Skip to contents

Confidence interval for a standard deviation

Source: R/statpsych1.R
ci.sd.Rd

Computes the traditional confidence interval for a population standard deviation using a sample estimate of the standard deviation. The traditional confidence interval assumes normality and is hypersensitive to minor violations of this assumption. This function should be used only if the data appear to come from an approximate normal or mildly platykurtic distribution.

Usage

ci.sd(alpha, sd, n)

Arguments

alpha

alpha value for 1-alpha confidence

sd

estimated standard deviation

n

sample size

Value

Returns a 1-row matrix. The columns are:

  • Estimate - estimated standard deviation (from input)

  • LL - lower limit of the confidence interval

  • UL - upper limit of the confidence interval

Examples

ci.sd(.05, 4.65, 50)
#>  Estimate       LL      UL
#>      4.65 3.884303 5.79452

# Should return:
#  Estimate       LL       UL
#      4.65 3.884303  5.79452