Parametric and nonparametric confidence intervals for estimating the difference of means of two skewed populations
In this paper, we have reviewed and proposed several interval estimators for estimating the difference of means of two skewed populations. Estimators include the ordinary-<italic>t</italic>, two versions proposed by Welch [17] and Satterthwaite [15], three versions proposed by Zhou and Dinh [18], Johnson [9], Hall [8], empirical likelihood (EL), bootstrap version of EL, median <italic>t</italic> proposed by Baklizi and Kibria [2] and bootstrap version of median <italic>t</italic>. A Monte Carlo simulation study has been conducted to compare the performance of the proposed interval estimators. Some real life health related data have been considered to illustrate the application of the paper. Based on our findings, some possible good interval estimators for estimating the mean difference of two populations have been recommended for the researchers.
Year of publication: |
2013
|
---|---|
Authors: | Kibria, B. M. Golam ; Banik, Shipra |
Published in: |
Journal of Applied Statistics. - Taylor & Francis Journals, ISSN 0266-4763. - Vol. 40.2013, 12, p. 2617-2636
|
Publisher: |
Taylor & Francis Journals |
Saved in:
Saved in favorites
Similar items by person
-
Testing for Seasonal Stability in Unemployment Series : International Evidence
Banik, Shipra, (1999)
-
Testing for Seasonal Stability in Unemployment Series: International Evidence
Banik, Shipra, (1999)
-
Finite Sample properties of Seasonal Integration Tests
Banik, Shipra, (1998)
- More ...