STAB22H3 Lecture Notes - Lecture 7: Fetal Alcohol Spectrum Disorder, Particulates, Preterm Birth

38
STAB22H3 Full Course Notes
Verified Note
38 documents
Document Summary
Observational study - observe individual & measure variable of interest but x influence response e. g. framingham heart study -> find trend/correlation/ association x cause & effect => 1. Retrospective study: collect data after events i. e. look back -> +ve vs -ve = easier to obtain enough data for rare event. & less time consuming vs more confounding and bias cuz select subjects w/ observation that confirms hypothesis 2. Experiment - impose treatment @ individual -> observe response e. g. lady testing tea -> find cause & effect; even if randomly ch subjects -> gp large enough = average controlled factors -> x affect by third variable. Response variable: output of interest -> at least 1 response variable measured. Explanantory variable: input, variable that experimenter thk might help to explain value of response variable -> experimenter manipulate. Blinding: avoid bias result from knowing what treatment was assigned. Single-blind - every individual @ either 1 class x know what treatment was assigned.