Portfolio
Fifty Shades of Medical Research
This presentation outlines the key components of good research practices. Here I emphasized the importance of reproducibility of research and ethical methods. Moreover, I discussed common misconceptions, such as the hierarchy of study designs, Simpson’s paradox, p-hacking, and stressing the difference between statistical and public health significance. View this presentation here!
Screening Test Evaluation Web App
This web-based app works on the platform of Shiny application, which facilitates the evaluation of screening tools by calculating various diagnostic efficiency statistics. The app provides a user-friendly interface and is useful for researchers, healthcare professionals, or anyone interested in assessing the effectiveness of a screening test in a quantitative manner. The simplicity of the interface makes it easy to use, even for those without extensive statistical knowledge. Check out the app here!
GWR analysis: A statistical method to account for spatial heterogeneity
In this study we introduced geographically weighted regression as a statistical method to account for spatial heterogeneity in data analysis. Unlike ordinary linear regression (OLR), which assumes a global association across the study area, GWR acknowledges and incorporates local variations in associations, providing valuable insights for policymakers to design targeted interventions based on localized data patterns. Continue reading
The Good, The Bad & The Ugly of Medical Research
In this session, I discussed the importance of medical research and highlights both its positive and negative aspects. I emphasized that good research should be reproducible with validated scientific, statistical, and ethical methods. I also introduced the “Four Horsemen of Research Apocalypse,” which includes issues like ghost and guest authorship, research fraud, predatory journals, and salami slicing.