Balance Assessment for Matched Data

Comprehensive balance assessment for matched treatment and control groups including standardized mean differences, variance ratios, and distributional balance tests.

Usage

assess_balance(
  matched_data,
  treatment_var = "is_treated",
  balance_variables,
  weight_var = NULL,
  thresholds = list(mean_diff = 0.1, variance_ratio = 2)
)

Arguments

matched_data

Data.table with matched observations

treatment_var

Character. Name of treatment indicator variable. Default: "is_treated"

balance_variables

Character vector. Variables to assess balance for

weight_var

Character. Name of weight variable (for weighted balance). Default: NULL

thresholds

Named list. Balance thresholds: list(mean_diff = 0.1, variance_ratio = 2)

Value

A list containing:

balance_table

Data.table with balance statistics for each variable

overall_balance

Overall balance summary

distributional_tests

Results of distributional balance tests

balance_plots

Data for creating balance plots

Examples

if (FALSE) { # \dontrun{
balance_results <- assess_balance(
  matched_data = ps_match$matched_data,
  balance_variables = c("age", "education", "wage", "sector"),
  thresholds = list(mean_diff = 0.1, variance_ratio = 2)
)
} # }