Advances and Applications in Statistics
Volume 39, Issue 2, Pages 95 - 108
(April 2014)
|
|
FIXED REGRESSORS IN BOOTSTRAP REGRESSION
Mezbahur Rahman, Faysal Ibne Satter and Han Wu
|
Abstract: In regression analysis when the samples are small, model assumptions are often violated, a nonparametric method is implemented. Bootstrap estimates are often used in estimating the standard errors and in estimating the sampling distributions of the estimators in making inferences. Bootstrap sampling can be performed for fixed and random regressors. Here we investigate several commonly used confidence intervals for regression parameters for both the options fixed and random regressors. Four different fairly small data sets are used in this study. |
Keywords and phrases: asymptotic normal theory interval, bias corrected interval, bootstrap percentile interval, normal theory interval, standardized bootstrap interval, studentized bootstrap interval. |
|
Number of Downloads: 364 | Number of Views: 815 |
|