SSA Canberra is delighted to announce that Mr Zhi Yang Tho from the ANU is the winner of the 2022 Dennis Trewin Prize! Zhi Yang received $1,000 for his presentation on Joint Mean and Correlation Regression Modelling for Multivariate Data.

The Dennis Trewin Prize, named after the former Australian Statistician, is an award for early career research in statistics or data science conducted within the ACT or regional NSW. This year, the Prize was modified from a video submission format to live presentations, resulting in a more engaging and exciting event. We had three brilliant presentations on the day, and the selection panel had much difficulty deciding the placement of speakers.

We would like to thank the participants for sharing their research, Dennis and the selection panel for their deliberations on the day, and the Australian Bureau of Statistics (ABS) for sponsoring the Prize.

By Nelson Chua, on behalf of the SSA Canberra council

SSA NSW Branch – October Meeting

On the evening of 26th October, Professor Brian Cullis from the University of Wollongong presented a talk titled “Optimal Design of Comparative Experiments and the ODW R Package”. 

Brian’s talk was divided into two main parts. In the first part, he gave an overview of optimal designs for multiphase experiments, particularly model-based designs using the linear mixed model with correlated treatment effects. The optimality of the designs are based on some criteria, such as D-, G-, and A-optimality. Since these designs typically involve many components, finding an optimal design is usually done through an intensive search process of the design space. As such, having computer software that can generate optimal designs given a set of requirements is desirable, which is a primary aim of the ODW R package that Brian and his collaborators are pursuing. 

In the second part, he introduced core functions of the ODW R package and demonstrated flexible uses of the package to search for optimal designs of various types. 

Finally, Brian gave several examples of when the package was used to obtain designs for an experiment regarding genotypes of desi chickpea. The seminar concluded with questions from the audience about both the technical details and applications of the ODW package.

Linh Nghiem
NSW Branch Newsletter Correspondent 

For the SA Branch September 2022 meeting, Prof Adrian Barnett (QUT) gave a presentation over zoom from Queensland; his talk a damning exposition on Bad Statistics in Medical Research. 

We began with an example of a Covid 19 study published in The Lancet, where 75% of subjects were excluded because they were still hospitalised or not confirmed as infected. Despite this major statistical flaw, the paper has been cited over 25,000 times. During the pandemic there has been pressure to publish research about Covid-19 quickly, however, it is plain to see the danger in scientists lowering the standard of their research, when the requirement for trust in science has never been so high. 

Adrian presented a number of examples of the misunderstandings of stats common in medical research, some baffling and even humorous. These included the assumption that continuous raw data must be normally distributed in order for statistical analysis to be valid; excluding outliers (when they can be the most interesting part of the data!); and an over-reliance on p-values to provide all the information about the analysis. The interpretation of p-values can also be faulty, but faulty interpretations are often accepted with the attitude of: “It's okay, since everyone else treats them this way too”. 

Next, we considered the distribution of z-values extracted from confidence intervals in a medical journal and noted that this had the shape of a normal distribution with a chunk missing in the middle - indicating that studies with a negative result (non-significant results) are often excluded from publication. Adrian went as far as to call this research misconduct. 

Adrian's suggestions for fixing the problem include:

• Abandoning p-values and statistical significance in reporting;
• Stop funding projects to make prediction models - unless a statistician is a principle investigator.

We also considered the introduction of a statistical audit, where a random selection of 100 papers per year are selected for audit by a statistician, to check the research and also whether they can reproduce the results. 

A statistical robot might also be put to use: an algorithm that can flag potential statistical problems in a paper and possibly detect fraudulent results. 

The presentation also included a number of lamentable figures from real publications, such as pie charts that distort the data, and a 3D bar chart that resembled pieces of fudge. 

The presentation ended with a lively discussion, with many weighing in on the problem of bad stats and how we might fix it. In response to the question, if we don't rely on p-values, then what do we use? Adrian conceded that sometimes results are complicated, but ultimately science is hard - and we should celebrate that.

By Annie Conway

Nominations are now being accepted for the 2022 Tjanpi Award, the annual student prize for best student paper in environmental statistics, sponsored by the SSA Environmental Statistics Section.  To be eligible a student must be:

• An author of a paper that has been accepted in the previous 12 months, having made a substantial contribution to the work
•  A student as of June 30 2022
• A current member of the SSA and the Environmental Statistics Section

The winner will receive $500 and will be asked to present in an invited session on environmental statistics at the next annual stats conference (in Wollongong, 2023).

Please submit your nominations to eo@statsoc.org.au, with Tjanpi Award submission in the header, by 5 PM AEDT, Thursday 8 December 2022, including:

- Full name, institution

- Paper, as one pdf file.

- Letter of support from supervisor or other academic at the institution, confirming student status of applicant and describing the student's role in the paper.

Image: Central Australian landscape dominated by Tjanpi, photo by Sara Winter 

Tjanpi is the Pitjantjatjara word for Triodia, a spiny tussock-forming grass that dominates the vegetation across more than 20% of Australia’s land mass.  It is a long-lived plant that makes deep roots and can withstand the hardiest of conditions.  It can grow over decades into characteristic ring formations three metres in diameter.  As a source of food and shelter, Tjanpi is fundamental to life in some of Australia’s most extreme conditions, being central to highly diverse ecosystems dominated by termites and ants, as well as reptiles, birds and small mammals.  It has also been traditionally used by Indigenous people for a range of purposes, including building shelters, making an adhesive resin, basket weaving, fishing and using its seeds as a food source. 

Tjanpi is an analogy for the Environmental Statistics student award – because the development and application of appropriate statistical techniques is fundamental to good environmental research, and our hope is that the recipient of this award will grow over the coming decades to become central to a diverse range of interesting research endeavours!

Would you like to become a committee member for the SSA Mentoring Program?

The SSA mentoring committee has begun planning the third year of its successful mentoring program. The program provides an opportunity for emerging statisticians to develop personal and professional skills, as well as providing connections between statisticians from across the nation.

This wonderful initiative is developed and managed by the SSA Mentoring Committee (pictured below). We are a team of statisticians eager to increase the support available to our community for success in the workplace. The committee is dedicated to delivering a program that meets the needs of its diverse range of mentees. To do this, the committee is seeking expertise, experience and perspectives to ensure diverse representation of the statistics community. Do you have the fresh new voice we’re looking for? Please get in touch!

To express your interest in joining the SSA Mentoring Committee, or for more information, please email us via ssa.mentoring@gmail.com with a couple of paragraphs about yourself and why you are keen to get involved.

In September, the NSW Branch hosted four industry speakers and approximately 50 students at the Courtyard Cafe at the University of Sydney. Despite some drizzles on the day, the event was well-attended and had a vibrant atmosphere. This event was partially sponsored by the Sydney University Postgraduate Representative Association (SUPRA). 

The event began with an acknowledgement of the country by our Young Statistician Representative, Yidi Yan. The President of SUPRA, Yuning Zhang, and the Vice President of NSW Branch Dr Thomas Fung, introduced the audience to their respective organisations and encouraged the students to stay in touch with some upcoming events which will offer great benefits to the students. I had the pleasure of introducing each of our four great panellists. 

The first panellist was Rachel Ragell from Lion. Rachel presented two case studies in her talk, one on greenhouse gas emission reduction and the other on beer demand forecast in the post-COVID recovery. These case studies allowed the audience to get a glimpse into how statistics can be applied to widely different industry projects. Rachel ended her talk with some valuable career interview tips for the students. 

The second panellist was Dr Stanislaus (Stani) Stadlmann from QBE Insurance. Stani presented his career journey which began in Gottingen, Germany but ultimately brought him to Australia. Through his own experience, Stani shared with us how he planned his career in statistics and data science and reminded us of the importance of finding our own passions and empathy for others. 

The third panellist was Dr Nancy Briggs from UNSW Stats Central consulting unit. Even though UNSW is an academic institution, Nancy had previously worked in various industries and loved talking to people about their statistical challenges. Statisticians often have the luxury of working on many different projects and Nancy is certainly one of them! She has done work on projects in psychology, image object detection and cancer therapies. These projects highlight the importance of statistics as a powerful tool in scientific discoveries.

Our last panellist was Dr Earo Wang from Canva. She connected with the audience immediately because the company is well-known in the student community as the go-to tool for visual design. Earo introduced her role at Canva as a data analyst working on website utilisation forecaster. Such a role seemed perfect for Earo as her doctoral training was in the area of time series forecasting. 

After each of our panellists have introduced themselves, the event proceeded with a networking session so the eager students can have the opportunity to chat with our panellists. The session is well-supported by a catering of pizzas and beers and as far as I can tell, everyone at the event had a great time and got important career-related questions answered. 

Kevin Wang
Councillor of the SSA NSW Branch

Science and Technology Australia (STA) will hold “Science meets Parliament” (SmP) again next year. This event provides the opportunity for scientists and technologists to interact with the Government's key decision makers. SmP 2023 comes at a time when science, technology, engineering and mathematics are high on the political agenda. SmP 2023 will build on the strong tradition of fostering relationships and understanding between scientists and technologists, MPs and Senators.

As a member of STA, the Statistical Society of Australia (SSA) is invited to be represented at SmP. This is a unique development opportunity which SSA would like to give to one of its "Early Career or Student Statisticians”. Our definition of an "Early Career or Student Statistician” (ECSS) is a person enrolled for a degree who is studying either full-time or part-time without age limit, OR a person who graduated with a Bachelor's degree within the past five years, OR a person awarded a postgraduate degree within the past year. The Society will pay the registration fee and reasonable travel expenses for our representative at SmP.

Next year's event consists of two parts: SmP will start with the opportunity for high quality professional development, delivered on a virtual platform across three days from 7-9 March 2023 during business hours. Sessions will also be recorded and available for two weeks after the event.

Part II will take place in person either on 15 March or 22 March 2023, depending on next year’s Parliamentary Sitting Calendar. The day on the “Hill” will feature:

• Meetings between SMP Delegates and Members of Parliament (meetings for STEM professionals only)

• Additional training and personal development sessions

• Attendance at Question Time

• National Press Club Address

• National Gala Dinner in the Great Hall at Parliament House

In 2022, SmP organised 64 meetings with scientists and Parliamentarians. It was attended by 528 delegates and featured 68 speakers. Attendees of previous SmPs described the event as “an awesome primer”, “a very useful opportunity to engage with important political stakeholders” and “an excellent experience to develop skills and network at the highest levels”.

If you are a Society member who meets the ECSS criteria and would like to attend SmP, please email your informal application with your CV to the Executive Officer (eo@statsoc.org.au) before Thursday, 3 November 2022. Your application email should include a little bit about yourself, why you would like to attend SmP, how attendance could potentially enhance your career, and what you as a representative of SSA can contribute to SmP.

The successful applicant is expected to provide a report about their SmP experience by Thursday, 6 April 2023. This report will be published in the SSA newsletter and on the SSA website.

More information about SmP is available here.