EPPS 2302 Chapter Notes - Chapter 1: Confounding, Response Bias
Document Summary
Confounding - effects of 2 or more individual variables that are not separated. Lurking - individual variable that was not considered but affects the dependent variable. Cross-sectional studies - observation about individuals at a specific time or over short period of time. Case control - individual looks back in time or researcher looks at previous records. (retrospective) Cohort - identifies a group of individuals to participate in the study and are observed over a long time. Some drop out later in the process and disrupt the data process. (prospective - future looking) Stratified sample - separating the population into nonoverlapping groups called strata. Systematic sample - obtained by selecting every kth person from population. Random number between 1 and k = p. Sample: p, p + k, p + 2k, p + 3k, p + (n-1)k. Samples individual tendency to favor one part of population.