COMPSCI C8 Lecture Notes - Lecture 3: Decimal Mark, Observational Study, Confounding
Document Summary
Group by some treatment and measure some outcome. Simplest setting: a treatment group and a control group. If the outcome differs (cid:271)et(cid:449)ee(cid:374) these t(cid:449)o groups, that"s e(cid:448)ide(cid:374)(cid:272)e of a(cid:374) association (or relation: e. g. , the top-tier chocolate eaters died of heart disease at a lower rate (12%) than chocolate abstainers (17%) If the two groups are similar in all ways but the treatment, a difference in the outcome is also evidence of causality. If the treatment and control groups have systematic difference others than the treatment itself, then it might be difficult to identify a causal link. When these systematic differences lead researchers astray, they are called confounding factors. Such differences are often present in observational studies: observational study: the researcher does not choose which subjects receive the treatment, controlled experiment: the researcher designs a procedure for selecting the treatment and control groups. When subjects are split up randomly, it"s u(cid:374)likely that there (cid:449)ill (cid:271)e syste(cid:373)ati(cid:272) differences between the groups.