Subgroup Analysis

Click on a button corresponding to a subgroup analysis method. Refer to the table below for guidance.

Process	Choose this process for...
Interaction Trees	Finds a subgroup of rows with a differentiated treatment effect by testing interactions of predictor variables with a treatment variable, splitting on the best interaction effect, and keeping only the best branch of the tree after each split.
Local Control	Estimates treatment effects within clusters of subjects. You can use this process interactively to select the number of clusters and cumulatively compare results.
Optimal Treatment Regime	Fitting both a response regression model and a propensity score logistic model to the input data set and computing a contrast response that can be modeled or used directly to assign optimal treatment.
Virtual Twins	Fits a model to counter-factual data to find a subgroup of rows with enhanced treatment effects

See the JMP Genomics Starter main page for other process categories.