HSCI 330 Lecture Notes - Lecture 7: Odds Ratio, Prospective Cohort Study, Melanoma
Document Summary
Information bias: need only to think about the study population. Some participants get put into the wrong box in the 2x2 table for exposure and disease. The problem is caused when we have imperfect knowledge about a participant"s true exposure or disease status. This is called misclassification of exposure or disease status. Sensitivity of exposure classification = 720 / 900 = 0. 80. Specificity of exposure classification = 910 / 1100 = 0. 83. Sensitivity of exposure classification is the probability that someone who is truly exposed will be classified as exposed in the observed data. Specificity of exposure classification is the probability that someone who is truly unexposed will be classified as unexposed in the observed data. Sensitivity of exposure classification = 48 / 60 = 0. 80. Specificity of exposure classification = 126 / 140 = 0. 90. Sensitivity of disease classification is the probability that someone who is truly diseased will be classified as diseased in the observed data.