Mohamed A. Mahmoud, G. Robin Henderson, Eugenio K. Epprecht, William H. Woodall
A multiple of the sample range is often used instead of the sample standard deviation when estimating the standard deviation of a normally distributed random variable because of the range’s computational simplicity. Although many sources argue that the loss of efficiency when using small sample sizes is minor, this study finds previously published tables of relative efficiencies are either mislabeled or inaccurate and advises against using range-based methods. Recommendations are made for situations when a number or samples are taken over time.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados