Abstract
The purpose of this study is to investigate the influences of sample size n, level interval d and number of levels L on the distribution properties of the estimates, that is, the mean μ and the standard deviation σ, of the fatigue enderance limits obtained by Probit and staircase methods, by Monte Carlo simulation tequnique.
Major conclusions are summarized as follows;
(1) The distribution of the sample mean μ estimated by both methods agrees well with the normal distribution.
(2) The distribution of the sample standard deviation σ can be approximated by the log-normal distribution.
(3) There is little influence of the level interval d and the number of levels L on the mean μ of the sample mean μ by Probit method, and μ is in the range of ±0.01 in all cases, when the population mean is zero and its standard deviation is unity. This is also the case of μ by staircase method.
(4) The mean σ of the sample standard deviation σ approaches the population value of unity from the larger side (σ>1) with an increase in sample size n in the case of Probit method, but σ approaches to unity from the lower side (σ<1) with an increase in n in the case of staircase method.
