Advances and Applications in Statistics
Volume 2, Issue 3, Pages 311 - 322
(December 2002)
|
|
ASYMPTOTIC AND BOOTSTRAP METHODOLOGIES ON THE ESTIMATION OF THE SONOROUS SENSATION
M. Salicru (Spain), S. Vives (Spain), J. Sanchez (Spain) and F. Oliva (Spain)
|
Abstract: While the last two decades
have seen significant reductions in the noise
produced by all sorts of vehicles, the rapid
growth in transport – particularly air and road
– has resulted in well over hundred million
people being exposed to noise levels above 55 dB
on the front facade of their houses. With the
aim of improving the accuracy in the estimation
– both point estimation and confidence intervals
– of the sonorous sensation, we propose several
improvements based on asymptotic and on
bootstrap methodologies. In order to evaluate
the different approaches, we have performed a
simulation study. In summary, the estimator and
the confidence interval based on the application
of the d -method
seems to be the best option, because it achieves
a good compromise among coverage, precision and
computational effort. |
Keywords and phrases: sonorous sensation, confidence intervals, d -method, Monte Carlo simulation, bootstrap methodology. |
|
Number of Downloads: 6 | Number of Views: 1210 |
|