Corrección a Valores críticos de 22 variantes de pruebas de discordancia para los datos desviados en muestras normales con tamaños de hasta 100 y aplicaciones en las ciencias e ingenierías [Rev. Mex. Cienc. Geol., 23 (2006), 302-319]

Resumen

We had not clarified that in the above mentioned paper as well as two later papers [Rev. Mex. Cienc. Geol. 25 (2008): 82-96 and 369-381], the critical values reported for test N14 (skewness or third moment test) were of the one-sided type. They were simulated from the actual value of the statistic TN14 and repetitions of 100,000 in 2006 or 1,000,000 in 2008. Using the absolute value of the statistic TN14, we have recently simulated more precise and accurate two-sided critical values (2,000,000 repetitions) for confidence levels of 70%, 80%, 90%, 95%, 98%, 99%, and 99.5% (or equivalently, significance level α of 0.30, 0.20, 0.10, 0.05, 0.02, 0.01, and 0.005, as well as for sample sizes of 5(1)100(5)200(10)500(20)1000(50)1500(100)2000(500)5000(1000)10000(5000)30000. These two-sided critical values are available along with their individual error estimates as txt or Excel files, to anyone interested on request to any of the authors.