Calculate Balance Score

This function calculates a covariate balance score by calculating (by default) the harmonic mean of the p-values from individual covariate balance tests. This does not require the assumptions that the covariates are independent.

Usage

calculate_balance_score(
  p_values,
  na.rm = TRUE,
  balance_metric = "harmonic_mean"
)

Arguments

p_values

A numeric vector containing p-values from individual covariate balance tests. All values must be between 0 and 1.

na.rm

Logical; if TRUE (default), NA values are removed before calculation. If FALSE and NA values are present, the function will return NA.

balance_metric

A character string specifying the metric to use for calculating the balance score. Valid options are "harmonic_mean" (default) and "product".

Value

A numeric value representing the overall balance score. Higher values indicate better covariate balance.

Examples

# Example with well-balanced covariates
p_vals <- c(0.8, 0.9, 0.85)
calculate_balance_score(p_vals)
#> [1] 0.848037

# Example with poorly balanced covariates
p_vals <- c(0.01, 0.02, 0.03)
calculate_balance_score(p_vals)
#> [1] 0.01636364

# Handling NA values
p_vals_with_na <- c(0.8, NA, 0.85)
calculate_balance_score(p_vals_with_na, na.rm = TRUE)
#> [1] 0.8242424