Deciding Which Estimator to Use

Which model you should use depends on what you are trying to model and the type of data you have. The table below can serve as a useful reference when deciding which model to use for a given dataset and causal question.

ModelStructCausal EstimandsSupported Treatment TypesSupported Outcome Types
Interrupted Time Series AnalysisInterruptedTimeSeriesATE, Cumulative Treatment EffectBinaryBinary, Continuous
G-computationGComputationATE, ATT, ITTBinaryBinary, Continuous, Time to Event
Double Machine LearningDoubleMachineLearningATEBinary, Count, Categorical, ContinuousContinuous
S-learningSLearnerCATEBinaryBinary, Continuous, Count
T-learningTLearnerCATEBinaryBinary, Continuous
X-learningXLearnerCATEBinaryBinary, Continuous, Count
R-learningRLearnerCATEBinary, Count, Categorical, ContinuousContinuous