The objective of this workshop is to highlight the systematic and universal nature of the vast majority of statistical flaws and their solutions. The course aims to explain the importance of biostatistics for science reproducibility/reliability and teach good practices. The lectures are tailored to biologists and concentrating on logic thinking behind biostatistics. Particular emphasis is placed on experimental design, analysis and presentation.
- Biostatistics: A short (and digestible) introduction to biostatistics; Statistical inference: testing an hypothesis; Statistical design: power, independence of variables, randomization
-Analysis: Selection of statistical tests; Parametric vs. non-parametric tests; Regression and correlation; Multiple comparisons: ANOVA and beyond; Introduction to linear models; Repeated measures; The problem of p-values.
- Data presentation: Graphical display; Error-bars, scatter plots and box plots: the dos and don'ts; Which information to disclose?
- Hand on workshop & 'consulting' session: Critical analysis of biostatistics of published articles. The last afternoon is dedicated to common discussion on biostatistical questions and problematics related to students' research projects. Questions will be sended to the teacher some weeks before the course through a form. |