Multivalued Treatment Effects

The term multivalued treatment effects broadly refers to a collection of population parameters that capture the impact of a given treatment assigned to each observational unit, when this treatment status takes multiple values. In general, treatment levels may be finite or infinite as well as ordinal or cardinal, leading to a large collection of possible treatment effects to be studied in applications. When the treatment effect of interest is the mean [Page 856]outcome for each treatment level, the resulting population parameter is typically called the dose–response function in the statistical literature, regardless of whether the treatment levels are finite or infinite. The analysis of multivalued treatment effects has several distinct features when compared with the analysis of binary treatment effects, including the following: (a) A ...