= Causal Inference = '''Causal inference''' is an experimental design used to isolate causation and then make use of predictive statistics. <> ---- == Causation == To study the impact of some '''treatment variable''' on an '''outcome variable''', all other variables must be held constant. If a relationship can be observed ''ceteris paribus'', then causation has been identified. ---- == Experimental Design == A true causal experiment involves the application of a treatment. By comparing outcomes between an experimental group and an identical control group, causation can be observed and measured. === In Social Science === The unit of measurement in social science is almost always ''people''. It isn't possible to source ''identical'' people for application of a treatment. As a result, true causal experiments are impossible in social sciences. Instead, '''statistical inference''' is applied. A tolerance for Type I error ''(false negatives)'' is set as a '''critical value'''. Experimental data is measured and, if the [[Statistics/TestStatistic|test statistic]] is greater than the critical value, the null hypothesis can be rejected. === Given Ethics === There are non-negligible ethical barriers to conducting many experiments when the unit of measurement is people. This is mitigated only prohibitively, i.e. by use of '''Institutional Review Boards''' ('''IRBs'''). In other words, the body of causal experiments is limited by what is IRB approve-able. ---- == Observational Study == Observational studies allow subjects to make decisions regarding the application of a 'treatment'. This absolves most ethical concerns, especially in social science. But it is no longer necessarily true that the experimental and control groups are random and comparable. There are multiple non-redundant '''confounders''' which are related to the treatment and/or the outcome. These must be controlled for. ---- CategoryRicottone