Statistical Sciences 2244A/B Chapter Notes - Chapter 7-8: Confounding, Longitudinal Study
30 Sep 2016
School
Department
Professor
Document Summary
Observational study: observes but does not influence response. Experiment: deliberately changes a variable to see the change in response. Confounding: effects of 2 different variable cannot be distinguished from each other. Convenience sample: choosing the people closest at hand - not a good sampling design. Voluntary response sample: lets people chose if they want to participate - also not a good sampling design. Everyone in the population has an equal chance to be selected. First divide the population into major groups (with specific characteristics) then use srs. Sample survey/cross-sectional studies: observational study randomly sampling the entire population at one point in time. Undercoverage: some groups of the population are left out of the sampling process. Case-control studies: random sample of the individuals with the condition (case) is compared with a random sample of individuals without the condition (control). Often uses a retrospective approach (looking into the past) to see expose factors that differ within the two groups.
