HSS 4303 Lecture Notes - Lecture 9: Endometrial Cancer, Vaginal Bleeding, Cohort Study
Document Summary
Midterm review (6 questions: diabetes test is being applied in a pop of 5000 men. Previous evaluation of the diabetes test in a different pop showed a sensitivity of 70% and a specificity of 80%. Early studies were case-control studies and they indicated a strong harmful effect of estrogen. The controls typically used were women with gynecological cancers other than endometrial cancer. This would yield too high a proportion of estrogen users among cases but a correct estimate of the proportion of estrogen users among controls: twd null if reasoning above is opposite. Prior to computers, generating large numbers of random numbers was a lot of work. Complicated algorithms and practices needed to truly establish a random allocation of a subject to a group. Goals of randomization: eliminate bias (primarily selection bias, balance groups with respect to prognostic variables (both known and uknown i. e confounders, forms basis for statistical tests.