This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Reduced by: • Using objective, prospective, primary data • Blocking over time and batch and randomizing allocation within batch • Blinding of observer doing exams, running tests, reading films • Increasing objectivity of measurements, validating reliability of subjective scales Reduced by: • Increasing study size (4 fold increase halves imprecision) • Reducing measuring imprecision • Replicating imprecise measurements • Blocking, stratifying, pairing, or pre-post design may improve • Reducing subject heterogeneity • Adjusting initial size for potential loss (~20%) Confounding Bias Reduced by: • Using random assignment. • Using a concurrent control group • Restriction or matching of subjects on known confounders • Analytic control of known confounders • Minimizing differential losses to followup Systematic Bias Random Imprecision Sampling / Selection Bias Study Design Study Size Measurement Bias...
View Full Document
This note was uploaded on 01/27/2012 for the course VMS 576 taught by Professor Johngay during the Spring '12 term at Washington State University .
- Spring '12