In a shock test, two kinds of incomplete data often occur. One is an initial incomplete data damaged by first minimum shock. The other is a final incomplete data not damaged by last maximum shock. These incomplete data make it difficult to analyze the fragility of products statistically. Neglecting these incomplete data provides worse accuracy in statistical analysis. It has already been demonstrated that the Johnson’s method and the improved Johnson’s method are applicable to a shock test result including the final incomplete data or the initial incomplete data. This paper presents a new method to be able to analyze statistically even a shock test result including both the initial incomplete data and the final incomplete data. The new method is to sum up the expectation values of the existence about all data, and enables to calculate order numbers which it is difficult to calculate using the ordinary method. It is confirmed that the accuracy of statistics is improved by applying this method to shock test results of eggs.

View full abstract