The 23rd Annual JB Douglas Postgraduate Awards Day was held on Monday, 5 December 2022 at the UTS Aerial Function Centre and was a great success. In the afternoon, 6 PhD students presented their theses, showcasing the depth, breadth and excellence of research performed by postgraduate research students in statistics within the NSW state. The joint winners of this year’s award are Emily Neo from the University of Sydney Business School and Chris Lisle from the University of Wollongong.
Emily presented a work on partial correlation screening, which is important in dimension reduction techniques for high-dimensional data achieving uncertainty quantification in situations where it is usually difficult to have it. She presented a technique that is defined in the field of statistics but can compare with machine learning approaches. Her presentation was clear, complete and she showed a great ability in answering the many questions she received.
Chris was able to introduce the problem in a clear way, we like he told the audience the story of the project and its context. The project he presented had a clear statistical interpretation, that means his diagnostics can be easily implemented in the field. He acknowledged his work as part of a team with different expertise and this is at the heart of applied statistics.
Each winner received a prize of $500.
After the students’ presentations, Professor Marijka Batterham from the University of Wollongong delivered the NSW Annual Lecture, titled “Encouraging Statistical and Data Literacy in Nutrition and Dietetics.” Marijka started her talk by giving an overview of how dietitians are usually trained. While most dietitians are highly specialised in some areas (such as clinical, food service, public health, sport, etc.), all are required to do introductory statistics to meet competency standards. Indeed, statistical literacy plays a vital role for dietitians to evaluate the evidence, make clinical decisions, and participate in nutrition research.
In the next part of the presentation, Marijka presented numerous examples demonstrating challenges of research on human nutrition, including missing data, compliance, and agreements between measurements, as well as some possible statistical solutions to tackle these issues. These challenges open up opportunities for many novel applications of statistical methods, ranging from classical statistics to modern machine learning, to address questions regarding nutrition and dietary.
Nevertheless, in her ongoing research, Marijka demonstrated there was generally a statistics anxiety among undergraduate health science students, many of whom found statistics useful but the content “incredibly hard” or “overwhelming.” Furthermore, lack of statistical support, including access to statistical software, can be a major barrier for a dietitian’s literacy. While free software for comprehensive data manipulation, visualisation and analysis, such as R and Python, are available, they require steep learning curves and may not be suitable to clinicians, who do not use statistics regularly and only often need some basic descriptive statistics and insights to ask the right questions to team members with a background in biostatistics. A current solution to this need is free point and click graphical user interfaces to R; nevertheless, their functionalities can still be limited or their interfaces are not so user-friendly to beginners. Marijka pointed out that computational skills still need to be improved in clinical disciplines, especially as we embrace reproducibility and more advanced computational methods.
The event concluded with the dinner and fun trivia. On behalf of the NSW Branch Council, the branch’s president Clara Grazian thanked everyone for attending and supporting the JB Douglas Awards Day as well as all the events of the branch this year.
We look forward to seeing you at our events in 2023!