDavid Vere-Jones, Ray Brownrigg and JunkoMurakami. Junko is one of DVJ’s postdocs on theHidden Markov Models program, which is coordinatedby DVJ and supported by NZIMA. Johanna Prebbleis very close to completing her MSc on the NewZealand Health Survey, and has recently started workat Statistics New Zealand. Johanna is the second oftwo successful Masters students (following KylieMason a year earlier) supervised by Richard Arnold,supported by awards from Public Health Intelligencein the Ministry of Health.
Various staff have been travelling, includingStefanka and Dong on separate overseas trips for acouple of months during the 2005/6 summer. RichardArnold attended the 25th Workshop on BayesianInference and Maximum Entropy Methods in Science
and Engineering in San Jose, in August. Richardpresented a paper on the application of Bayesianstatistics to seismology (earthquakes as signaturesof the stress in the earth’s crust). Estate Khmaladzealso travelled extensively over the summer. InDecember Estate was a keynote lecturer at theEURANDOM Workshop “Economics and financeof extremes”, and an invited speaker at the EuropeanScience Foundation Workshop on Model Specificationin Santander, Spain. While in Europe, Estate alsoworked with Miguel Delgado (Madrid) and JohnEinmahl (Tilburg), before returning to EURANDOMto give a series of lectures on the differentiability ofset-valued functions and their use in statistics. Onthe way back to New Zealand, Estate fitted in workingvisits to see Hira Koul (Michigan State) and RogerKoenker & Stephen Portnoy (Illinois).
In other news, Megan unfortunately broke her legby falling when leaving a party in late November; itwas a steep driveway, honestly, before you speculatefurther. In December 2005, Stefanka Chukova andMark Johnston helped to organise the OperationalResearch Society of New Zealand’s 40th conference,held at VUW. One highlight was that Stefankaorganised two Stochastic OR sessions, the first ofwhich was chaired by John Haywood. As I told theslightly bemused audience, they were making historyjust by being there, since there had never been anysessions devoted just to stochastic OR in the previous39 years! Tony Vignaux fittingly gave the leading talkin that first Stochastic OR session; fitting since Tonyhas done so much to champion OR in NZ, andStochastic OR in particular - a fact acknowledgedwith thanks by many at the conference. Othermembers of our group who presented papers at theconference included Richard Arnold, Stefanka, John,Ivy Liu, and Dong Wang. After helping to organisethe conference, Mark couldn’t actually get to it sincehe was in England at the time, at a PlanSIG planningand scheduling workshop held at City University,London. Ivy has a visitor coming to VUW for a monthin May 2006: Bhramar Mukherjee from the Universityof Florida. Finally, to conclude on a very happy note,we were pleased to enjoy a (too short) visit from YuHayakawa at the end of February 2006. Yu returnedto VUW for 10 days, to work with Stefanka andRichard. While she was here we managed tostatistically test (and refute) the hypothesis that she’dlost some of her noted capacity for fun-filled nightsout since she left VUW in early 2004. It was clearthat Yu has settled in well to her AssociateProfessorial position at Waseda University in Tokyo,and that the demands of that role have not affectedher ability to party.
![Molecular Profiling of Pancreatic Adenocarcinoma and ... › content › canres › 63 › 10 › 2649.f… · [CANCER RESEARCH 63, 2649–2657, May 15, 2003] Molecular Profiling](https://static.documents.pub/doc/80x56/5f0c1d237e708231d433cead/molecular-profiling-of-pancreatic-adenocarcinoma-and-a-content-a-canres.jpg)
