Confidence Interval Methods for Parameter Estimation of the Odoma Distribution with Engineering Application
Science & Technology Asia · 30.06.2026 · 29-44
DOI: 10.70730/sta.v31i2.261836 ↗Kurz gefasst
Fachartikel in der Zeitschrift Science & Technology Asia, online erschienen am 30.06.2026. Verfasst von Panicha Kaskasamkul und 1 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Four confidence intervals (CIs) for parameter estimation in the Odoma distribution, which is used in lifetime data analysis, were developed and evaluated in this study. These included the likelihood-based, Wald-type, bootstrap-t, and bias-corrected and accelerated (BCa) bootstrap methods. We used a Monte Carlo simulation study and a real data set to compare the CIs. The criteria in CI comparison are based on their empirical coverage probability (ECP) and average length (AL) in various situations. The explicit formula of the Wald type CI was derived in this study. The simulation results indicated that the ECPs of the likelihood-based and Wald-type CIs approached the nominal confidence level of 0.95 in almost all cases. In cases of small sample sizes (n = 10, 20, or 30), the bootstrap-t and BCa bootstrap CIs yielded ECPs below 0.95. With the increase of sample sizes, the ECPs of the bootstrap-t and BCa bootstrap CIs converged to the nominal confidence level, while the parameter values also influenced the ECP. For low parameter values, the ECPs of the likelihood-based and Wald-type CIs approached the nominal confidence level of 0.95. On the other hand, the ECPs for the bootstrap-t and BCa bootstrap CIs showed low coverage at high parameter values and small sample sizes. The performance of these proposed CIshas been confirmed by their application to the strength data of aircraft window glass, with results corresponding with those from the simulation study.
Zitieren
APA
Kaskasamkul, P., & Panichkitkosolkul, W. (2026). Confidence Interval Methods for Parameter Estimation of the Odoma Distribution with Engineering Application. Science & Technology Asia, 29-44. https://doi.org/10.70730/sta.v31i2.261836
Vancouver
Kaskasamkul P, Panichkitkosolkul W. Confidence Interval Methods for Parameter Estimation of the Odoma Distribution with Engineering Application. Science & Technology Asia. 2026:29-44. doi:10.70730/sta.v31i2.261836
BibTeX
@article{Kaskasamkul2026,
author = {Kaskasamkul, Panicha and Panichkitkosolkul, Wararit},
title = {Confidence Interval Methods for Parameter Estimation of the Odoma Distribution with Engineering Application},
journal = {Science & Technology Asia},
year = {2026},
pages = {29--44},
doi = {10.70730/sta.v31i2.261836}
}
EUVIMED