Adriano Polpo, Associate Professor at UWA, spoke in the June meeting about Optimal Sample-Size-Dependent Significance Levels. The talk was based on the work (DOI: 10.1080/00031305.2018.1518268) published on the The American Statistician special issue about “Statistical Inference in the 21st Century: A World Beyond p < 0.05”. It can be seen that this is a hot topic given the number of publications in this issue is over 40. But Adriano and his coauthors address or revisit the nuances of hypothesis testing in the classical approaches in order to prepare the way for their approach to what has recently been a vexed question involving whether or not or how to use p-values?

For those who do not know Adriano, he has recently taken up his position at UWA having come from the Federal University of Sao Carlos, Brazil, where he spent 12 years. He originally got his PhD from the University of Sao Paulo in the year 2005 and has also taken post-doctoral study at Florida State University.  He was born in Sao Paulo, Brazil, but also has dual citizenship with Italy.  On the other hand, he does not confess to speaking Italian. Interestingly he tells me that he learnt English by using the computer.

The statisticians know of the inconsistency, or paradox, in the current classical tests of significance that are based on p-value statistics that are compared to the canonical significance levels (10%, 5%, and 1%). Adriano and his colleagues argued that researchers do not need to completely abandon the p-value, rather, they should instead stop using significance levels that do not depend on sample sizes. A testing procedure was presented, with a significance level that is a function of sample size, obtained from a generalized form of the Neyman–Pearson Lemma.

At the conclusion of his talk several people dined afterwards and continued discussions at the  Bateman Chinese Malaysian Eating House.

Brenton Clarke

On the 12^th November 2019 Cathryn Lewis who is Professor of Genetic Epidemiology and Statistics at Kings College London, University of London received her Frank Hansford-Miller Fellowship medal, and led the audience in a tour through some of her research in a talk titled “Polygenic risk scores to predict risk of disease, hype, hope and statistical reality”. Cathryn has been working at her present position for 20 years and describes herself as someone who works on applied statistics.  This is very fitting as Frank was very enthusiastic for applied statistics and the benefit that it could bring to the whole of society.   Cathryn trained as a mathematical statistician, receiving a BA from Oxford, then a Masters and PhD from the University of Sheffield, before moving to do a postdoctoral work at the University of Utah, Salt Lake City, USA. Here she was involved with discovering the link between the BRCA1 gene and the risk of breast cancer, which has motivated a lot of interest in genetics and risk ever since. She returned to England taking up her current position which she has worked at apart from short sojourns off for having children.  

Cathryn noted that while she was in London reading about Frank she was struck by the quote “He was a rather eccentric Englishman who adopted Australia as his home after he retired from the UK”. On arriving in Perth, she could now see exactly why he had done this and volunteered to say that if we have any tutorial posts in 10 years time she would certainly be interested.

The synopsis of Cathryn Lewis’ talk started with an introduction to genetics and polygenic risk scores, analysis methods, challenges, and a return to quote Frank Hansford-Miller.

She introduced the discovery of DNA and highlighted the work of Rosalyn Franklin, somewhat overlooked when discussing the discovery of DNA in 1953 by Crick and Watson.

She then talked about why a statistician may be interested in DNA by alluding to the fact that there were 3 billion base pairs in the DNA which would take us 50 years to type out one letter at a time, but pointed out there were 0.1 % of these pairs distinguished us as individuals.  While 0.1% is small this percentage of 3 billion is quite an amount of data.  She went on to discuss the link or hypothesis that various alleles could lead to increased risk of disorders such as depression.  The genetics technology now looks at sites across half a million sites at one time examining the whole genome at one time.  There are matrices of data, and each row is examined at one row at a time, and individuals are examined for cases and controls.  There are adjustments for ancestry. Essentially for each SNP a logistic regression is carried out so that one gets an odds ratio for a particular increased risk of a particular allele.  The studies involve P-values which are adjusted for the large numbers of tests carried out.

In comparison to other disorders “hits” due to particular gene types only began for depression in 2015.  Risks on a polygenic scale now lead to risk scores for every individual that give you information on risk of the disorder.  Methods involve weighted scores across all SNP’s..  There are considerations of correlation and independence

Results involve scores that have roughly a normal distribution, and then interest is in the tails.

 How do we predict outcomes for cases and controls?

We were led through various statistical methodologies of Cohen’s D and Nagelkerke’s R.

Depression has a prevalence disorder of 15%.    Some discussion ensued about course of illness.

A discussion of risk in other areas put the ideas of screening for various disorders in perspective, and it was pointed out that predictions of disorder were not the same for African as opposed to European cases whereupon ancestry is important.

There is still a need to increase the predictive ability.  We have gone from a yes/no genetic description to a polygenic risk on a normal scale.

To remind us of the aptness of her work Cathryn pointed to a quote from

Frank “We have been become too technical and should ask “How is this piece or research going to help humanity””

and challenged us all to play a part in making statistics relevant to society.  She closed her talk with a picture of the double helix in the DNA tower at Kings Park.

Brenton R Clarke

Following the initial round of public consultations, the Australian Research Council (ARC), Australian Bureau of Statistics (ABS), New Zealand Ministry of Business, Innovation and Employment (MBIE) and Stats NZ have now published the ANZSRC Review Consultation Draft. The proposed changes to both the Fields of Research (FoR) and Socio-Economic Objectives (SEO) classifications are now available for comment until 10 February 2020.

The Consultation Draft and details on how to make a submission are available on the ARC website.

The ANZSRC Review Consultation Draft proposes significant changes to both the Fields of Research (FoR) and Socio-Economic Objectives (SEO) classifications based on initial consultations. A new Division-level (‘2-digit’) classification for Aboriginal and Torres Strait Islander, Māori, and Pacific Peoples research is proposed for both the FoR classification and the SEO classification. These new Divisions would allow for greater recognition and visibility of this important and unique sector of the Australian and New Zealand research landscape.

Other proposed changes to the FoR classification include splitting Division 11 Medical and Health Sciences into two Divisions and removing Division 10 Technology by distributing its Groups among other Divisions. It is also proposed that the Sector level of the SEO classification be removed to align with the FoR hierarchy.

This draft is not final and the ANZSRC Review Steering Committee is allowing time for changes before finalisation of the new ANZSRC classification. Feedback is sought on:

• Whether the revised classifications accurately capture the current Australian and New Zealand research landscape.
• Whether any errors or ambiguities have been introduced in the drafting process.
• Concordance between the old and revised ANZSRC classifications, including

• where codes have been deleted, where would that research be classified in the revised ANZSRC?
• where new codes have been created, where would that research have been classified in ANZSRC 2008?

To assist in a balanced evaluation of the draft, submissions in support of changes are also welcomed. This will be the last opportunity for public comment on the ANZSRC draft. It is anticipated that the final updated ANZSRC will be published in mid-2020.

If you have any questions regarding the ANZSRC Review or the Consultation Draft, please contact the ARC on 02 6287 6755 or ARC-ANZSRCReview@arc.gov.au.

Kind regards,

ANZSRC Review Team

Australian Research Council I Phone: 02 6287 6755

I can't believe we are reaching the end of another year already. Where did the time go? This year has seen many things happening in the statistical community and within SSA. Below I have listed just a few:

First of all we held a record number of workshops and other events. SSA’s branches and sections truly outdid themselves with their offerings to members, organising events for the many different interest groups within the society. I noticed especially the many events for our students or early career statisticians, all aimed at creating a kick-start to a successful career in statistics.

It is the many, many volunteers at branch and section level that made this possible and we can’t thank them enough.

The Betty Allen Travel Award was launched for the first time, and we received a good number of outstanding applications, making it very difficult to make a decision.  The lucky winner will be notified in January. Then there was the inaugural Louise Ryan Best Presentation Award, awarded at YSC2019 by the wonderful Louise Ryan herself.

We now have a new ANZSTAT list on our very own website, drawing the statistical community closer together, which is fantastic.

SSA’s relationship with the Australian Bureau Statistics reached a new level with an official partnership between the two organisations being contracted and recently extended into next year! This is fantastic news indeed and opens many opportunities for strategic collaboration and funding of special projects. This partnership has also supported the International Data Science in Schools Project, which aims to create a 2 year, 240 hour course in Data Science for year 11 and 12 students, and an accompanying course to train teachers to give it, to be delivered internationally.

The SSA continues to improve its value proposition to members. We now have member discounts for Wiley, OUP, Taylor & Francis, Routledge, CRC Press, shinyapps and Significance magazine. We have created a page of videos of branch talks that are only available to members. A new webinar series started in the second half of 2019 and has proven immensely popular. One of the highlights would have to be the recent webinar with Sir David Spiegelhalter. We are delighted to be able to offer recordings of our webinars on the SSA website for our members. The link to Sir David’s webinar will be made available shortly. Please check the webinar page from time to time.

The results of the Federal Election in May took all of us by surprise and almost seven months on we are still grappling with questions about opinion polls and how they got it so wrong, but we will try to get to the bottom of this.

Let’s keep the momentum going into the next year!

Thank you for your continued support of the Statistical Society of Australia. You are the Society and you make coming to work every day a pleasure for me.

During what is for many a holiday time, let us look out for one another, keep safe and share plenty of smiles with those around us.

Very warm greetings and a happy, peaceful and prosperous new year.

Marie-Louise Rankin, 
Executive Officer, SSA

Bayes on the Beach, the 13^th biennial meeting of the Bayesian Section of SSA, was held in Surfers Paradise, Queensland, in November. It featured keynote presentations and tutorials by Prof Antonietta Mira (Università della Svizzera Italiana, Lugano, Switzerland), Prof Sudipto Banerjee (University of California, Los Angeles, USA), A/Prof Renate Meyer (University of Auckland, New Zealand), Prof Nial Friel (University College Dublin, Ireland), Prof Heikki Haario (Finnish Meteorological Institute, Finland), Prof Mark Briers (Alan Turing Institute for Data Science, UK), and Dr Jegar Pitchforth (Booking.com BV), as well as many excellent talks and posters. A particular highlight was a game of beach cricket, in memory of Prof Richard J. Boys, a previous speaker at Bayes on the Beach who sadly passed away earlier this year. Many thanks to the organising committee and to our sponsors for making this event possible, and to all attendees for their enthusiastic participation.

A game of beach cricket in memory of Richard Boys. Photo by Cheryle Blair.

The symposium on Data Science for Social Good at QUT featured a keynote presentation by Prof Sir Peter Donnelly (Genomics PLC and the University of Oxford) as well as talks on conservation, the environment, healthcare, genetics, data science for non-profits, food security, public policy, and artificial intelligence. This was followed in the evening by the launch of the QUT Centre for Data Science.

Executive Committee of the Bayesian Section of SSA

@BayesSSA

https://www.statsoc.org.au/Bayesian-Statistics

Science Meets Parliament (SmP19) – a conference run by Science & Technology Australia - has been bringing scientists and parliamentarians together since 1999. The objectives of the forum are to inform members of parliament how science can contribute to public policy, and provide insight for scientists to the political, media and parliamentary processes that govern the Commonwealth of Australia. Around 200 scientists and communicators were selected to attend this event to discuss the current and future role of science in politics. The Statistical Society of Australia (SSA) selected Carmen Lim from the University of Queensland and Janan Arslan from the University of Melbourne to represent the society at SmP19.

The two-day event (26-27^th November) was held at both Hotel Realm and Parliament House in Canberra. The first day of the conference was packed full of presentations and workshops (particularly for first-time attendees), with captivating speakers, such as Professor Fiona Wood AM (Director of Burns Service of WA & 2005 Australian of the Year) and Dr Alan Finkel AO (Australia’s Chief Scientist) taking to the podium. Day One was also a great opportunity for the 200 attendees to connect with each other, which lead to exhilarating conversations. The day ended with a Gala Dinner, which was MC’ed by SCOPE TV presenter Lee Constable. Presentations were given by The Hon Karen Andrews MP (Minister for Industry, Science, and Technology), The Hon Brendan O’Connor MP (Shadow Minister for Employment and Industry, Science, Small and Family Business), and Marina Sara (Chemical Engineer, ANSTO), who shocked our delegates by revealing we work approximately 80,000 hours in our careers. Each table at the dinner contained several scientists as well as a parliamentarian. This was our first opportunity to connect and communicate with policymakers.

L to R:
Carmen Lim and Janan Arslan enjoying a night out at SmP Gala Dinner.

L to R:
Adrian Barnett, Susanna Cramb, Carmen Lim and Janan Arslan. Statisticians storm parliament house!

The second and final day was all about meeting our assigned parliamentarians. Every scientist was assigned to a parliamentarian based on either common scientific interests or by the electorate. Carmen was paired with Dr Mike Freelander MP while Janan was paired with Dr Adam Bandt MP. Janan also attended the National Press Club Address presented by Professor Lisa Harvey-Smith – an award-winning astrophysicist and the Australian Government’s Women in STEM Ambassador - in which she discussed current gender biases and committing to bringing equity in science, engineering, and mathematics. Additional tours were made available for delegates, such as the Parliament House Geology and Beekeeping Tours. The day concluded with a parliamentary panel and discussion that, unfortunately, was short-lived when the bells began to ring throughout Parliament House. The conference was a revelation. It emphasised the importance of scientific communication, highlighted Parliamentary processes, and offered an opportunity to build strong working relationships between scientists and members of parliaments.

L to R:
[First person - Unknown], Dr Mike Freelander, Susanna Cramb, and Carmen. Carmen and fellow scientists meeting Dr Freelander to discuss their respective scientific projects.

L to R:
Tracey Ellis, Peter Baines, Dr Adam Bandt, Janan Arslan, and Amy Winship. Lots of smiles after talking about our work with Dr Adam Bandt.

We want to thank the SSA for supporting our attendance at SmP19. We thank the ever so amazing Marie-Louise Rankin and Adrian Barnett for supporting us throughout the process and guiding us during the conference. This has truly been a remarkable experience. With meeting people across all the STEM careers and parliamentarians privately in small groups, we not only had the opportunity to share our research projects with prominent members of the community but also heard the scientific journey of many others. We connected with fellow scientists and felt completely at home during the entire event. We were reminded of the core values of being a scientist: to seek the truth and make a difference in the world. The most valuable lesson that we learnt was scientists and politicians are actually driven by common goals: we all want to make a difference. We conclude this article with some of Janan’s and Carmen’s highlights and remarkable quotes from the event.

Quotes:

“You could be the best visionary in the world, but if you’re in a soundproof room, what is the point?” – Professor Fiona Wood

“Collaborating with mathematicians and statisticians helps to shape our work.”– STEM Professionals in other fields

“We do not need to fix women. We need to fix the system.” – Professor Lisa Harvey-Smith

“We have to show the politicians where our impact is.” – Dr Alan Finkel

Carmen’s highlight:

My highlight was perhaps learning how to pitch my work in a minute. We fill in existing gaps in scientific knowledge by creating solutions and that has helped me to focus my pitch when speaking with the parliamentary leaders I met in Canberra. I met Dr Mike Freelander, a paediatrician turned MP for the Labour party. I was given an opportunity to discuss the issues around my research on substance abuse in vaping and got a chance to hear his political viewpoints on this. His vision to serve his electorate was truly inspiring. I encourage all scientists to get in touch with their local MP and see how your research can be translated into impactful research. Politicians are apparently good storytellers. SSA members, please consider attending SMP in 2020. You will have an experience of a lifetime. 

Janan’s highlight:

The moment I walked into the Hotel Realm, I could immediately feel the buzz and energy in the air. As scientists, we sometimes find ourselves isolated from the world, for we are so focused on our own passions and research. You forget that there are others out there with similar aspirations and desires. Finding yourself with likeminded individuals who understand what it is like to wake up at 4 am and go to bed late at night all in the name of seeking scientific truth is refreshing, to say the least. While I consider myself to be an excellent communicator, I thoroughly enjoyed Dr Will Grant’s and Dr Rod Lamberts’ pitching workshop, in which I had to pretend to go back in time and explain a modern piece of equipment without being burnt at the stake (e.g., mobile phone). It may seem unorthodox at first, but the lesson of speaking in relatable, reassuring, and easily understandable language came through very strongly. These lessons from the first day proved to be very effective throughout the second day of the conference. I was able to have an incredible conversation with my assigned politician, Dr Adam Bandt – co-deputy leader of the Australian Greens Party - which has since led to securing further meetings with him to discuss the projects I (and my supervisor) are currently working on. I consider myself very fortunate to have attended the conference and to have been presented with such an excellent opportunity to improve myself as a scientist, expand my network of colleagues, and build a working relationship with a member of parliament. So, thank you SSA for such an amazing opportunity.

Janan Arslan & Carmen Lim

South Australian Branch SSA December 2019 Meeting

The SA Branch was pleased to welcome Distinguished Professor Marti J. Anderson, a fellow of the Royal Society of New Zealand and a recent recipient of a prestigious James Cook Fellowship, to give the tenth E.A. Cornish Memorial Lecture. Marti holds the Professorial Chair in Statistics in the New Zealand Institute for Advanced Study (NZIAS) at Massey University in Auckland. She is an ecological statistician whose research is inter-disciplinary: from ecology to mathematical statistics. Her core research is in community ecology, biodiversity, multivariate analysis, models of ecological count data, experimental design and resampling methods, with a special focus on creating new applied statistics for ecology that can yield new insights into global patterns of biodiversity.

Her talk was based on a fundamental question in ecology – How do species respond to spatial or environmental gradients? To answer this question, one needs to think about how to best model these responses. A typical response is unimodal and a couple of classical models including generalized linear models (GLM) with one or more polynomial term(s) and a bell-shaped (Gaussian) curve were discussed to fit unimodal response patterns in ecology. However, there are problems with polynomial models because they are quite constrained and can generate unrealistic predictions (even yielding negative numbers). Meanwhile, Gaussian models don’t account for asymmetry.

Instead, something more flexible like generalized additive models, e.g. splines may be considered, but these do not provide interpretable parameters. Such flexible spline-type models also have only previously been applied to binary-type data.

Marti showed examples of real data of depth gradient distributions of different fish species in the NE Pacific. She explained her first principle of modelling is to re-visit data-types commonly encountered and re-visit genuinely observed patterns in such variables along large-scale gradients. The data were very messy and contained large numbers of zero values at certain depths.

The goal was to decide on a flexible nonlinear parametric mathematical form to model the mean response of species to environmental gradients. This needs to be coupled with a suitable statistical distribution to model the error structure. Model frameworks were discussed and four mean functions were introduced: Beta (modified), Sech (modified), HOF (Huisman, Olf, Fresco) and Gaussian mixtures. Various error distributions which could account for excess zeros and overdispersion were coupled with these four mean functions and models were compared using the AICc. Marti showed an example of the best fitting model for the Shortspine thornyhead (Sebastolobus alascanus), which was the Sech function combined with a zero-inflated negative binomial (ZINB).

Visualisations of these models for multiple species simultaneously were also discussed, including overlays of mean distributions, ordered ‘floating’ distributions and ordered ‘strip’ distributions.  A further example was given using data from the Continuous Plankton Recorder (CPR) from the North Atlantic. Beautiful visualizations displayed northern shifts in the latitudinal distributions of plankton species, showing how these have changed between 1960 and 2005: a cold-water species has contracted polewards, while a warm-water species has extended its range northwards.

Developments and extensions of these models are an area of current research and include cross-validation, estimation of variation in parameters, Bayesian approaches; extension of error distributions to include linked zero-inflated models, contagious distributions, under-dispersion, etc; modelling simultaneous responses of multiple species(Y), accounting for inter-specific associations; consideration of more than one gradient (X), including interactions; and ordination of species (modes, dispersions) in environmental space.

For more information contact m.j.anderson@massey.ac.nz

Yiwen (Wendy) Li

We're pleased to report that our two-day series of workshops in Nov 2019, covering both Python and R, was a success.  Our Python workshops proved to be particularly popular. 

We started on 18 November with Machine Learning with Python, hosted jointly with Eliiza and delivered by a team of presenters from Eliiza led by Patrick Robotham.  This was the first time we have offered these workshops and they proved to be popular, with over 30 attendees.  For those who missed out, don't worry, we plan to run them again next year. There was a lot of interest particularly in image classification techniques, for which we plan to develop some more material for next time.

We continued on 19 November with a repeat of our R skills workshops, hosted jointly with MIG.  This year we were joined by Emi Tanaka from Sydney who delivered the R Markdown workshop, while Damjan Vukcevic presented on Building R packages.  We had almost 20 attendees for each one and many requests for more workshops, both at a more introductory as well as a more advanced level.

Patrick Robotham & Damjan Vukcevic

VIC Branch final event for 2019

The Victorian branch final event for 2019 was a series of short talks and panel discussion on current activities within Australia and globally to promote strengthening stats and maths through gender diversity.

The first presentation was by SSA Vice President Dr Jessica Kasza on her work to establish a Safe SSA committee and updated code of conduct to prevent and respond to harassment within the statistics community. She led a response team (females and males from all career stages) at the joint meeting of the International Society of Clinical Biostatistics and Statistical Society of Australia (Melbourne 2019), a first for the SSA conference, where delegates could report to team members any unacceptable behaviour.

Next up was Professor Claudia Czado from the Technical University of Munich who spoke about the Global Challenges for Women in Math Science program. Here awards for scientific and entrepreneurial achievements are given to female students and ECRs to promote them to continue in the mathematical sciences disciplines. Professor Jessica Purcell from Monash University then introduced the Women in Mathematical Sciences Special Interest Group (WIMSIG) which was established in 2013, and held a highly successful conference in Adelaide in 2017. WIMSIG has launched a pilot mentoring program and is running a conference at Monash University in 2020 to celebrate women in Australian Mathematical and Statistical Sciences. The last speaker was Anna Quaglieri, one of the former main organisers of R-Ladies Melbourne. The R-Ladies Melbourne meet up group has now over 1300 members, and as well as events, has an active discussion board. To finish off the evening all speakers contributed to a lively panel discussion regarding the initiatives/groups/awards that each speaker has established.

Professsor Julie Simpson