General Matching Function for Multiple Imputation Data

This function facilitates propensity score matching on datasets, including those generated through multiple imputation, to assess covariate balance across treatment groups. It leverages the WeightIt and MatchThem packages to calculate propensity scores and apply matching techniques. Users are referred to these packages for detailed methodological specifications and additional functionality.

Usage

match_mi_general(
  data,
  X,
  baseline_vars,
  estimand,
  method,
  subgroup = NULL,
  focal = NULL,
  sample_weights = NULL
)

Arguments

data: A data frame or a 'mids' object containing the data for matching.
X: A string naming the treatment or exposure variable within `data`.
baseline_vars: A character vector naming the covariates to include in the propensity score model.
estimand: A string specifying the target estimand ('ATE', 'ATT', or 'ATC').
method: The method for estimating propensity scores, per the WeightIt package's documentation.
subgroup: (Optional) A string specifying a variable by which to subgroup the data for within-group matching.
focal: (Optional) Specifies the focal treatment group, useful for 'ATT' or 'ATC' estimands.
sample_weights: (Optional) A string indicating the variable in `data` representing sample weights.
stabilize: Logical. Indicates whether to stabilize weights, defaulting to TRUE.

Value

A list of matched datasets for each subgroup level (if `subgroup` is used) or a single matched dataset. Each element is typically a data frame or an object reflecting the matched data structure, depending on the matching method applied.

References

Detailed methodology and additional options can be found in: - WeightIt package for propensity score estimation. - MatchThem package for matching within imputed datasets.

Examples

if (FALSE) { # \dontrun{
# Assume `df` is a data frame with treatment, covariates, and sample weights
matched_data <- match_mi_general(data = df,
                                 X = "treatment_var",
                                 baseline_vars = c("covariate1", "covariate2"),
                                 estimand = "ATE",
                                 method = "nearest",
                                 stabilize = TRUE)
} # }