HLTB15H3 Lecture Notes - Lecture 3: Randomized Controlled Trial, Blood Pressure, Observational Error
Document Summary
Retrospective: using data that"s already been collected. Why do we want to know whether treatment a is better than treatment b: allows us to apply it in society clinical setting. Random error: probability that observed result is due to chance. Chance: unknown sources of variation: equally likely to bias study in either direction. Variability in the data that: cannot be explained by the design, remains after systematic error (bias) is eliminated. How do we reduce the likelihood of random error: get a larger sample size, not only good for more similarities at start of trial but also good to reduce random error. Bias: any systematic error in design, conduct or analysis of a study that results in a mistaken estimate of an exposure"s effect of the risk of disease: flaw that keeps on affecting the study. Can occur at any stage (e. g. design, data collection, analysis) Can affect internal validity of a study (selection bias, information bias, confounding)