论文信息 - Use of Probability Weighted Moments in the Analysis of Means

Use of Probability Weighted Moments in the Analysis of Means

In the analysis of means the estimate of σ is usually calculated using range. In this study it is proposed that the linear estimate of σ based on probability weighted moments should be used. A simulation study based on 1000 values suggested that the linear estimate of σ using probability weighted moments is more close to the original value. In the simulations different distributions such as normal, t, Cauchy and exponential were assumed.

[1] H. Scheffé. The relation of control charts to analysis of variance and chi-square tests. , 1947, Journal of the American Statistical Association.

[2] Peter R. Nelson,et al. Multivatiate normal and t distributions with þjk=ajkajk , 1982 .

[3] G. Barrie Wetherill,et al. Sampling Inspection and Quality Control, 2nd edn. , 1978 .

[4] J. H. Sheesley. Simplified Factors for Analysis of Means When the Standard Deviation is Estimated from the Range , 1981 .

[5] J. R. Wallis,et al. Probability Weighted Moments: Definition and Relation to Parameters of Several Distributions Expressable in Inverse Form , 1979 .

[6] Peter R. Nelson,et al. Power Curves for the Analysis of Means , 1985 .

[7] F. Downton,et al. Linear estimates with polynomial coefficients. , 1966, Biometrika.

[8] L. S. Nelson. Exact Critical Values for Use with the Analysis of Means , 1983 .

[9] E. S. Pearson. Biometrika tables for statisticians , 1967 .

[10] F. Mosteller,et al. Understanding robust and exploratory data analysis , 1985 .

[11] E. S. Pearson,et al. Charts of the power function for analysis of variance tests, derived from the non-central F-distribution. , 1951, Biometrika.

[12] Ralph B. D'Agostino. Linear Estimation of the Normal Distribution Standard Deviation , 1970 .

[13] John P. Hoyt. Two Instructive Examples of Maximum Likelihood Estimates , 1969 .