HSCI 330 Lecture Notes - Lecture 5: Observational Error, Relative Risk, Clinical Trial
8 Apr 2020
School
Department
Course
Professor
Document Summary
Bias is a measure of the extent of distortion of conclusions about an exposure-disease relationship (effect measure). True risk ratio in source population = 1. 2. This study is valid because the distribution of observed results in repeated studies is centred on the true source population effect. Distinguish the difference between validity and precision in epidemiologic studies and describe it using the terms systematic error and random error. Results of repeated studies using samples from the source population. Both study designs are valid (i. e. multiple samples using the same design give effect estimates that are centered on the true ratio in the source population). There are two types of error that make what we see in the sample different from what is true in the source population. They are systematic error (bias) and random error (precision). This is an example of a study with systematic error. Gives multiple results that are centered at 1. 4 is not valid and is biased.
