ISSN 0112-2649stats.org.nz

The New Zealand Statistical Association

NewsletterNumber 79 July 2017

Welcomes

• Welcome from the President• Editorial

Conferences

• Coming Conferences

Stories of Interest

• Obituary: Alastair Scott• Review: Introductory Bayesian MOOCs

Stats Education News

• Statistics Education Group• NZSA Education Committee

Local News

Statistics at:• AgResearch• AUT: Biostatistics• AUT: School of Engineering, Computer and Mathematical Sciences• Department of Conservation• Massey University: Albany• Massey University: Palmerston North• Plant and Food• Statistics New Zealand• Statistics Research Associates• The University of Auckland• University of Otago: Department of Mathematics and Statistics• University of Otago: Division of Health Sciences

Welcome

by IAN WESTBROOKE

NZSA President

Departures, conferences and education, are someof the themes in this newsletter, along with newsof statisticians in NZ.

Most significant was the loss of one of our leadingstatisticians. I sent this message to Alastair Scott’sfamily on behalf of NZSA:

Our members salute Alastair Scott’s memory - as astatistician with strong international standing whoalso contributed much to statistics in New Zealandand to the NZ Statistical Association. Alastair servedas President in 1989 and 1990, and continuedto make insightful contributions regularly at ourconferences right up to our latest conference latelast year, both formally and informally. He wasunfailingly interested in helping and encouragingstatisticians - younger and more mature.

On a personal note, Alastair gave some of the firststatistics lectures I attended in my first year atAuckland University, and his first year teachingthere, to many hundreds in a cavernous theatre inthe then brand-new “Brutalist” building where hisdepartment is still situated today.

Our association expresses its condolences to Alastair’sfamily, colleagues and friends. His legacy lives on -through his contributions to statistics internationallyand in NZ, and with his personal contact, teaching,collaboration and encouragement to many of ourmembers and beyond.

There are a number of retirements noted

throughout the newsletter, with the retirement ofRoss Ihaka featuring at our conference later thisyear. It is great to see that many of those whohave formally retired are continuing to contribute- following in our own ways the great example setby Alastair.

I’d like in particular to express the Association’sthanks to recent retiree Howard Edwards, forhis contribution to the Association, especiallyhis role as treasurer over the past few years,and his extra efforts in the transitions from ourprevious treasurer and to our new treasurer, DanielGerhard.

Good reading

Ian

2

Editorial

by MARIE FITCH

Kia ora koutou,

I hope the year is going well for you.

As always this newsletter includes updates on thecomings and goings of our New Zealand statisticscommunity.

Those of you whose lives intersected in some waywith Professor Alastair Scott, will, I hope, take timeto read the tribute to “one of the finest statisticiansNew Zealand has produced”. Little did I realisesitting in his lectures as a 20 year old that I wouldbe collating this newsletter and recognising hislife-long contributions to our discipline. Thankyou Ilze, Chris, Chris and Atokohu for penningsuch a fitting tribute.

Thanks also to Richard Arnold for contributinghis article Reviewing Introductory Bayesian MOOCs.I would welcome similar contributions to futurenewsletters.

I am also pleased to be able to include for the firsttime since I have been the editor, contributionsfrom ... Well I will leave that for you to work out,there are three of them! Thanks to these new orreturned contributors and of course thanks to allthe regular contributors, without their contributionthis newsletter pretty much would not exist!

The next newsletter will be published in early 2018and will include reflections on the December NZSAconference. (If you would like to volunteer towrite those I’d love to hear from you!) If you haveanything that you would like to contribute to thatnewsletter please let me know and plan to haveyour contribution to me by mid December (2017).

Nga mihi,

Marie

3

Conferences

2017 IARSC-ARS/NZSA Conference

Call for Papers and Conference Announcement

We are pleased to announce the 68th Conferenceof the New Zealand Statistical Association. Theconference will be held in conjunction withthe Asian Regional Section of the InternationalAssociation of Statistical Computing of theISI. The conference will take place in theworld-class Owen G. Glenn Business School atthe University of Auckland from the 10–14thDecember 2017. Registration details, as well asinformation about accommodation, can be foundat www.nzsa2017.com. Registration is open now.

The aim of the conference is to provide a forumfor the discussion and exchange of ideas, newconcepts and recent methods in statistics. Beinga joint conference, there will be a blend oftraditional statistical theory and practice alongwith new developments in statistical computing,

especially as related to the R statistical system. Theconference marks the retirement of Ross Ihaka, oneof the co-founders of R.

Four half-day workshops will be held on Sunday10 December, the day preceeding the mainconference: "Faster R code" by Thomas Lumley,"Getting to Know Grid Graphics" by Paul Murrell,"Analysing spatial point patterns using spatstat" byRolf Turner, and "Graphics in R" by Chris Wild.

Call for Papers

Participants wishing to present a paper atthe conference should submit an abstract inEnglish (plain text or LaTeX) not longer than500 words, including title, name of authoror authors, keywords and contact information.The submission should be uploaded on theconference website www.nzsa2017.com/abstracts/The deadline is 31 August 2017.

VenueThe conference and conference dinner will be held in the University of Auckland’s world-class Business School.

Owen G. Glenn Building, City Campus, The University of Auckland, 12 Grafton Road, Auckland, New Zealand.

This modern 11-storey building has 74,000 m2 of floor space and a 1,000 carpark space in its basement. There are two 600-seat lecture theatres, and two 300-seat and one 150-seat lecture theatres.

Auckland, called the ‘City of Sails’, is New Zealand’s largest city with a population of about 1.5 million people (almost 1/3 of NZ’s population). The Waitemata Harbour is within walking distance of the university and it stretches eastward to sever-al popular swimming beaches (including Mission Bay and Kohimarama). The quaint Devonport suburb is a short ferry ride from the Auckland Ferry Building. Some islands worth visiting by ferry are Tiritiri Matangi, Rangitoto and Waiheke. Also within walking distance of the university is the Auckland War Memorial Museum, one of the two most important NZ museums, in the Auckland Domain. The Parnell and Newmar-ket shopping precincts are close by.

For other ideas, some useful websites are:

www.AucklandNZ.comwww.AucklandTourism.co.nzwww.NewZealand.com

Auckland has a temperate climate. Being summer in December, the sun can be particularly strong, and we recom-mend a good hat, sunglasses and some suntan lotion!

AccommodationAuckland offers a wide range of accommodation choices, ranging from backpackers to 5-star hotels. Some general websites include:

www.booking.com www.trivago.co.nzwww.tripadvisor.co.nz www.wotif.co.nz

A block-booking with University Hall has been secured by the University of Auckland; please see the conference website.

CommitteesNext IASC-ARS organiser Philip Yu, Hong Kong University

General ChairThomas Yee University of Auckland

Local Organizing Committee (LOC)James Curran (chairperson), University of AucklandRolf Turner University of Auckland

Scientific Program Committee (SPC)Ciprian Giurcaneanu (chairperson), University of AucklandClaudio Conversano University of Cagliari, ItalyRobin Hankin Auckland University of TechnologyGiampiero Marra University College London, UKIan McKeague Columbia University, USAOsamu Komori University of Fukui, JapanAntony Unwin University of Augsburg, Germany

Consultative MembersRepresentative of the NZSA: Ian Westbrooke, Department of Conservation, New ZealandRepresentative of the Department of Statistics at Auckland: Ilze ZiedinsRepresentatives of the ARS of IASC: Chun-houh Chen, Jung Jin Lee.

Joint Meeting of 10th Asian Regional Section (ARS) of the International Association for Sta-tistical Computing (IASC) and the NZ Statistical Association (NZSA)

10--14 Dec 2017

On the retirement of Ross Ihaka

http://www.iascars2017.comhttp://www.nzsa2017.com

The University of Auckland, New Zealand

Contact

nzsa2017@stat.auckland.ac.nziascars2017@stat.auckland.ac.nz

IASC-ARS

NZSA 2017

Workshops (half-day)

Faster R Code Thomas LumleyGetting to Know Grid Graphics Paul MurrellAnalysing Spatial Point Patterns Rolf Turner Using spatstatGraphics in R and iNZight Chris Wild

4

Other Conferences

Biometrics by the BorderKingscliff NSW 26-30 November 2017 held at Mantra at Salt Beach, Kingscliff in far northern NSWhttp://www.biometricsociety.org.au/

Sixth Wellington Workshop in Probability and Mathematical Statistics4-6 December 2017, Victoria University of Wellington http://sms.victoria.ac.nz/Events/Dec2017WWPMS

SEEM 2017: Statistical Ecology and Environmental Monitoring conference 2017Queenstown, New Zealand Dec 6-8 2017 The website is www.stats.otago.ac.nz/conferences/seem2017

5

Stories of interest

Alastair John Scott, Emeritus Professor of Statistics(November 17, 1939 – May 25, 2017)

by ILZE ZIEDINS, CHRIS WILD, CHRIS TRIGGS, ATAKOHU MIDDLETON, DEPARTMENT OF STATISTICS,UNIVERSITY OF AUCKLAND

Alastair Scott, one of the finest statisticians NewZealand has produced, died in late May, aged 77,on the North Shore of Auckland. He is survived byhis wife, Margaret, children Andrew and Julie, sixgrandchildren, and his sister Marilyn. There wasstanding room only at the University of Auckland’sMaclaurin Chapel when he was farewelled onWednesday May 31.

Alastair served the University of Auckland withdistinction from 1972 until 2016. After hisformal retirement in 2005 he held only fractionalappointments but continued working nearly fulltime, regardless. His research was characterised bydeep insight and he made pioneering contributionsacross a wide range of statistical fields. He wasacknowledged, in particular, as a world leader insurvey sampling theory and in the developmentof methods to efficiently obtain and analyse data

from medical studies. His methods are applied ina wide range of areas, notably in public health.Beyond research, he contributed prolifically tothe statistical profession in academia, government,and society.

But Alastair wasn’t just a world-leading statistician.He was also sociable, charming, mischievous, wasalways first with the news, loved a good gossip,but was always supportive of others. “Everybodywho met Alastair became a lifelong friend,” sayscolleague Chris Triggs. Adds colleague ThomasLumley: “Alastair had a national and internationalreputation for being friendly and helpful.”

Alastair grew up in Te Awamutu playing rugby andcricket, and had a lifelong love of the games. Heplayed in the high school curtain-raiser for the1956 Waikato vs Springbok match in Hamilton,playing as hooker. He also played representativecricket for Waikato Colts against the touringAustralian national team in 1956-57.

In his last year at Te Awamutu College, Alastairwas Dux, and won the school’s first universityscholarship via a hothouse exam involving 10papers, two in each of five subjects, sat overa period of less than three weeks. In 1958,he started studying maths at the University ofAuckland. While living at student hall O’Rorke, hefell for fellow resident Margaret Hedley, a sciencestudent from Te Kuiti. Margaret would later teachhigh-school chemistry and biology.

Also in 1958, David Thomson, who was alsostudying maths, met the pair at O’Rorke Hall.“Alastair was one of the best people you could ever

6

hope to meet, indeed a best friend, and I can saythe same about Margaret,” he says. “Alastair waspatient, especially when explaining concepts toothers who had not understood something. Asa student, I often benefited from his knowledgeand his willingness to explain meanings of things Ifound difficult.” David and Alastair later shared astudent flat, playing bridge (coached by Alastair),going skiing at Ruapehu, and running a basketballteam in a YMCA competition.

Alastair gained his BSc in Mathematics in 1961 andhis MSc in Mathematics in 1962. He then won aPhD scholarship in pure mathematics (algebra) tothe University of Chicago. During the nine-monthgap before going to the United States, Alastairmarried Margaret (he later told a colleague thatone of the chief pleasures of marriage was that henever had to go dancing ever again) and workedfor the Applied Mathematics Division of the-thenDepartment of Science and Industrial Research(DSIR) in Wellington. It was while at the DSIRthat Alastair became fascinated by statistics anddecided to switch his subject at Chicago from puremathematics to statistics.

At Chicago, Alastair’s doctoral supervisor wasDavid Wallace, now an Emeritus Professor there.In turn, David’s supervisor had been John Tukey,who played a key role in the development andstudy of statistics in the mid-1900s. Wallaceencouraged Alastair to take a holiday job at BellLabs, the research laboratory where Tukey worked.

At Bell Labs, Alastair met David Brillinger,who became a collaborator and lifelong friend,and future statistical computing luminary JohnChambers was an office mate. Alastair gained hisPhD in 1965, his thesis titled Allocation of Effortin the Design of Selection Procedures. When Davidmoved to the London School of Economics (LSE),Alastair wrote him a letter asking if there were anyjobs in its Department of Statistics.

Indeed, there were - and Alastair ended uptaking over the job vacated by fellow Aucklandgraduate George Seber! Alastair’s time at theLSE, from 1965 to 1972, was formative. If theDSIR and Bell Labs had been Alastair’s statisticalnurseries, Chicago and LSE were the sources of hisformal statistical education. At LSE, Alastair cameunder the influence of James Durbin, a British

statistician and econometrician known particularlyfor his work on time series analysis and serialcorrelation in regression, and Alan Stuart, whoseresearch focused on sampling theory, efficiencyof tests, and measures of association. Alastairalso met Fred Smith, now Emeritus Professor forDemography at the University of Southampton,who also became another lifelong friend andresearch collaborator.

The LSE years were a time of hard work butalso the foundation for lifelong friendships andcollaboration. Fred remembers that he, Alastairand David Brillinger “had in common recentmarriages and the births of our children. Weshared experiences and baby clothes and carriagesas well as interests in politics and sports. Therewas a very good common room in LSE where theconversation flowed across many topics, switchingfrom politics to economics, then sport and familymatters, as well as statistics.” LSE at the time“had the greatest collection of statistical talent ina London college” and had a “radical image” thatsuited Alastair’s political interests, says Fred.

Alastair played tennis with Fred at the LSEsportsground, employing a “vicious slicedbackhand”, and while in the LSE staff cricketteam, which played other London colleges, wasknown as a “demon leg spinner.” One memorablesummer, Fred and the Scott family were part ofa group who drove a minibus from Southamptonto Oslo, Norway, to attend a meeting of Europeanstatisticians. “I can’t remember the meeting”, saysFred, “but I still remember the trip.”

Alastair made a conscious decision to make hiscareer in New Zealand: he and Margaret wantedtheir children to grow up as New Zealanders. TheScott family returned to New Zealand in 1972, andAlastair joined what was then the Department ofMathematics and Statistics.

Distance proved no barrier. Alastair held visitingpositions at Bell Labs, the Universities of NorthCarolina, Wisconsin, and California Berkeley in theUS, and at the University of Southampton in theUK. Throughout his career he was offered a streamof job offers from prestigious universities overseas,but turned them down.

In 1994, the University of Auckland’s statisticsstaff, led by George Seber, had a very amicable

7

divorce from the Department of Mathematicsand Statistics. But due to George’s ill-healthat the time, it was Alastair who became thefirst head of the Department of Statistics. Thepair set the tone for the department that stillexists: hard-working, but welcoming, and social.Says David Thomson: “Alastair believed thatlecturers needed to demonstrate humility andrespect for their students, and wherever possible,to show a human face to those they were teaching.”George, who retired in 2002, and Alastairwere instrumental in building the Departmentof Statistics into the largest such department inAustralasia.

Rachel Cunliffe, a MSc in Statistics graduatewho is now co-director of the Department’sCensusAtSchool project, was one of Alastair’sstudents. “When he was my lecturer, he madeit a joy to be in his classes. I remember sittingin my seat smiling because he found a way totalk about his twin grandchildren in almost everylecture. When there wasn’t a way to do so naturallyas part of our material, he’d bring up an overheadtransparency showing us the latest photos.” Rachelsays that Alastair “taught me to work hard, but tonot make life all about work. He was the best kindof human: incredibly smart and thoughtful, yet fullof warmth and humour with a continual sparkle inhis eye - someone who was undoubtedly enjoyinglife and genuinely always made time for others.”

Statistics sits within the Faculty of Science. Deanof Science John Hosking says while Alastaircontributed in many obvious ways to the faculty,“his main contribution was less visible and morepersonal. Alastair was an important mentor toa huge range of people, particularly youngerresearchers. His affable personality and abilityto deal with anyone at any level with a smile andan appropriate word made him a natural for thisrole, and the sage advice that followed was alwaysincredibly useful and effective.”

The generosity didn’t stop there, says John:“Alastair was also notably free with his wallet at theSenior Common Room, regarding a professorialsalary as coming with an obligation to share hisgood fortune with others yet to reach that level.”In 2005, the year Alastair officially retired, aconference in Auckland in his honour attracted

the largest concentration of first-rank internationalstatisticians in New Zealand in one place at onetime. In his “retirement” years - as we mentionedearlier, he was still working - he remained a centralfigure in the Department and the common roomwas still his office.

From 2005 to 2008, Alastair was on New Zealand’sfirst advisory committee on Official Statistics, thechief source of external advice for the Minister ofStatistics about the health of the Official Statisticssystem. He also advised government statisticsagencies in Australia, Canada, and the US.

Alastair was a Fellow of the Royal Society of NewZealand, the American Statistical Association, theInstitute of Mathematical Statistics, and the RoyalStatistical Society, and an honorary life memberof the New Zealand Statistical Association (NZSA).In 2006 he received the prestigious WaksbergAward from the American Statistical Associationand Statistical Society of Canada for outstandingcontributions to survey methodology. That wasfollowed in 2012 by the NZSA’s premier prize, theCampbell Award, for sustained contribution to thepromotion and development of statistics.

In November last year, in Christchurch, Alastairwas awarded the Royal Society of New Zealand’sJones Medal, which recognised his lifetimecontribution in the mathematical sciences.According to Thomas Lumley, “Statistics is mostlynot about maths, so for a statistician to get amathematics award requires significant theoreticalcontributions.”

On the night, Alastair said he felt a great dealof pleasure in receiving an award named forNew Zealand’s most celebrated mathematician, SirVaughan Jones, as the latter was in the first classthat he had taught at the University of Aucklandafter returning from London in 1972. Also in thatclass was the late Keith Worsley, whom Alastairsaid was his “most successful PhD student” andwho later became a world leader in neurostatistics.

Alastair John Scott had a full life professionallyand personally. He was a valued colleague andfriend. We will miss him greatly and we extendour heartfelt condolences to Margaret, Andrewand Julie and his wider family and friends all overthe world.

8

Review: Introductory Bayesian MOOCs

by RICHARD ARNOLD, SCHOOL OF MATHEMATICS AND STATISTICS, VICTORIA UNIVERSITY OF WELLINGTON

richard.arnold@vuw.ac.nz

For large classes doing standard introductorymaterial, Massive Open Online Courses (MOOCs)provide unparalleled efficiency for both studentsand teachers. Each works in his or her own time.The lecturer only needs to deliver lectures once,but can also hone, polish and rework material. Thestudent can view and re-view material as slowly oras quickly as they like. Students don’t hold eachother up, or feel as if they are left behind.

Whether MOOCs end up replacing a substantialproportion of conventional University teachingremains to be seen: they suffer from the samedrawbacks of conventional distance learning(reduced student engagement, the lack of themotivating experience of being in a room withpeers, lack of direct access to the lecturer, aswell as the problem of verifying who has actuallycompleted an assignment or quiz). Howeverthe new and developing technology associatedwith MOOCs (new video presentation techniques,integration of written materials with video andonline content, peer discussion forums) do meanthat MOOCs are a significant step forward fordistance learners. For universities they providea showcase for teaching staff, attracting newstudents to enrol, and also a convenient way toallow students to access introductory pre-requisitematerial prior to enrolment in a regular taughtprogramme.

So what MOOCs are available for students ofBayesian statistics? Here I review two introductoryBayesian statistics courses which are available onthe Coursera platform (www.coursera.com). Theyare ‘Bayesian Statistics: From Concept to DataAnalysis’, taught by Herbie Lee from Universityof California at Santa Cruz, and the ‘BayesianStatistics’ module of the ‘Master Statistics withR’ programme taught by Mine Çetinkaya-Rundel,David Banks, Colin Rundel and Merlise Clyde ofDuke University.

The Coursera platform has some very nice featuresfor its students. Courses are bundled intoweek-long blocks, with short videos (3-10 minutes

typically), full video transcripts, supportingdocuments, and quizzes (mostly multi-choiceanswers, though some allowing numerical or textanswers to questions). Videos can be playedat variable speed (happily without altering theinstructor’s voice pitch), and instructors canarrange for the video to pause and ask the studenta question at key points.

Students pay around US$50 per month for accessto a course – which entitles them to submit quizzesfor credit, and participate in peer discussionforums. For the Santa Cruz course, this coversthe full 4 week course. The Duke Bayesian coursecould also be completed in 4 weeks, thoughit is one of 5 courses in a ‘Specialization’ inintroductory statistics with R – which at therecommended pacing would take 5-6 months tocomplete. Students can of course go faster if theywish. At the end of the course students receivea certificate. It is also possible to gain accessto the course content without payment, but thisremoves access to assessments, feedback and peerdiscussion forums – and there is no recognition oflearning available at the end.

The two courses contain broadly similar content– Bayes Theorem in the familiar and simplediscrete and continuous settings seen in anyBayesian course (Beta-Binomial, Gamma-Poisson,Normal-Normal). Both spend time talking aboutconjugacy, Bayesian hypothesis testing, and theBayesian approach to linear regression. Bothspend time comparing the frequentist and Bayesianapproaches, with the Duke course going furtherand spending time on Lindley’s Paradox.

Courses on Bayesian statistics have two veryparticular demands which distinguish themfrom other courses, even from frequentistcourses: namely requirements for mathematicsand computational methods. The former presentscomplications for statistics programs of all kindsworldwide: we have an increasing demand forstatistical training for students who have a lowlevel of mathematical knowledge. The latter is

9

more easily addressed by teaching students thecomputational methods they need as and whenthey need them.

The two courses differ in their pre-requisite levelsof mathematics. The Duke course assumes onlyvery basic mathematics, whereas the Santa Cruzcourse explicitly states that first year Universitycalculus is required – although more at thelevel of familiarity than mastery. Both coursesnevertheless spend time developing expressionsfor posteriors, involving the necessary concepts offinite and infinite sums and integrals. Consistentwith its higher mathematical pre-requisite, theSanta Cruz course had more of the feel ofa traditional university course with instructorHerbie Lee writing out extensive expressions andderivations using a Lightboard. In contrast, theDuke course instructors stood in front of theircontent slides, where fully formed mathematicalexpressions appeared – for the most part asadditional information for those students whowere interested in the detail.

Both courses use R for analysis, providing detailedcode and demonstration examples of Bayesiananalyses. The Santa Cruz course also providesequivalent instruction for each example in Excel.These examples are an excellent start for studentswho want to go off and do something practical

after the course – the Duke course in particulargiving an extensive worked example of modelselection and averaging in linear regression usingthe BAS package in R.

These courses both provide a learner with amotivating start for learning more about Bayesianstatistics – the different styles of presentationare engaging and the amount of thought andpreparation that has gone into the two coursesis very apparent. The Duke course is conceptuallyambitious – with its coverage of frequentistparadoxes and model selection – and a nicefeature is a set of four short interviews withBayesian academics and practitioners. TheSanta Cruz course covers less ground, but doesso with a well-paced higher technical intensity,building on an assumption of greater mathematicalknowledge.

I’d recommend both courses to students who wantto know what Bayesian statistics is about – whetheror not they intend to take it further. Thoseintending to carry on with further study will bewell prepared for higher level courses in Bayesianstatistics.

Note: This review first appeared in the ISBA bulletinand is used here with their permission.

10

News from the Statistics Education Teams

Statistics Education News

by MAXINE PFANNKUCH

International NewsThe 10th International Research Forum onStatistical Reasoning, Thinking and Literacywill be held from 2 to 8 July 2017 in Rotorua,New Zealand. Maxine Pfannkuch, StephanieBudgett and Pip Arnold are organizing theconference. Chris Wild is the keynote speakerand Anna Fergusson and Anne Patel, AucklandUniversity, will be presenting their research. Thetheme of the Forum is Innovations in statisticalmodeling to connect data, chance and context. (seehttp://srtl.info)

USCOTS 2017, 18-20 May, the US conferenceon Teaching Statistics was held at Penn StateUniversity. Chris Wild was the opening keynotespeaker. His topic was Oh say can you see?on graphics and components of visualizations.https://www.causeweb.org/cause/uscots/uscots17

World Statistics Congress 2017, 16-21July in Marrakech, Morocco. TheInternational Association for StatisticalEducation (IASE) has organized about 10Invited Paper Sessions for this conference(see http://www.isi2017.org/)

IASE 2017 Satellite Conference, 11-14 July inRabat, Morocco. The theme of the conferenceis Teaching statistics in a data rich world. See:http://iase-web.org/conference/satellite17/

The Tenth International Conference onTeaching Statistics (ICOTS-10), 8-13 July 2018in Kyoto, Japan. Preparations are now underwayfor this conference. Abstracts for proposed papersare due 15 August 2017. See: http://icots.info/10/

Local NewsPeople in statistics education research.

Through an invitation, Anna Fergusson AucklandUniversity, is presenting again at the AmericanStatistical Association’s Teachers’ Workshops aspart of the Joint Statistical Meeting in August2017 in Baltimore. As part of her preparationfor this Meeting, Anna gave a very wellreceived webinar on Statistical reasoning with DataCards (see K-12-Statistics-Education-Webinars athttp://www.amstat.org/asa/education/)

Anna Fergusson and Malia Puloka will bepresenting their research at the 2017 NewZealand Association of Mathematics Conference inChristchurch. Pip Arnold, Cognition Education,has been invited to the USA in October toparticipate in an updated version of the publicationGAISE (Guidelines for Assessment and Instructionin Statistics Education), which is published by theAmerican Statistical Association.

CensusAtSchool Project. The CensusAtSchoolproject is currently running the 2017 survey ofNZ school students, with 13,000 participatingso far. Chris Wild and Rachel Cunliffeare Co-Directors of the Project and AtakohuMiddleton writes press releases. Media coverageof initial findings from the survey on screentime, packed lunches and pocket money has beensubstantive. Rachel was interviewed on radio andTV about the findings. Anne Patel is responsiblefor the teaching resources on the website. See:www.censusatschool.org.nz

FutureLearn MOOC - Data to Insight: AnIntroduction to Data Analysis. Chris Wild,University of Auckland, with the assistance of Mike

11

Forster again ran this eight-week course, fromMarch to May 2017 for 10,500 registered students.The course will start again in October this year.

The course is centered on Chris Wild’s innovativesoftware iNZight.

NZSA Education Committee, June 2017

by MIKE CAMDEN

Changing digital technology and its implicationsfor work and society appear frequently in themedia. These changes have implications forstatistical education. Our initial ideas are onthe CensusAtSchool website. The paper raisesquestions about how statistical educators can act toanticipate change and guide the further evolutionof school statistics. We welcome feedback frommembers on this issue.

We agreed four priorities for this year. They are:

• Statistics education in the primary sector• Establishing an online statistics teaching

journal for New Zealand• Datasets: Sourcing, processing and making

fit for assessment, data that engages students• Future of our statistics curriculum at school

level.

Work is progressing on these.

Some specific issues of interest to us at presentare:

• the NCEA external assessments in statistics;i.e., the exams

• the NZAMT15 conference in Christchurch inJune this year: http://www.nzamt2017.com/.We plan to dialogue with teachers at a forumthere.

The Ministry of Education commissioned thissummary of background information for its reviewof curriculum and related resources: Review andMaintenance Programme (RAMP) Mathematics and

Statistics; Themes in the research literature; AlexNeill and Rosemary Hipkins. The report can beaccessed at https://ncea.tki.org.nz. It reflects ourapproach to statistical education well.

Do you need a tune-up, or (as is more likely) knowsome people who need one? Try this new animatededucational resource from colleagues across theTasman: http://statstuneup.com.au/.

John Harraway retired from Otago University’sDepartment of Mathematics and Statistics at theend of April, and keeps some teaching and researchactivities. John has a very long history of veryactive involvement in statistical education, bothhere and internationally. He helped to bringICOTS 3 here in 1990, took leadership in manyinternational statistical education conferences, andchaired the IASE 2013 to 2015. He has been anactive member of our education committee in ourwork in recent years on curriculum, assessment,resources, and software.

John’s innovative work includes making theresearch of some of his Otago colleagues fromother disciplines available as videos and datasets,and promoting access to and use of Genstatfor Teaching and Learning. He remains thechair of the Advisory Board of the InternationalStatistics Literacy Project. John has recently beenpromoting the magnificent ISLP set of resources athttps://iase-web.org/islp/Resources.php.

Best wishes, John, and keep up your interest instatistical education!

12

Local News

Statistics at ...

AgResearch Statistics Group

by JOHN KOOLAARD

AgResearch’s Bioinformatics and Statistics Teammet for a two-day meeting at the end ofMay in Palmerston North. This was the firstface-to-face meeting of our staff from all thecampuses for nearly three years. You may noticeEsther Meenken in the photo, who earlier thisyear joined the stats group at Lincoln, afterhaving worked at Plant and Food Research.

AUT:

Department of Biostatistics andEpidemiology

by STEVE TAYLOR

The department has recently welcomed threenew staff members: Robert Borotkanics, RoseSisk and Philip Prah. Robert studied inthe US and joined us from Australia. Hecompleted his doctoral training from the JohnsHopkins Bloomberg School of Public Health andbrings many international collaborations to thedepartment. Rose joined last year as an intern andthen accepted the role of Research Officer. Shestudied statistics at the University of Glasgow andworked for two years as analyst for NHS Scotland

before moving to New Zealand. Philip also joinedus from the UK where he worked for eight yearsas statistician on various studies including thelarge NATSAL study of sexual behaviour. Amongother roles, Rose and Philip will be working on thePacific Islands Families Study, which has recentlycommenced three new HRC-funded projects.

Last year, the department awarded Steve Taylora doctoral scholarship and he has recentlycommenced a PhD looking at randomised trialsin neurorehabilitation. Nick Garrett continuesas head of department and Alain Vandal returnsfrom sabbatical in July.

13

School of Engineering, Computerand Mathematical Sciences

by KATE LEE

The analytic team would like to welcome ournew lecturer, Nuttanan Wichitaksorn (Nate).Nate’s current specialization is Econometricsand Statistics with an emphasis on a Bayesianapproach and applications in economics, finance,health science, and social science. His researchinterests also lie in the fields of computationalstatistics using Monte Carlo methods, financialanalytics, skew distributions, copula modeling,robust quantile regression, and GPU-based parallelcomputation. Before joining AUT, Nate hasbeen a research fellow at Thailand DevelopmentResearch Institute and was a Lecturer in Statisticsat University of Canterbury, New Zealand wherehe is now an Adjunct Fellow. Prior to that, heobtained a PhD in Econometrics from Universityof Sydney, Australia, and two MAs in Economicsfrom Rutgers University, USA, and ChulalongkornUniversity, Thailand.

The Mathematical Sciences Research Group at AUTplans to run the 4th AUT Mathematical SciencesSymposium, 23-24 November 2017, in AUT CityCampus. More details will be announced soon.Researchers in Analytics, Applied Mathematics andStatistics areas are welcome to attend.

Dr Sarah Marshall attended the SAS UsersConference (SUNZ) in Wellington on 11th May2017. This event provided the opportunity tohear about new developments at SAS as well asinteresting industry-based applications of statisticsand analytics.

Professor Jeff Hunter delivered an Invited Talk on

Mean First Passage Times in Markov Chains – Howbest to compute? at the 17th Applied StochasticModels and Data Analysis Conference (ASMDA2017) held at De Morgan House, the home of theLondon Mathematical Society, London, over theperiod 6-9 June 2017.

Dr Robin Hankin has created a new learningresource for students of modern physics consistingof 200 video-taped lectures discussing themathematical basis of Einstein’s theory of generalrelativity. He has also been working in thefield of formal likelihoods in competitive eventsand has submitted a paper to the Journal ofStatistical software outlining a generalisation ofBradley-Terry likelihoods as applied to AustralianMasterchef.

Dr Murray Black presented six sessions on thestatistics used to analyse data obtained fromfood tasting at the Rotary National Science andTechnology Forum in January. This forum washeld over two weeks with AUT being one of theparticipating universities. There were six groups of28 Year 12 secondary school students at the Forumfrom all around NZ sponsored by Rotary clubs.Included were six students from Australian schools.He also has been involved in an advisory capacitywith respect to the assessment of Pacific Islandersfrom Samoa, Cook Islands and Fiji in the raisingof statistical capacity in the workplace. Theseassessments form part of a national Certificate inOfficial Statistics.

Dr Kate Lee delivered an invited talk at thecomputational inverse problem workshop, 13-23June 2017, at the MATRIX, Melbourne, Australia.This workshop aims to address open challengesand recent advancements in computational inverseproblems.

14

Department of Conservation

by PAUL VAN DAM-BATES

The Department of Conservation (DOC) statisticalunit is going through major changes this year. Wehave lost 50% of our staff to retirement. Aftermany distinguished years with the Department,Ian Westbrooke has called it quits. Although theDepartment suffers a real loss without him we willcontinue to leach information from him when heisn’t too busy as the NZSA president. Luckily wewelcome Helene Thygesen as the new principalscience advisor (statistics). Helene has a PhDin biostatistics from Amsterdam and has worked

in cancer research, biosecurity and various otherareas of applied statistics. She will be attending theStatistical Ecology and Environmental Monitoring(SEEM) conference in Queenstown this Decemberand likely working out of the Hamilton office.Sophie Kincaid, a recent graduate from theUniversity of Auckland, has also joined DOC foranalysis and reporting on community based kiwicall monitoring data. As always we deal withproblems in ecology as well as recreation andhistoric data which keeps things very interesting.

Massey University Albany

by ADAM SMITH

There are currently five academic staff in statisticsat Massey University’s Albany Campus in Auckland,after the recent departure of Drs HowardEdwards and Daniel Walsh.

Prof Marti Anderson is busy working on researchfunded by a Royal Society of NZ James CookFellowship, which includes collaboration withPerry de Valpine at UC Berkeley; the researchis entitled: New multivariate statistical modelsof ecological communities: estimation, powerand prediction. Marti is also working on aMarsden-funded project (as co-PI, with Dr CliveRoberts of Te Papa) on the phylogenetic andfunctional diversity of New Zealand’s marine fishes.Marti, along with Matthew Pawley, Adam Smith,and some other colleagues and students, presentedat the International Temperate Reefs Symposiumin Pisa, Italy, in July 2016.

Dr Barry McDonald took part in the Mathematicsin Industry Study Group Workshop in July 2016,in Kyushu and Tokyo, as part of a delegationof New Zealanders active in the Mathematics inIndustry community in New Zealand. Funding wasprovided by the Japanese Government through

Kiwinet. In August, Barry visited the Pew Researchcentre in Washington DC to study their methodsfor demographic analysis of religion data.

Dr Beatrix Jones travelled to Sardinia in June2016 to give the presentation Decoupled Shrinkageand Selection for Gaussian Graphical Models at theInternational Society for Bayesian Analysis (ISBA)meeting. She is now the editor of the ISBA Bulletin,ISBA’s quarterly newsletter. Anyone wishing towrite about New Zealand Bayesian activities forthe Bulletin should feel free to get in contact.

Dr Matthew Pawley (with co-PI Dr DavidAguirre) obtained an Unlocking Curious Mindsgrant from MBIE for a project aimed at stimulatingparticipation in STEM by Maori tamariki. Theproject taught kai tiaki (guardians) from NgatiWhatua o Kaipara to use matauranga taupuhikaiao (traditional ecological knowledge) tomonitor their kaimoana. Specifically, studentslearnt how drones, ground-penetrating radar andeDNA might be used to survey toheroa. Afollow-up hui at Massey University (Albany) taughtthe students how to analyze the spatial data, anddescribed how data mining methods might be used

15

to detect shellfish populations from drone footage.Mat is also collaborating with our computerscientists to work on methods of automaticallyclassifying individual dolphins using photographsof their fins.

Dr Adam Smith has been busy preparing for ascientific expedition to New Caledonia, Fiji, andTonga, on board the RV Braveheart (39 m) inJuly-August 2017. The trip is being organisedby Dr Tom Trnski of Auckland Museum, and

will also involve a few other Massey colleagues,including Dr David Aguirre, Dr Libby Liggins,and Ms Emma Betty. The expedition aimsto survey marine ecosystems at some of theworld’s most remote and undisturbed islandsand atolls. Adam will focus on using thelatest stereo-baited-remote-underwater-video anddiver-operated-video methods to survey sharks andother fishes, in collaboration with and funded bythe Global FinPrint project.

Massey University, Palmerston North

by JONATHAN GODFREY

For the most part life has been fairly quiet for theStatistics and Bioinformatics Group of the Instituteof Fundamental Sciences. Among the ongoingcycle of reviews and restructuring, our groupcontinues to do all the things we’ve done beforeand weather the storms; our current strategy forweathering the storm is to let Martin Hazeltonfill in as Head of Institute for the next 14 months.Perhaps the most excited (editorial licence usedthere) among us is Mark Bebbington who takesover as group leader for that time.

Geoff Jones travelled to Vietnam in June totake part in a workshop planning the next stageof a multi-country study into the endemicity ofH5N1 (bird flu) organised by the UN Food andAgricultural Organization. Jonathan Godfrey isabout to travel to The Netherlands and the USAin search of blind students willing to test out thelatest in interactive graphics designed specificallyfor blind R users. He will attend the AnnualConvention of the National Federation of the Blindin Orlando – a conference that expects to attractaround 3000 people, most of whom will be blind.

Dates for our regular outreach events are set

down for 2017. 19 teams have registered forthe M3S Massey Manawatu Maths and StatsCompetition for Year 12 students, which is beingheld on 30 June. The competition follows on fromthe Maths in Industry Workshop (MINZ) beingheld on campus that week so that the year 12students and their teachers can attend the finalproblem presentations. The Massey Maths andStats Forum and Stats Research Forum is beingheld on 24 November; only preliminary planninghas been done to date, so look out for furtherannouncements if this event is of interest to you.

Penny Bilton successfully defended her PhDthesis Tree-based Models for Poverty Estimation,supervised by Geoff Jones, Siva Ganesh andSteve Haslett, and graduated in May; NadeekaPremarathna has submitted her PhD thesis and islooking forward to her defence in July.

We still see our most recent retirees, Chin-DiewLai and Doug Stirling, albeit less frequently; weare in good spirit; and there is enough good newsfor us in the aforementioned reviews to remainoptimistic about our future as a group.

16

Plant and Food Research

by DUNCAN HEDDERLEY

We have had some comings and goings at Plant andFood. Peter Alspach, based at our Motueka site,retired in February. Peter had worked for PFR andtwo of its predecessors (DSIR and HortResearch)since 1979, and was a valued collaborator andmentor to the scientists there, to plant breedersthroughout PFR, and to the rest of the biometricsteam. Peter’s retirement left a large gap – at leastfour of us are covering bits of work he used to do.He is aiming to spend more time with his family,and continuing his love of the outdoors and theenvironment, biking and tramping.

We also bade farewell to Esther Meenkenwho popped across the road to our sister CRI,AgResearch. Esther’s most recent work includeddevelopment in uncertainty estimation applied tocrop modelling activities. This of course left someof us guessing.

On the other side of the ledger, we have justbeen joined by Maryam Alavi, who did herPhD at Auckland University, and has been doingpost-doctoral work there with the Chemistrydepartment on monitoring air quality.

The PFR biometrics team (L to R Carmel Woods,Duncan Hedderley, Mark Wohlers, Ruth Butler,Andrew McLachlan, Peter Alspach, Lindy Guo,Peter Jaksons, Linley Jesson, Kate Richards)

17

StatsNZ

by RICHARD PENNY

It has come to our attention that it is a very longtime since you had a report from Statistics NewZealand, so there is a lot of news.

Firstly, as those of you that consult our web pages(and that should be all of you reading this) willknow our branding has changed. We now useStatsNZ as our moniker. While orange is our newblack we can assure you we aren’t now competingwith Serco. Our task remains as ever, producing arange of numbers to inform.

The biggest news is the November 2016earthquake. Like after the earthquake in2011, StatsNZ managed to continue to producescheduled outputs with little or no delay, a creditto the dedication of staff across the 3 centres wherewe are based. As many of you know the earthquakeseverely damaged our Wellington building. Infact it’s so bad there is a real possibility that verylittle inside can be retrieved. At least those inChristchurch in 2011 got many of their possessionsout. We are now in temporary digs in Wellingtonwith the meeting rooms in a separate building.Those of you who know Wellington will appreciatewhy getting to a meeting is sometimes describedby my colleagues as ‘bracing’, ‘challenging’ and afew other words I cannot put to print. Also ourAuckland office has moved to the New ZealandMint building on Greys Avenue.

As there have been a lot of staff changesin the Statistical Methods area since our lastreport I thought it would be useful to listall the statisticians we have as we don’t listpeople in StatsNZ on our website. As I amoften asked by academics about former students,and vice versa, I got people to list wherethey got their degree, though there is somenonresponse (probably MAR). Drum roll please.In SM as I write are: Aaron Beck, AbbyMorgan (Canty), Alison Livingston (Massey),Alistair Ramsden (Auck), Allyson Seyb (Otago),Anapapa Mulitalo, Andrew Black (VUW),Andrew Richens (Canty), Anna Lin (Canty),Anna MacDonald, Antony Gomez (VUW), CalebMoses, Chen Chen, Chris Bryant (Canty), ChrisHansen, Christine Bycroft (Otago), Felibel

Zabala (Auck, Philippines), Frances Krsinich,Gareth Minshall, Ho Chang Choi (Canty),Jamas Enright (VUW, Massey), Jeni Darnbrough,Joe Winton, John Crequer (Canty, Lincoln),Julia Hall (Otago), Larry Zhang (Regina),Lena Rodnyanskiy (Canty), Lilian Morrison(Massey), Lyco Wen, Matthew Brown, MichelleFeyen (Otago), Naeimeh Abi (Canty), NancyWang (Otago, VUW), Natalie Mawson (VUW),Nathaniel Matheson-Dunning (Auck), PatrickGraham (Massey, VUW), Penny Barber, PeterQiu (Canty), Rebecca Green (Massey), RichardPenny (Canty), Rochelle Morgan, RosaliaRohwer (Canty), Ryan Buchanan, SharonSnelgrove (VUW), Sini Miller (Lincoln, Turku),Stephen Challands, Susmita Das (Virginia -Charlottesville), Temaleti Tupou, Thanh Nguyen,Tianying Chu, Tracey Savage (Canty), VicDuoba, Vinayak [Vinny] Anand-Kumar (Auck),Vince Galvin, Wendy Dobson (Bath), WilmaMolano (Philippines). As you can see we are abig group of statisticians.

SM people have been to a fair number ofconferences recently. Allyson Seyb was on theorganizing committee for the International TotalSurvey Error Workshop (ITSEW16) in Octoberlast year in Sydney. She attended along withPatrick, Felibel and Christine. They presented,chaired and were discussants for various sessions.Nathaniel has recently presented a paper atthe recent New Technologies and Techniques forofficial statistics (NTTS17) in Brussels. Coming upis the ISI World Statistical Congress in Marrakech,where Tracey is organizing a Special TopicSession on Strategies and methods for dealingwith population over-coverage in administrativedata, and Patrick is presenting a paper in thesession. Locally we have a large presence at the2017 PANZ conference in Christchurch as 2018 isCensus year and we have major developments forfuture Censuses to report on. Christine, Anna andNathaniel are presenting and Ho-chang, Megan,Aaron and Sini are attending that one. AlsoVinny is presenting at the NZAE conference inWellington.

18

Recently published are two StatsNZ WorkingPapers, Accurate calculation of a Gini index usingSAS and R by Vic Duoba and Nairn MacGibbon,and Estimating infant mortality by ethnicity -New methods for dealing with inconsistent ethnicreporting and small numbers by John Bryant andAnne Howard. Both can be downloaded fromthe StatsNZ website, a good source for data,information, news and jobs at StatsNZ.

StatsNZ are developing a new Business DataCollection, which will provide quarterly indicatorsof sales, purchases, salaries and wages, and profitfor businesses across most industry sectors inNew Zealand. In March this year we released,for the first time, the initial series from thenew collection on the StatsNZ Innovation Site.These data are produced using an ‘administrativedata first’ approach, to reduce the burden onour respondents. In the long term, thesedata will feed in to a new quarterly income

measure of GDP (GDPI). We are releasingthese data on an ‘experimental’ basis to getfeedback on the approach and methodologywe’re adopting. For more details please contactchen.chen@stats.govt.nz. As part of StatsNZ’svision to unleash the power of data to changelives, data from the Survey of Family, Income andEmployment was made available to researchersat the Victoria University of Wellington’s Institutefor Governance and Policy Studies. The aim ofthe research was to track individuals’ earnings,assets, and wealth between 2002 and 2010. MaxRashbrooke, his father Geoff Rashbrooke, andStatsNZ analyst Wilma Molano looked into thedata to extrapolate a nationwide snapshot ofindividuals’ wealth. Their findings suggest limitsto social mobility.

For more on their findings, see the Stuff articleMost people have few assets and less than $10,000cash in the bank: new research.

Statistics Research Associates

by ROBERT DAVIES

Statistics Research Associates is a small statisticsresearch and consulting company originally setup in 1999 by statisticians formerly with VictoriaUniversity and DSIR.

The main news is that David Harte has rejoined us2 days a week. He is still at GNS 3 days a week. Heintends to continue research on self-exciting pointprocesses in the two days with us but is preparedto give the occasional course on R.

John Maindonald has resumed work in whichhe was involved more than two decades agowhen he was employed by DSIR, finding newinsights that come with time and new tools. Oversummer, he helped in the supervision of a studentwho was employed by Plant and Food to collateand run summary analyses on all data availableinternationally on cold treatment disinfestationfor four fruit fly species. Another interestingproject, again with Plant and Food, concerns theelectrical properties of fruit and other organicproduce. In his spare time he is giving a course

on "Thinking Critically about Data" for VUWContinuing Education and is a moderator at thisyear’s Maths in Industry week.

Alistair Gray works mainly for governmentagencies. Currently he is working with NRB Ltdon the design and implementation of the secondnational panel survey of marine recreationalfishers commissioned by the Ministry of PrimaryIndustries. This is a mixed mode household surveywith fishers reporting weekly over the comingfishing year.

Peter Thomson continues his involvement withNIWA (contract research) and the VUW School ofEconomics and Finance (adjunct Professor).

Robert Davies is doing occasional work for OpusInternational Consultants, currently on the qualityof aggregate used in road construction.

John celebrated a significant birthday this year.Robert and David also celebrated significantbirthdays, but of lesser significance than John’s.

19

Department of Statistics, The University of Auckland

by ATAKOHU MIDDLETON

The Ross Ihaka Lecture Series– a new, annual event

In March, to mark Ross Ihaka’s impendingretirement, we inaugurated the annual Ross IhakaLecture Series – four Wednesday evening lecturesover a month. They were tremendously successful,attracting large and youthful audiences. Watch thespeakers online: R wizard Hadley Wickham, NewZealand Herald digital editor Harkanwal Singh,Rice University’s Genevera Allen, and Ross Ihakahimself.

New faces

Professor Judi Hewitt has joined our team.She is a soft-sediment benthic ecologist and aprincipal scientist in marine ecology at NIWA,the National Institute of Water and AtmosphericResearch. Her interests include coastal andestuarine marine ecology; biological and ecologicalmapping; ecological impact assessment and designand implementation of ecological monitoringprogrammes. Judi will be embedded in theDepartment every Friday, and we are lookingforward to the new ideas and collaborations thather presence will bring. She is a recently electedFellow of the Royal Society of New Zealand and thefirst woman to be a Professor in our Department.Judi’s email here is jhew020@aucklanduni.ac.nz.

Dr Binyamin Oz from Israel has joined theDepartment of Statistics as a Postdoctoral ResearchFellow to work alongside Associate Professor IlzeZiedins. He says poetry exists in modellingstrategic behaviour. “You take some abstract ideaor a message you want to deliver, and you doit with a minimal amount of words (in poetry)

or mathematical symbols and expressions (inmodelling). You always try to do it in the mostelegant and concise way that still captures yourabstract message.” More here.

A familiar face returns:Dr Ben Stevenson has returned to us afterdoctoral study at Scotland’s University of StAndrews. Ben, who did his BSc(Hons) and MScdegrees in the Department of Statistics, says hedidn’t need to think twice when he was offered ajob working alongside the department’s AssociateProfessor Rachel Fewster. Details here.

Fighting serious disease with statisticsDr Yalu Wen, who won last year’s Worsley EarlyCareer Award, has described how seeing two of hergrandparents die of cancer put her on her careerpath. More here.

20

What Kiwi kids are thinking, feeling and doing

CensusAtSchool/TataurangaKiTeKura 2017 isunderway, with 331 teachers from 694 schoolsregistered at the time of writing. So far, the datagathered has led to three media releases, all ofwhich have enjoyed plenty of media attention.Tuck shops are losing out to home-packed lunches,and it appears that most school kids have noscreen-time limits. CAS has also found out that sixin 10 schoolchildren reported that they receivedpocket money, an allowance or a cash gift the weekbefore participating, getting a median of $15.

Departures and retirements

David Scott retired at the end of semester oneafter 22 years in the Department of Statistics. Buthe’s still doing consulting, as well as tweaking theR packages he either wrote or supports and postinghis statistics-driven rugby union and rugby league

predictions on the department’s popular blog, StatsChat.

Brian McArdle has retired. At his farewell in April,many spoke of their gratitude for his contributionto statistical ecology, the help he gave to graduatestudents, his absolutely outstanding lecturing, histremendous sense of humour, and his kindness.

Mark Holmes has taken up an AssociateProfessorship and ARC Future Fellowship at theUniversity of Melbourne. At his farewell, his quirkysense of humour shone with a performance of hisown version of Frank Sinatra’s My Way, in additionto a very gracious farewell speech. Mark hasbeen a champion for improving the mathematicalpreparedness of our incoming students, andfor giving our students a solid grounding inmathematical statistics. Both Brian and Mark willbe greatly missed.

University of Otago

Department of Mathematics andStatistics

by AUSTINA CLARK

After 52 years continuous teaching, Associate ProfJohn Harraway has formally retired from theUniversity of Otago in April this year. John willcontinue to teach in semester 2 and carry out someresearch projects. Over the years John has taughtmore than 36,000 students; his service and legacywill be long remembered.

Division of Health Sciences

by CLAIRE CAMERON

Biostatistics in the Health Sciences Division ticksalong. In Dunedin, the big news has been thatElla Iosua had her wee baby boy, Anlon, in March.A number of us attended the celebration of JohnHarraway’s retirement at the end of April. After 52years, he has almost said goodbye to the University.In May, we had a couple of people interviewedfor the position of Director of the BiostatisticsUnit (that doesn’t yet exist) and we are waitingpatiently to hear about the outcome from that.We all (across the three campuses and, sometimes,Peter Herbison in Bannockburn) continue to meetmonthly to discuss various aspects of statisticalpractice.

21

The Statistical Consulting Centre,

Department of Statistics, The University of Auckland, New Zealand,

presents the following two-day course:

 

Analysis of time-to-event (survival) data Professor Thomas Lumley

 

Tuesday 12th and Wednesday 13th September, 2017 

 

This workshop will cover data exploration, data summaries, and regression modelling for time-to-event data. There will be both lecture and practical sessions. Topics:

- Concepts: censoring, truncation, competing risks, choice of time scale - Summaries: the Kaplan--Meier curve; mean, median, and proportion surviving;

the hazard rate; graphical exploration - Two-sample testing: the logrank test and its strengths and weaknesses - The proportional hazards model: right censoring, left truncation - Time-varying predictors - Modelling recurrent events

Participants should be familiar with linear and logistic regression, and should bring a laptop with suitable statistical software. R is preferred, but assistance may be available with Stata and SAS. Instructor: Thomas Lumley is Professor of Biostatistics at the University of Auckland. He has taught survival analysis at introductory and advanced levels, and has developed software implementations and new methodology.

Registration: Registration will open early July. Regular registration fees will be available until 15th August and late registration until 8th September.

You can find registration details and other updates on: https://www.stat.auckland.ac.nz/en/about/statistical-consulting-centre/workshops1.html

For addition information email: rk.barraclough@auckland.ac.nz

22