estmean.multilevel.subgroup: Estimate the ATE of the mean difference in one multilevel subgroup defined by the proportions

Description

Scores are adjusted to the opposite sign if higher.y == FALSE; scores stay the same if higher.y == TRUE; this is because subgroups defined in estmean.multilevel.subgroup() start from the lowest to the highest adjusted scores, and higher adjusted scores should always represent high responders of trt=1

Usage

estmean.multilevel.subgroup(
  y,
  x.cate,
  x.ps,
  trt,
  score,
  higher.y,
  prop,
  ps.method = "glm",
  minPS = 0.01,
  maxPS = 0.99
)

Value

estimated ATEs of all categories in the one multilevel subgroup; vector of size equal to the length of categories in the multilevel subgroup

Arguments

y: Observed outcome; vector of size n (observations)
x.cate: Matrix of p.cate baseline covariates; dimension n by p.cate (covariates in the outcome model)
x.ps: Matrix of p.ps baseline covariates (plus a leading column of 1 for the intercept); dimension n by p.ps + 1 (covariates in the propensity score model plus intercept)
trt: Treatment received; vector of size n units with treatment coded as 0/1
score: Estimated CATE scores for all n observations from one of the six methods (boosting, linear regression, two regressions, contrast regression, random forest, generalized additive model); vector of size n
higher.y: A logical value indicating whether higher (TRUE) or lower (FALSE) values of the outcome are more desirable. Default is TRUE.
prop: Proportions corresponding to percentiles in the estimated CATE scores that define subgroups to calculate ATE for; vector of floats in `[0, 1]` always starting with 0 and ending with 1: Each element of prop represents inclusive cutoffs in the multilevel subgroup and the length of prop is number of categories in the multilevel subgroup
ps.method: A character value for the method to estimate the propensity score. Allowed values include one of: 'glm' for logistic regression with main effects only (default), or 'lasso' for a logistic regression with main effects and LASSO penalization on two-way interactions (added to the model if interactions are not specified in ps.model). Relevant only when ps.model has more than one variable.
minPS: A numerical value (in `[0, 1]`) below which estimated propensity scores should be truncated. Default is 0.01.
maxPS: A numerical value (in `(0, 1]`) above which estimated propensity scores should be truncated. Must be strictly greater than minPS. Default is 0.99.