Research Thesis Topic
Building statistical support tools for researchers that maximise their existing domain expertise, encourage statistical learning and continually improve reproducibility.
Science relying on experiments are at risk of non-reproducible outcomes particularly associated with low power and estimated effects that are exaggerated in magnitude. The cost of designing and running experiments can lead researchers to try and minimize the number of sampling units they use. Additionally, experiments may need approval from ethics committees who must assess the risk of over or under-use of subjects while achieving the desired outcomes through the implementation of well-powered experiments.
This project aims to develop an online tool that improves reproducibility by providing scenario alternatives for researchers based on their initial parameter specifications, and educational support to help them choose the most effective form of analysis and reporting.
This will improve the confidence and statistical understanding of researchers which will assist in managing the risks associated with conducting small experimental studies and enhance the experimental protocol formed prior to data collection. This app will promote sample size estimation as a range, rather than a single number point estimate, and will encompass a scenario-based decision process for the user.
Potential research students interested in developing a project proposal in this field are invited to contact us to discuss their ideas.
- School of Agricultural and Environmental Sciences
- Statistics
- Doctor of Philosophy (DPHD)
Please review the admission requirements for the academic program associated with this Thesis Topic