Obtaining point estimators of parameters from confidence intervals or joint confidence regions Article

cited authors

  • Chen, Z

fiu authors

abstract

  • There are several different ways to obtain point estimators for the parameters of a population distribution. The maximum likelihood estimation and moment estimation are the most commonly used ones. Using point estimators only to estimate unknown parameters is somewhat risky because the probability that the estimation is wrong is almost 100%. Interval estimation, on the other hand, can reduce this risk considerably. The purpose of this paper is to propose a new method for obtaining point estimation of parameters. The point estimators discussed here are obtained by squeezing a confidence interval or joint confidence region of the parameters. The proposed method is easy to use in some cases. The estimators obtained by using this method possess some unbiasedness property. It is also shown that the point estimator obtained by this method is more reasonable than the maximum likelihood estimator when the population distribution is skewed. © World Scientific Publishing Company.

publication date

  • February 1, 2007

Digital Object Identifier (DOI)

start page

  • 21

end page

  • 28

volume

  • 14

issue

  • 1