powered by
The sample range R is a biased estimator for the population standard deviation. For a normally distributed sample of size n, R/d2(n) is unbiased.
R
n
R/d2(n)
d_2(n)
the sample size
a numeric value representing the multiplier to render R unbiased
Montgomery, D.C. (2000) Introduction to Statistical Quality Control, Wiley
d_3
# NOT RUN { x <- rnorm(10, sd = 3) R <- diff(range(x)) R/d_2(10) # expected value of this is 3 # }
Run the code above in your browser using DataLab