The fairground shotgun
Random error and systematic error. We describe two types of errors, the random error due to measurement randomness and inaccuracies, and the systematic error or bias.
Science without sense...double nonsense
Evidence Based Medicine pills
Science without sense...double nonsense
Evidence Based Medicine pills
Read MoreThe fairground shotgun
Together but not in each other’s pockets
Hybrid designs The main observational hybrid designs are described: cohort nested cases and controls studies and cohort and case-control studies.
Generalizations are hazardous
Berkson’s fallacy. Berkson’s fallacy is described, which occurs when we find a spurious association between two variables due to a improper sample.
Do not be misled by the outliers
Robust scale parameters. We describe robust scale parameters, such as the median absolute deviation or the winsorized standard deviation, less susceptible to extreme values.
A simple relationship
Simple regression. The concept of regression is defined and the most used simple regression models are detailed: linear, logistic and proportional hazards regression.