Approximate confidence interval for standard deviation of nonnormal distributions

The exact confidence interval for σσ is hypersensitive to minor violations of the normality assumption and its performance does not improve with increasing sample size. An approximate confidence interval for σσ is proposed and is shown to be nearly exact under normality with excellent small-sample properties under moderate nonnormality. The small-sample performance of the proposed interval may be further improved using prior kurtosis information. A sample size planning formula is given.