Analysis of Two-Factor Experiments Under an Inverse Gaussian Model
研究了在失效时间服从逆高斯分布时，两因子实验的分析方法，包括参数估计、假设检验和置信区间，并用绝缘材料强度数据演示应用。
Abstract This article treats the analysis of factorial experiments under an inverse Gaussian distribution for the failure times. A reciprocal-linear model for the factor effects is motivated from the context of the underlying Wiener process. Explicit solutions to the likelihood equations are derived, and important properties such as strong consistency and limiting normality are established. A least squares approach using the reciprocals of the sample cell means is also studied and compared with the maximum likelihood method. Other aspects of the investigation include likelihood ratio tests, an analysis-of-reciprocals analogue of the usual normal theory analysis of variance, and confidence intervals for contrasts. An application of the procedures is illustrated with a data set of strength measurements of an insulating material.