Advancing public health through Bayesian thinking
#SSACAB
|
#scienceforafricafoundation
|
#wellcometrust
|
#wits
|
#stellenbosch university
|
#NIMR
|
#CUHAS
|
#PublicHealth
|
#BayesianStatistics
|
#BayesianModelling
|
#Biostatistics
|
#HealthResearch
|
#DataScienceForHealth
|
#StatisticalModelling
Professor Eustasius Musenge, Deputy Director and co-principal investigator of the Sub-Saharan Africa Consortium for Advanced Biostatistics (SSACAB) and Professor Innocent Maposa a SSACAB collaborator based at Stellenbosch University, served as facilitators at the recent “Bayesian Modelling using R” workshop in Tanzania. Hosted by the Catholic University of Health and Allied Sciences (CUHAS) in Mwanza, the workshop presented advanced statistical methods to over 30 participants from across Africa.
Bayesian statistics is a way of combining what we already know with new data to get a better, more realistic answer. Traditional statistics only looks at the data collected in a study, but Bayesian methods blend the information from the data with any useful prior knowledge.
This produces an updated estimate, known as the posterior, which reflects both past understanding and current evidence. In the course, students used R to define their prior information, run Bayesian models and generate predictions.
This matters because many real-world problems cannot be understood from limited data alone, especially in public health settings where information is often incomplete or uncertain. By incorporating prior knowledge, Bayesian methods allow researchers to make stronger, more reliable inferences, even in data-poor environments.
“This workshop highlighted how valuable Bayesian modelling has become in public-health research, especially for incorporating prior knowledge and working effectively with complex or uncertain data. It showed me how Bayesian methods can strengthen conventional models and produce more reliable estimates. I am excited to build on this training in my postdoctoral work, particularly by applying these techniques to longitudinal and hierarchical data to answer complex public-health questions with greater precision and relevance,” said Dr Neema Mosha, Post-Doctoral Fellow in Biostatistics.