The 63rd Annual Conference of the South African Statistical Association 2022

The 63rd Annual Conference of the South African Statistical Association 2022 took place in George, Western Cape from November 28 to December 2, 2022. This event is a prominent gathering for statisticians, analytics experts, and data scientists from South Africa and beyond. The conference covered a wide range of topics in theoretical and applied statistics, highlighting the vastness and relevance of statistics in our daily lives.

One memorable aspect of the conference was the presentation by a renowned figure in the field. As the Director of the Modelling and Simulation Hub, Africa (MASHA) and an associate Professor at the University of Cape Town (UCT), she discussed her groundbreaking work on studying the spread of infectious diseases and developing effective public health policies through mathematical and statistical models. Her contributions, including the application of these models to assist the South African government in responding to the COVID-19 pandemic, showcased the vital role of statistics in healthcare decision-making at both individual and policy levels.

Another intriguing presentation focused on cancer diagnostics and early detection. A University of Washington professor shared their research, which utilized statistical models to understand population cancer trends and inform health policies, particularly in prostate cancer. By analyzing data from population registries, the study revealed valuable insights into disease natural history and overdiagnosis, ultimately guiding national guidelines for prostate cancer early detection.

During the COVID-19 pandemic, an eye-opening study conducted by a lecturer from the University of Pretoria highlighted the challenges faced by students during this period and in online classes. Many students were disengaged and emotionally drained. However, the lecturer took a proactive approach by dedicating a portion of class time to check in with students and address their emotional well-being. The study demonstrated that incorporating affective learning alongside cognitive learning significantly improved students' willingness to learn.

Additionally, the preliminary results of the 2022 Census were presented, showcasing the adoption of a digital method for population counting. The presentation explored the new digital approach to data collection and provided an overview of the current state of society as reflected in the census data.

A thought-provoking round table debate on data science, data literacy, and the future of statistics sparked lively discussions. Questions arose regarding the distinctions between statistics and data science, the perception of data science as a mere buzzword, and whether statisticians were facing an identity crisis. Even the experts at the conference were unsure about definitive answers, but it was clear that both statistics and data science have their unique roles depending on the objectives at hand. One interesting perspective from a professor emphasized that statisticians possess capabilities beyond what so-called data scientists can do. There is still no definite separating line for these two professions or rather titles since statistics is actually a science of data.

All of the presentations were incredibly impressive and provided valuable information. However, due to the nature of some presentations being more theoretical, it wouldn't be suitable for me to discuss each one in detail here. Nevertheless, I must mention my own presentation, as I had the honour to present my Masters research findings. The research compared the classification performance of the traditional statistical methods to machine learning techniques such as artificial neural networks, support vector machines, and the weighted k-nearest neighbour using various dimension sizes of balanced and unbalanced simulated data. While the literature explored the comparisons of these methods, their performance varies depending on the dataset used. This poses a challenge for researchers in determining which technique will perform best under specific conditions, and that's a challenge that this research addressed. The results of this research are due for publication soon.

Attending the conference served as a gentle reminder of the vast untapped potential in the realm of data and statistics in the library environment. While the library's current involvement mainly revolves around descriptive statistics, there is so much more that can be achieved by investing in data mining within this field. Predicting the future and allocating resources effectively could greatly benefit from such efforts. However, there seems to be some resistance and a lack of sufficient value placed on data within the library. Despite the current trend of claiming to be data-driven, there is a need to assess whether we truly embody this mindset. Embracing the opportunities presented by big data requires readiness and willingness to explore its potential in this field.