where ν = − n 1 is degrees of freedom of 2 χ distribution. Several authors have carried
out numerical investigations of the accuracy of McKay’s confidence interval. For instance,
Iglewicz and Myers [11] had compared McKay’s confidence interval with the exact
confidence interval based on the noncentral t distribution and they found that McKay’s
confidence interval is efficient for n ≥10 and 0 0.3