lecture 18 - Lecture Goals Lecture 18 Bias and Confounding...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Lecture 18 Bias and Confounding Lecture Goals { Define the difference between bias and confounding { Define different types of bias and confounding { Describe how bias and confounding can be controlled To Show Cause We Use { Koch’s Postulates for Infectious Disease { Hill’s Postulates for Chronic Disease and Complex Questions z Strength of Association z Biologic Credibility z Specificity z Consistency with Other Associations z Time Sequence z Dose-Response Relationship z Analogy z Experiment z Coherence To Show a Valid Statistical Association { We need to assess: z Bias : whether systematic error has been built into the study design z Confounding : whether an extraneous factor is related to both the disease and the exposure z Role of chance : how likely is it that what we found is a true finding (next lecture) Bias { Systematic error in the study design { Results in the incorrect estimation of the association between exposure and outcome { Two classes of bias: { Selection bias { Information bias Selection Bias { Error in identifying the study population { The way in which individuals are chosen results in an apparent association that is different from the true association between exposure and outcome { Subcategories: z Response Bias z Exclusion Bias z Berksonian Bias
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 Response Bias { Are people who participate different than those who refuse to participate? { Volunteers may be different from those who are enlisted { Impacts the validity of the study Exclusion Bias { Occurs when eligibility criteria for cases is different than criteria for controls z i.e. Some medical conditions may be acceptable in cases but not in controls { If methods of selection are related to exposure status, then bias will result Berksonian Bias { Cases and controls are selected from a hospital population { Cases and controls may have different characteristics than others with the outcome { Reason for admission { Length of hospital stay { Hospital ward (oncology, intensive care, pediatrics) { ER vs. in-patient Ins ance stat s Information Bias { Systematic error in the measurement of outcome and/or exposure { Inadequate means for obtaining information about study subjects { Inaccurate data = inaccurate measures of association Types of Information Bias { Recall bias { Interviewer bias { Observer bias { Surveillance bias { Hawthorne effect
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 04/06/2008 for the course PUBLIC HEA 832:335 taught by Professor Schneider during the Spring '08 term at Rutgers.

Page1 / 6

lecture 18 - Lecture Goals Lecture 18 Bias and Confounding...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online