How can open science help in health research? (postponed)
PLEASE NOTE: THIS EVENT HAS BEEN POSTPONED
Public trust in science depends on both real and perceived threats to the rigor and reproducibility of its methodological process.
Hosted by the Australian Centre for Public and Population Health Research, this seminar will explore common biases and statistical errors in public health scientific literature, as well as ways to improve transparency and reliability by prioritising open science methods.
Presented by Dr Cynthia Kroeger, this seminar will discuss p-hacking, a biased process of obtaining p-values, and the methods which preclude its practice will be described. Additionally, she will explore issues that include: the “differences in nominal significance” error, overlooking regression to the mean, and the misuse of nonparametric tests in the presence of heteroscedasticity.
About Dr Cynthia Kroeger
Cynthia is a postdoctoral fellow and interdisciplinary researcher in the Charles Perkins Centre at the University of Sydney. She is also a member of the Evidence, Policy, and Influence Collaborative at the Charles Perkins Centre and the Sydney Initiative for Truth.
Her expertise includes research integrity, open science, and behavioural interventions for obesity.
Currently, Cynthia is examining biases in published public health research and the impact of trauma on the autonomic nervous and metabolic systems in people with obesity.