A new confidence interval in measurement error model with the reliability ratio known |
| |
Authors: | Liang Yan |
| |
Affiliation: | School of Mathematics and Statistics, Beijing Institute of Technology, Beijing, China |
| |
Abstract: | For the slope parameter of the measurement error model with the reliability ratio known, this article constructs a fiducial generalized confidence interval (FGCI) which is proved to have correct asymptotic coverage. Simulation results demonstrate that the FGCI often outperforms the existing intervals in terms of empirical coverage probability, average interval length, and false parameter coverage rate. Two examples are also provided to illustrate our approach. |
| |
Keywords: | Correct asymptotic coverage false parameter coverage rate fiducial generalized confidence interval measurement error model. |
|
|