Skip to contents

Compute the standard error of a proportion

Source: R/stats.R
se_prop.Rd

Computes the standard error of a proportion for a logical or numeric vector.

Usage

se_prop(
  x,
  na.rm = FALSE,
  min_var = 5,
  low_var_action = c("warn", "ignore", "error")
)

Arguments

x

a logical or numeric vector. If numeric, it must include only 0s, 1s, and/or NAs.

na.rm

logical. Should missing values be removed?

min_var

numeric. Minimum variance (n * p * (1 - p)) for valid normal approximation of the binomial. See Details.

low_var_action

character. Action to take when variance is below min_var.

Details

Standard error of a proportion is calculated using the formula:

$$SE = \sqrt{\frac{p(1 - p)}{n}}$$

This formula assumes that the binomial sampling distribution underlying the observed proportion can be approximated by a normal distribution. This assumption is valid when the proportion variance (np(1 - p)) is sufficiently large, but may not hold when variance is lower. se_prop() therefore issues a warning when the variance is less than 5; this behavior can be overridden using the min_var and low_var_action arguments.