I couldn’t avoid it any longer – on 26 May it was my turn to speak at the monthly meeting of the Canberra Branch of the Statistical Society of Australia. My topic was the same as the one I used at the Australian Statistical Conference in July 2014, but no-one minded because I had such an awful timeslot at that conference that everyone was hearing the talk for the first time! I definitely felt I had a supportive audience, who were nonetheless keen to ask questions as we went along, clarifying points about the way the data was collected (it is SO hard to avoid self-selection, informative drop-out etc etc etc when surveying students!) I was able to use the longer time allowance to cover all aspects of the lexical ambiguity study done in collaboration with Peter Dunn and Rene Hutchins, including tutor perceptions student achievement and modelling the factors that impact upon student achievement.
This book by Joop Hox was loaned to me by a research student. Hox tackles a range of topics under the umbrella of multilevel analysis, including a really instructive graphical display of the random coefficient model in chapter 2. The very range of topics included, everything from longitudinal data to multilevel path models meant that my attention waxed and waned as the topics were of more or less interest to EM. I was slightly frustrated by the presence of computer package output without clear indications of where it had come from: much of it was from HLM and MLWiN, with R receiving honourable mentions. I was pleasantly surprised to find a leaf write-up of the interpretation of Nagelkerke’s Rsquared for logistic regression – not many authors seem to bother to do that! The one weak chapter I thought was the one on modelling correlated errors with dummy variables … But it may be that this is simply another way of writing down a MANOVA.
I hope my student can bring other books that are similarly interesting and well-written. This one has a website at www.joophox.net
Sarah Helton is about to start her Honours thesis on this proposed topic, also aired on 12 May. She’s going to look into the food charities, analyse the nutrient content of their meals, and look at how that might influence health status. There’s Food Works Professional software to do the meal analysis. The bulk of her project will be qualitative rather than quantitative, but I really enjoyed hearing about the interesting projects that are out there to do in the Nutrition discipline.
A second Honours in Nutrition final presentation took place on 12 May, by Gavin Whitehouse. He had primary data from 72 team sports players and 19 athletes aged around 17. Sixty percent reported protein supplement use, mostly for muscle recovery and on average once a week. Online purchasing (nearly 1 in 5 reported this) was a particular focus of the study. His survey was validated by a U.S. Masters student, and the quantitative work he presented was purely descriptive.
I’m co-supervising Amy Shellard in this Honours project in Nutrition, and she gave her final presentation on 12 May. Gender and physical activity are the main factors Amy focused on, along with dietary intervention (a supplement and dietary advice). She presented a selection of her cross-sectional and longitudinal analyses with some nice images and use of tables. Questions were around the lack of a normal-weight group, the nature of the physical activity data collection, and the nature of the supplement. The issue of “what next” was also raised – I think it would be great to do a mixed model analysis which could provide a lot of insight. The possibility of informative drop-out is also a concern.
Professor Frank Bowden of ANU Medical School gave a fascinating talk on 24 April to be one of the first speakers. To the newly organised Health Research Institute. He opted to talk about one territory and one state in the end, beginning with his work in the 1980s on HIV in Melbourne, before moving to the Northern Territory in the early 90s to tackle some truly horrifying sexually transmitted diseases. He spoke proudly of his involvement in the elimination of Donovanosis from the NT, and encouraged us all to become involved in further efforts to rid the country of infections diseases, one disease at a time (www.1disease.org/ONI).
Assistant Professor Carlo Angelantonj of the University of Turin gave this talk on 22 April based on joint work with our ARC Research Fellow Mirian Tsulaia. Strings are one-dimensional objects that help us describe quantum gravity, unifying particle physics and Einstein’s gravity. Carlo gave a classic maths seminar, writing on the board (or wall, to be precise, as we were meeting in the INSPIRE Centre with the write-on walls). The heterotic bit is the different degrees of freedom that left and right moving modes can have. An important exercise is dimension reduction from 10 to 4 ( not 3!) but the details escaped me. Carlo made frequent use of bosons (heard of them) and fermions (not so sure about them!)
Michael Davies of UC-RISE also spoke on 27 March, with a pitch for the longest proposed thesis title! Michael and I have been working together for a number of months, and I’m very excited about the meta-analysis that is finally coming together as part of his research. I also now know how a key piece of his equipment works – an avatar is shown on a screen of your previous performances, to show your improvement or spur you on.
Kellie Toohey is a member of UC’s Research Institute for Sport and Exercise (UC-RISE). She presented the plan for her thesis in a talkfest on 27 March. She will profile the physical activity levels of a cancer survivor and contrast the effects of low-volume high intensity exercise versus moderate intensity exercise. Like so many Australian families, cancer has affected a family member, and this sort of research is so very necessary.
Q&A at the University of Canberra Research Festival on 19 March. I attended the session on “Teachers + students: why isn’t the equation adding up?” Issues arose around rote learning versus creative thinking in cultures around the world; and declining numbers in advanced maths unit and engagement in maths. Solutions revolved around drawing the links between maths and other subjects; and growing great advocates. Unfortunately I couldn’t stay till the end, but hope that something actionable will come of the meeting of minds that took place.