Quantifying the Uncertainty in Optimal Experiment Schemes via Monte-Carlo Simulations
In the process of designing life-testing experiments , experimenters always establish the optimal experiment scheme based on a particular parametric lifetime model. In most applications, the true lifetime model is unknown and need to be specified for the determination of optimal experiment schemes. Misspecification of the lifetime model may lead to a substantial loss of efficiency in the statistical analysis. Moreover, the determination of the optimal experiment scheme is always relying on asymptotic statistical theory. Therefore, the optimal experiment scheme may not be optimal for finite sample cases. This chapter aims to provide a general framework to quantify the sensitivity and uncertainty of the optimal experiment scheme due to misspecification of the lifetime model. For the illustration of the methodology developed here, analytical and Monte-Carlo methods are employed to evaluate the robustness of the optimal experiment scheme for progressive Type-II censored experiment under the location-scale family of distributions .
KeywordsObjective Function Asymptotic Variance Fisher Information Matrix Model Misspecification Lifetime Distribution
- Chan, P. S., Balakrishnan, N., So, H. Y., & Ng, H. K. T. (2016). Optimal sample size allocation for multi-level stress testing with exponential regression under Type-I censoring. Communications in Statistics—Theory and Methods, 45, 1831–1852.Google Scholar
- Johnson, N. L., Kotz, S., & Balakrishnan, N. (1994) Continuous univariate distribution (Vol. 1, 2nd ed.). New York: WileyGoogle Scholar
- Ka, C. Y., Chan, P. S., Ng, H. K. T., & Balakrishnan, N. (2011). Optimal sample size allocation for multi-level stress testing with extreme value regression under Type-II Censoring, Statistics, 45, 257–279.Google Scholar
- R Core Team. (2016). A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical ComputingGoogle Scholar
- Viveros, R., & Balakrishnan, N. (1994). Interval estimation of parameters of life from progressively censored data. Technometrics, 36, 84–91.Google Scholar