Statistical Inference for Quantiles of Two-Parameter Gamma Distribution: A Generalized Approach

• Malwane Ananda

  English

  Author

Hydrology, meteorology, environmental monitoring, longevity testing, and dependability are just a few of
the numerous fields that make use of the gamma distribution for data analysis. The quantiles of a two-parameter
gamma distribution are the focus of this statistical investigation. Particularly in domains like life testing and flood
frequency analysis, gamma quantile testing and estimate are necessary. Statistical inference methods included in
published works on the subject are all approximation techniques. Our research suggests two approaches to this issue.
An accurate statistical inference strategy based on the generalized p-value methodology is the first approach. The
method relies on precise probability assertions instead of approximations, making it accurate. The second method
follows the same principle as the parametric bootstrap method. We evaluate the suggested approaches using various
real-world data sets and contrast the outcomes with those of competing approaches. To evaluate how well the
suggested approaches work in comparison to other ways that are already out there, a brief simulation study is provided.
Whether we're talking about lower or higher quantiles, the simulation results show that these two new approaches
outperform the other current ones in terms of size and power.

• pdf