Advances and Applications in Statistics
Volume 48, Issue 3, Pages 185 - 204
(March 2016) http://dx.doi.org/10.17654/AS048030185 |
|
A NOTE ON PLANNING THE SIMPLE STEP-STRESS ACCELERATED LIFE TEST UNDER BAYESIAN METHOD AND TYPE-I RIGHT CENSORING
Saleem Z. Ramadan
|
Abstract: The objective of this study is to investigate the effects of the simple step-stress accelerated life test parameters under Bayesian method on the accuracy of the test. The parameters which investigate and affect the test accuracy are: the low stress level, the high stress level, the number of test units, the stress changing time, and the total test time. The effect of the prior information strength has also been considered.
Three sets of experiments were conducted, namely: multivariate linear regression, two-sample t-test of hypotheses for the difference between two means, and comparative analysis to assess the effects of the five test parameters on the test accuracy. The results found show that increase in either total test time or stress changing time or decreasing the low stress level increases the accuracy of the test regardless of the prior strength. In addition, it is obtained that increase in either the number of test units or the high stress level increases the test accuracy when the priors are vague while they have no significant effect when the priors are informative. Furthermore, the results emphasize the importance of involving all the test parameters in planning the SSALT. |
Keywords and phrases: simple step-stress accelerated life test, Bayesian statistics, multivariate-linear regression. |
|
Number of Downloads: 423 | Number of Views: 1286 |
|