Advances and Applications in Statistics
Volume 51, Issue 5, Pages 337 - 350
(November 2017) http://dx.doi.org/10.17654/AS051050337 |
|
COMPARISONS OF INTERVAL ESTIMATORS FOR THE LINEAR FUNCTION OF VARIANCES UNDER NON-NORMALITY
Lampang Saenchan
|
Abstract: This paper considers confidence intervals for the linear function of variances under non-normality. Use of the method of variance estimates recovery (the MOVER) and its application is based on the modification of Bonett’s non-normal variance method. Simulated coverage probabilities use different confidence interval methods which are computed for a variety of sample sizes and distributions. The results indicate that this approach outperforms other approximation methods in most cases of study. |
Keywords and phrases: linear combination, interval estimator, the MOVER. |
|
Number of Downloads: 401 | Number of Views: 1269 |
|