On 24^th July 2019 The South Australian branch was pleased to have Dr Murthy Mittinty present at the monthly branch meeting. Dr Mittinty is a senior lecturer in the School of Public Health at The University of Adelaide and interested in both methodological development and applications of statistical methods. He gave a talk on Quantitative Bias Analysis.

Dr. Mittinty talked about systematic and random errors in both observational or experimental data and how to handle them in observational studies.  The systematic error includes aspects such as selection bias, measurement error, confounding bias and unmeasured confounding. In a perfect randomized control trial, one can attempt to remove the biases such as measured and unmeasured confounding. Unfortunately, this is not the case with observational data , hence it is important to describe how these errors are handled. He talked about how these errors should be handled so one can predict better estimates or place more confidence in estimates. He further included three important question to answer before conducting a quantitative bias analysis: (1) when we should conduct one; (2) how we select which bias to address; and (3) how we select a method to model bias. Another question is how we interpret and present these results. He aimed to answer these questions with examples in his presentation. He talked about how to select a method to model biases and how to assign values to parameters of the bias model as well as software that can be used for conducting quantitative bias analysis. There were quite few discussions after the talk.

A dinner was held after the meeting at Jasmine Restaurant, 31 Hindmarsh Square, Adelaide.

Aarti Gulyani, SA Branch
University of South Australia

On 25 June, our members gathered for breakfast.  This was an event to connect young and old, share ideas about where statistics can take you, and also provide advice and career guidance.  Breaking with tradition for our branch, attendance was members-only and the event was held in the morning.  How did it go? 

We invited eight statisticians / data scientists from diverse backgrounds to be our mentors for the day.

Our invited mentors.  Top row, L to R: Kohleth Chia, Sandy Clarke-Errey, Harry Gielewski, Andy Kitchen.  Bottom row, L to R: Stephen Leslie, Margarita Moreno-Betancur, Dennis Trewin, James Wilson

After a brief icebreaker, we formed into small groups around each mentor.  The intimate and relaxed setting allowed members to get to know each other and hear about our mentors' experiences first-hand.

We had excellent feedback from those who attended.  They particularly liked:

●       The diversity of mentors' backgrounds, a good mix of academic and non-academic.

●       The morning scheduling, which suited many members with family responsibilities, who find our evening events harder to attend.

●       The delicious food, provided by ASRC Catering.

The discussion could have easily gone for longer, if only it weren't a workday...

We are grateful to our sponsors for the event, Eliiza and Bunnings.

Damjan Vukcevic

The SSA Biostatistics & Bioinformatics Section co-hosted their ‘Getting started in biostatistical consultancy’ workshop with the Clinical Epidemiology and Biostatistics Unit at the Murdoch Children’s Research Institute (MCRI) in Melbourne on Thursday 4^th July. This long-awaited event attracted a great amount of interest, reaching capacity shortly after the early-bird registration date; perhaps not surprising given the fantastic speakers we managed to attract to speak at this event! Sadly, the morning fog in Sydney caused some issues for our interstate participants planning to attend, although some valiantly made their way on later flights to catch up on the proceedings.

Associate Professor Susan Donath, Senior Biostatistician and Deputy Director of the Clinical Epidemiology and Biostatistics Unit at MCRI, kicked off the morning session with an overview of statistical consultancy from start to finish. Susan stressed the importance of being curious and asking many, many questions with “what is your research question?” being the most critical and often taking the longest to answer! Susan also highlighted the importance of having a network of biostatisticians to discuss problems with.

Next up was Dr Emily Karahalios, Senior Research Fellow at both Monash and Melbourne University, who discussed how to plan, organise and monitor projects. Emily stressed the importance of having face-to-face meetings at the beginning of consultancy projects to avoid confusion that can occur from emails. Emily also provided some fantastic resources for helping project planning and monitoring, noting the need to take time to plan your work and to be realistic with yourself (and others) about what can be achieved within a given timeframe.

After lunch, Professor Julie Simpson, Head of the Biostatistics Unit at Melbourne School of Population and Global Health and leader of the Biostatistics node of the Melbourne Clinical and Translational Science platform at Melbourne University, delivered a presentation on how to reach clients. As well as discussing finance models and modes of operation, Julie talked about how to build a team and work effectively with one another. She led a lively interactive session in which participants talked about the characteristics they bring to a team and what their strengths are.

Our final presenter of the day was Dr Emi Tanaka from the University of Sydney who discussed interacting with different (and difficult!) clients. Emi discussed the emotional impact that difficult clients can have on us and our work as biostatistical consultants and how we can overcome that. She noted that it is important to keep a communication trail so that all have a record of any agreements that were made.

The workshop ended with a lively panel discussion with many questions from our very engaged audience. We thought the workshop was a great success and hope the participants enjoyed it too!

Karen Lamb, Jaimi Greenslade and Sabine Braat on behalf of the Biostatistics & Bioinformatics Section

Farewell to William Dunsmuir after 25 years at UNSW

On 6th June the New South Wales branch was very pleased to be part of a milestone event held at the School of Mathematics and Statistics, University of New South Wales (UNSW). The event was to celebrate the career of Professor William Dunsmuir, who just retired after around 25 years at UNSW. William has also been a great contributor to the society and served as national president during the late 2000s.

   About 100 people attended a drinks and canapes party,followed by a testimonial event. Jake Olivier was the compere and pointed out that William spent most of Jake's eleven years at UNSW "about to retire". Head of school, Bruce Henry, spoke next and summarised William's contributions to UNSW.

   Next up was Doug Shaw who told everyone about how William led the response to the Melbourne 2008 debacle, when a conference company ran off with our registration fees and left us to pay the conference bills. Doug pointed out that after this difficult period the branches more congenial to each other.

   Then Sue Wilson (UNSW and Australian National University) headed to the podium and delivered a short talk titled  "An Irregular 45 Year Time Series". Sue told us that she and William first met at the Australian National University in 1974 when they were PhD students together. Some interesting photos of prominent Australian statisticians when they had a lot more hair than today were projected. 

   Following Sue was Feng Chen, who has been a departmental colleague with William at UNSW for the last several years. Feng talked about their joint work on generalised autoregressive conditional heteroscedastic models and sequential Monte Carlo algorithms.

   William's oldest friend at the event was David Scott of the University of Auckland, New Zealand, and he was one of the event's special guests. David told everyone about having known William since 1972 and how they hung out together at the La Trobe University campus in the early seventies getting very excited about "A Course in Probability Theory" by Kai Lai Chung.

After a few more facial hair photographs David talked about their time together at the Siromath consulting company in the mid-1980s and then becoming colleagues at Bond University in the late 1980s.

Towards the end of his presentation he shared some pictures of he and William on the Queen Charlotte Sound walk in New Zealand.

   Kylie-Anne Richards was the last of William's PhD students and she delivered a presentation titled "The Impact and Marks of Mentors". She expressed he gratitude for William helping her to achieve her dream of become an academic after starting a family. Hawke's self-exciting point processes and limit order book events were also discussed.

   The principal speaker was Richard Davis, who is chair of the Department of Statistics at Columbia University in New York City.
Richard's talk, titled "William and Me", described his close friendship and research collaborations over a period of 40 years. Richard and William first met as assistant professors at Massachusetts

Institute of Technology in the late 1970s. After William returned to Australia in 1980 they did not see much of each other for another decade. But in 1990, a time series workshop in Minnesota reunited them and led to a very successful collaborative partnership that saw them unlocking the mysteries of the moving average unit root problem followed, obviously, by GLARMA(0,1).

As is known in time series circles, this is a generalized linear autoregressive moving average process of autoregressive order 0 and moving average order 1. Later they worked on least absolute deviations, and have been continually publishing together since around 1995. Richard's bottom line was the great intuition that William brought to their research problems.

  The pre-dinner part of the Dunsmuirfest finished with William getting up and giving us a short talk tilted "Large sample theory for detecting the impact of marks in Hawkes self-exciting point processes". He told us that out of his career achievements, he gained the most pleasure from proving large sample theorems for time series. He presented a recent central limit theorem-type result and explained the steps. After the technical part, heclosed the event with some words of thanks. UNSW, the Statistical Society of Australia and family and friends were mentioned first.

He pointed out that he had worked with hundreds of people and taught thousands of students over the course of his career. Then he publicly thanked particular people for playing extra-special roles in his career and life: Bruce Brown, Niels Becker, David Scott, Pat Moran, Ted Hannan, Sue Wilson, Herman Chernoff, Doug Shaw, Robert Kohn, Richard and Cathy Tweedie, Richard and Patti Davis, and his wife, Barbara Blanche with whom he celebrates 35 years of marriage in July 2019.

Matt Wand
University of Technology Sydney

Dr Anthony Lee will be presenting a tutorial on sequential Monte Carlo methods in statistics. His visit has been partially funded by the SSA and the Australian Centre of Excellence for Mathematical and Statistical Frontiers (ACEMS). Details of times and locations are as follows:

Tuesday, 2 July 2019 at 4:30pm for refreshments (talk starts at 5)
room S301 (S Block Level 3) at QUT Gardens Point Campus, Brisbane, QLD
https://www.statsoc.org.au/event-3415826

Thursday, 18 July 2019 at 6pm (with refreshments and pizza dinner afterwards)
Science S4, 16 Rainforest Walk, Monash University, Clayton, VIC
https://www.statsoc.org.au/event-3426913

The call for abstracts has now opened for Bayes on the Beach, which will be held on the Gold Coast, QLD, from November 25-27. If you would like to present a talk or poster on any topic related to Bayesian statistics, please submit a 250 word abstract by email to bob.admin@qut.edu.au by August 16. Further details are available from the conference website, https://botb2019.wordpress.com/call-for-submissions/

Other Upcoming Conferences

The 12^th International Conference on Monte Carlo Methods and Applications (MCM 2019) will be held at the University of Technology, Sydney, NSW, from July 8-12.
http://www.mcm2019.unsw.edu.au

The 4^th conference of the East-Asian Chapter of the International Society for Bayesian Analysis (EAC-ISBA) will be held in Kobe, Japan, July 13-14, 2019.
https://eacisba2019.wordpress.com

The biennial conference on Bayesian Computation (BayesComp 2020) will be held at the University of Florida, USA, from January 7-10 next year.
http://users.stat.ufl.edu/~jhobert/BayesComp2020/Conf_Website/

The next instalment of the popular “ABC in…” series of workshops on Approximate Bayesian Computation will be held in Grenoble, France, from March 19-20, 2020.
https://sites.google.com/view/abc-in-grenoble/home

Matt Moores

On behalf of the Bayes Section of SSA

The New South Wales Branch hosted two seminar events in May, with the one that normally would be in April moved to 2nd May so that we could have a public lecture from Hadley Wickham. Thirteen days later we a seminar from Ian Renner in Parramatta.

Security was tight at a large University of Sydney auditorium for the Wickham lecture since there was a small over-subscription. Council members were employed as bouncers to check tickets and make sure we didn't contravene any fire laws. As it turned out, the auditorium was only about 72% full according to ecologist/ statistician David Warton. With a total capacity of 650 we now had an audience estimate of 468 without having to do any capture or recapture. The demographic was in the youthful direction and t-shirts with slogans such as "CAN {CODE}" were spotted. Hadley Wickham, chief scientist at R Studio, and a U.S.-based New Zealander then delivered a lecture titled "Tidy(er) data". His rules of tidy data were passed on:

1. Each variable is a column.
2. Each observation in a row.
3. Each cell is one value.

We had group exercises with data sets from the Billboard popular music charts and U.S. government agencies, where the speaker asked "What makes your uncomfortable?". Live tidy-ups were done using the tidyr package in R. In summary, it was a very successful and entertaining lecture from the Master of the Tidyverse.

On 15th May the branch caravan rolled out to a different part of Sydney - the Western Sydney University in Parramatta. Ian Renner from the University of Newcastle, New South Wales, was our speaker. Ian spoke about species distribution modelling and had a running example involving lynx populations in the Jura Mountains of France. Distinctions were made between presence-only data and multiple visits data. Inhomogeneous Poisson point process models and maximum likelihood were shown to play central roles. The least absolute shrinkage and selection operator (LASSO) was shown to lead to improved performance. Dung beetles got a mention as well - which they should given their importance for ecosystems. The main message of Ian's talk was sound conversation decisions based on data analysis, good models and statistical methodology.

Professor Matt Wand, NSW Branch

The Official Statistics Section is pleased to forward the following message from the IAOS:

The IAOS is pleased to announce the composition of its new Executive Committee (EXCO) for the period 2019-2021. In the statement below from Mr Mario Palma, IAOS President (2017-2019), you will find the full report.  

The term of the new EXCO will begin at the IAOS General Assembly, to be held at the ISI World Statistics Congress in Kuala Lumpur, 18 -23 August 2019. The current President-Elect Dr John Pullinger will take over as President and will form the new EXCO together with the selected candidates.

We wish John Pullinger and his team a very successful tenure.

Warm regards,

IAOS Executive Committee

-----------------------------------------------------------

Dear IAOS members,

I have great pleasure in sharing with you the final composition of the new IAOS Executive Committee for 2019-2021.

Following the IAOS Statutes, the 2018 Nominating Committee, in charge of selecting the candidates for posts of President-elect and four EXCO members, having one candidate for each position and taking into account a proper regional representation and gender balance as much as possible, has come up with the following list: 

Members of the EXCO:

1. President: Mr. John Pullinger (UK);

2. President-elect: Mr. Misha Belkindas (USA);

3. Member: Dr. Mohd Uzir Mahidin (Asia, second term);

4. Member: Mr. Peter Popoola (Africa, first term);

5. Member: Ms. Hasnae Fdhil (Arab Region, second term);

6. Member: Ms. Gemma Van Halderen (Oceania, second term, EXCO lead on 2022 Conference).

7. Member (ex-officio): Ms. Ada Van Krimpen (Director of the Permanent Office of the ISI).

Co-opted Members:

1. Mr. Rolando Ocampo Alcantar (Latin America and the Caribbean);

2. Mr. Jan Robert Suesser (Europe, IAOS representative to the ISI WSC);

3. Ms. Ayush Ariunzaya (Conference Advisor for the IAOS 2020 Conference in Zambia).

 

Special Invitees:

1. Dr. Kirsten West (Editor in Chief, SJIAOS, until August 2019);

2. Mr. Pieter Everaers (Editor in Chief, SJIAOS, from August 2019);

3. Ms. Teodora Brandmueller (SCORUS Chair);

4. Ms. Nancy McBeth (Special Advisor to the President);

5. Mr. Ronald Jansen (UNSD);

6. Mr. Oliver Chinganya (Programme Chair of the IAOS 2020 Conference in Zambia);

7. Mr. Kees Zeelenberg (Former Programme Chair of the IAOS-OECD 2018 Conference in Paris).

 

I would like to thank the Nominating Committee members for their work and wish the new EXCO for 2019-2021 the best of luck and success.

Mario Palma

IAOS President 2017-2019

A recipe for quantifying the impact of prevention

It was a full house for Professor Tony Blakely’s seminar on quantifying the impact of preventative health interventions, with almost 100 people in attendance. 

Tony began with the motivation behind his league tables, the 100 manila folder problem that is faced by government ministers when they need to decide on which health interventions to invest in. Next he demonstrated how they simulate, using multi-state lifetables, projections of mortality, morbidity and quality-adjusted life years (QALYs) for Australian and New Zealand populations under “business as usual”, or under an intervention of interest, such as taxation of tobacco. This simulation framework can also capture other rewards such as a reduction in health expenditure and productivity costs. He finished his presentation with a demonstration of VIVARIUM, which implements his simulation model in Python (soon available on GitHub), and the BODE league table R Shiny app for New Zealand, a visual tool comparing the impact of prevention interventions for policy makers.

Julie Simpson 

Young Statisticians talk – May 2019

Each year the WA Branch offers a scholarship worth $1,000 to a Young Statistician completing honours in that year at a Western Australian University. This year’s 2019 winner was Emily Whitney from Curtin University.  The prize comes along with a request to deliver next year’s May address to the WA Branch.

President of the WA Branch, Dr Brenton Clarke, delivers a $1,000 “digital handshake” to Emily Whitney, this year’s winner of the WA Branch Honours Scholarship.

The May meeting of the Western Australia Branch also heard talks from two young statisticians, Michael Dymock and Connor Duffin, both PhD students from the University of Western Australia. Michael talked about his current research with Group Based Trajectory Modelling with Monotonicity Constraints, which aims to better quantify the uncertainty in three-dimensional geological interpolation models and geophysical models through Bayesian inversion. Connor talked about his research on Modelling site-specific Australian daily rainfall with Bayesian mixture models which aims to model monthly rainfall across four different areas in Australia.

Group Based Trajectory Modelling with Monotonicity Constraints- Talk by Michael Dymock

Michael started his undergraduate Mathematics and Statistics degree in 2014 at the University of Western Australia, before completing his honours degree in 2018. Along with being awarded first class honours for his research project, Michael received honours scholarships from both the Statistical Society of Australia and the International Biometric Society (Australasian Branch). Through his work Michael has taken first place in both the 2017 Woodside Hackathon and the 2018 Worley Parsons Hackathon, and second place in the 2017 Visagio Hackathon.

The understanding and modelling of developmental trajectories in longitudinal data are of fundamental importance across many areas of research with applications ranging from the health and social sciences to that of marketing. Group based trajectory modelling, an application of finite mixture modelling, is often the first choice in approaching the naturally complex task of modelling these trajectories. The group-based strategy acknowledges the possibility of a heterogeneous population by fitting several groups to the data and subsequently treating each group as a distinct entity or sub-population.

In his talk, Michael explained that existing methodology for group-based trajectory modelling, implemented through the SAS procedure TRAJ, has been developed over the past two decades with the addition of numerous extensions such as the ability to jointly analyse multiple trajectories as well as the handling of missing data. However, there is no methodology currently in place to impose constraints of any kind on the trajectories, in particular, monotonicity constraints. Monotonicity constraints on polynomials play a role in data analysis, when it is known, from the underlying physical theory that the response behaves monotonically. In other words, if we know that the response must progress in the one direction (either increasing or decreasing) over the explanatory variable, it is useful to constrain the model to represent this trend accurately. However, due to a multitude of possible reasons such as data entry error and missing data, in these situations, sometimes unconstrained models fail to capture the monotone behaviour, and thus monotonicity constraints are required.

In his work, Michael implemented a new methodology for fitting group-based trajectory models with monotonicity constraints by using the Expectation Maximisation (EM) algorithm. The structure of the EM algorithm allowed him to separate the optimisation routine into two smaller optimisation sub-routines (one that computes the group membership probabilities and another that maximises the likelihood function). Furthermore, to illustrate the effectiveness of his methodology, he demonstrated the use of his implementation on a real-world example in the statistical programming language R. In his example he aimed to model the developmental trajectories of individuals' lung function, in particular, Forced Expiratory Volume measurements, over a period of approximately forty years. This example is of particular interest to us because we know from underlying theory that the response trajectories will be monotonically decreasing. However, Michael showed that this unconstrained model fails to capture the required monotone behaviour. After re-running the same analysis under monotonicity constraints, Michael was able to show that the implementation is able to effectively capture and model the required monotonic trajectories.

                                      

Modelling site-specific Australian daily rainfall with Bayesian mixture models- talk by Connor P. Duffin

Connor completed an Honours degree in Mathematics and Statistics from UWA in 2018, under the supervision of Edward Cripps. Connors research was in the field of Bayesian computational statistics, on modelling Australian daily rainfall. Having stayed in this field, Connor is currently pursuing a PhD at UWA, focussing on quantifying and explaining uncertainties in numerical oceanographic models.

Daily rainfall has a large impact on the social behaviour of human beings, and also has wide agricultural, biological, and economic effects. Being able to model location-specific daily rainfall, across the country, is therefore of utmost importance. There are three principal complexities in modelling daily rainfall at a single location: temporal evolution, zero and missing days, and extreme tail behaviour.

Connors work aims to investigate models that are able to capture these complexities across 151 individual rainfall measurement locations (sites) across Australia. Connor used the finite mixture model as a framework to model the discrete and continuous data that comprise these measurements. There were four main features of the data that Connor had to work around; the seasonality of rainfall, missing data when no measurements were taken, zero-days when there was no rainfall and fat-tails when there are days with intense rainfall. Taking this into account, Connor used the Finite mixture model incorporating temporal evolution model, which captured data with no rainfall and also for days with rainfall. The varying intensities of rainfall were captured through the gamma distribution.
Some notes:

• 1.      Dirac delta to capture zeroes.
• 2.      Mixture of gamma PDFs to capture non-zeroes.
• 3.      Gamma PDFs analogous to rainfall amounts? Difficult to interpret for K > 3 though.
• 4.      Connors approach: Use variation in K = 2, 3, 4, and 5 gamma densities to see how well the model captures tail behaviour.

Temporal evolution is incorporated through the use of a mixture-of-experts structure on the mixture weights. Connor used the Markov chain Monte Carlo to estimate the model.

The results were then analysed through posterior predictive checking, and optimal models where decided on through formal model diagnostics. Connor is eager to continue working on the model and has plans to further investigate if there are other factors that might affect the choice of the model.

  

After the talks, Michael and Connor where joined by fellow Statisticians for dinner at Tiamos restaurant, where they talked about their future plans and how eager they are to continue to optimise and implement their research on a more broader scale. 

Deneegan Subramanian

Statistical Journal of the Assocation of Official Statistics - June preprint edition out

A Pre-press edition of the June issue of the SJIAOS has been released: see https://content.iospress.com/journals/statistical-journal-of-the-iaos/Pre-press/Pre-press,

Kirsten West is retiring as editor after 5 years. Her closing editorial is at: https://content.iospress.com/download/statistical-journal-of-the-iaos/sji199002?id=statistical-journal-of-the-iaos%2Fsji199002 . The incoming editor is Pieter Everaes

Measuring Indigenous Identification

The March Issue of JIAOS is dedicated to this topic; the editor-in-chief reminds readers of past contributions by the journal and the Association on improving statistics of indigenous people. Australian contributiuons to this effort are conspicuous, from then- editor Fritz Scheuren's 2014 editorial leading with an Australian paper "Measuring Indigenous Populations Across Nations: Challenges for Methodological Alignment",by Bradley Petry and Erica Potts, to the three papers in the March issue drawing on Australian experience.

Australian contributors to the topic since 2014 include Kalinda Griffiths, UNSW; Richard Madden and Clare Coleman at the University of Sydney; Ching Choi and Len Smith, Australian National University; Ian Ring Elias, Brenda Lee, Vanessa’ Smylie, Janet Waldon, John Hodge, Felicia Schanche; University of Queensland; and Maggie Walter University of Tasmania. That Australian experience has been prominent in the international debate reflects a foundation in long collaboration between key ABS officers and academic demographers in the enumeration of the indigenous population, against the grain of national collection strategies of the time.

References:

The March Open Access issue of the SJIAOS

https://content.iospress.com/journals/statistical-journal-of-the-iaos/35/1

In particular, its guest editorial by Michele Connolly: https://content.iospress.com/download/statistical-journal-of-the-iaos/sji189008?id=statistical-journal-of-the-iaos%2Fsji189008

Anderson, I, Robson B, Connolly M, Al-Yaman F, et al., “Indigenous and tribal peoples’ health” (the Lancet-Lowitja Institute Global Collaboration), Lancet, London, England. 2016: 388(10040):131–57.

Stephen Horn, Chair Official Statistics Section