Calculate pooled variance for related binomial experiments with confidence intervals

This method calculates pooled variance for related binomial experiments with confidence intervals. It relies on the normal approximation / Central Limit Theorem to develop a confidence interval each binomial probability in the vector p_values. In order for this assumption to work well, n should be large. The method adjusts the calculated binomial variance to account for the normal approximation confidence interval, and then calculates pooled variance across all experiments.

Usage

BinomVarNorm(n, p_values, alpha = 0.05)

Arguments

n

integer. Number of trials per experiment.

p_values

numeric. Probability of success for each experiment.

alpha

numeric. Probability level for Z-value in normal distribution confidence interval calculation.

Examples

n <- c(10, 5, 30)
p_values <- c(0.2, 0.3, 0.4)
alpha <- 0.05

BinomVarNorm(n, p_values, alpha)
#> [1] 12.04754