The Biostatistics, Epidemiology, and Research Design (BERD) team attended the Conference on Statistical Practice virtually in February 2021. Our team benefited from short courses that were offered, with topics ranging from addressing missing values, mixed models to casual inference. We also gained new tools on visualizing and addressing missing data, identifying sources of variability using common forms of dependence among observations in mixed model designs and accounting for uncontrolled bias in non-randomized studies.
Our team attended conference sessions on diverse topics. We were particularly inspired by sessions offering current recommendations for moving beyond p-values and statistical significance in presenting statistical results. We will be sharing more about about P-values in our upcoming webinar on April 15, 2021.
Additionally, several sessions addressed the statistical needs unique to COVID-19, such as the ethical and the technical strain placed on epidemiologists charged with designing research studies, and statisticians tasked with analyzing data and producing results in a short time frame. Sessions on effective communication promoted the concept of shared knowledge and highlighted the importance of both the client and the statistician knowing and acknowledging their shared knowledge. In the session about collaborative grant writing, panelists discussed the importance of allocating time for biostatisticians to learn about the entire research to obtain best advice and providing evidence of sufficient commitment from the biostatisticians to increase chances of grant approval for research proposals.